AMD Curve Optimizer any guides / experience

nội dung

Just to update on my latest stability testing. I am now at the following on my 5800x B550-F with 1801 bios:

Core

curve offset w/pbo +0, scaler x2.

0 (perf #3/4)

-15

1 (perf #4/5)

-15

2 (perf #5/6)

-10

3 (perf #1/1)

10

4 (perf #7/8)

-30

5 (perf #6/7)

-20

6 (perf #1/2)

3

7 (perf #2/3)

-5

The above curve was tested +200, scaler x1, with both p95 and testmem5 running on a single core at a time. Compared to my earlier settings, I had to add a little more voltage to cores 5 & 6 after testing with Testmem5 at +200. Also core 7 was not p95 stable at -10. I don't know if cores 5&6 were 100% stable at these settings on p95 either, but what I found was that Testmem5 runs the clocks higher than the p95 setting I have been using since it uses less power. It seems like it is worth testing both. With this curve setting, TM5 is holding ~4900 MHz on most cores. Given that, I have backed off the +200 and put it back to +0. This makes no noticeable difference to the benchmarks I have run so far, and I have more confidence everything is going to remain stable now. I also upped scaler to x2 since that seems to be the default in this BIOS, but I'm still not sure if this setting makes any difference on my configuration. My conclusion here is that AMD has already tuned these things pretty close to the limit out of the box. So there is very little to gain from trying to add higher clock. Adding some curve offset helps with multi-core loads, but the difference is not huge, and it is quite difficult to properly test for stability when you do this.

In short, be careful when setting a curve offset, or even just adding pbo boost override, since it is very difficult to test each core individually for stability at all possible clocks.

harm9963

One last hurrah for bios 3003 before i update to a bios with AMD AM4 AGESA V2 PI 1.2.0.0 and support for Nvidia smart access memory.
Cold air benching with EK custom waterloop+TechN Zen3 waterblock :) Curve optimizer = -30 allcore Stable in everything i throw at it, and no WHEA errors.

View attachment 185204

Cinebench r23 multithread = 32229 points Cinebench r23 singlethread = 1729 points Cinebench r20 multithread = 12441 points Cinebench r20 singlethread = 674 points Cinebench r15 multithread = 5404 points Cinebench r15 multithread = 288 points

CPU-Z validator @ https://valid.x86.fr/dl125q

Some Asus realbench + Passmark performancetest numbers @ https://www.passmark.com/baselines/V10/display.php?id=135921464997 (This machine is ranked #36 out of 156355 results globally)

View attachment 185205
https://www.passmark.com/products/performancetest/index.php
Geekbench 4 @ https://browser.geekbench.com/v4/cpu/16005550 Singlethread = 8215 points Multithread = 74733 points

Geekbench5 @ https://browser.geekbench.com/v5/cpu/6050938

Singlethread = 1844 points Multithread = 20054 points Some heavy IBT high+very high and Y-Cruncher numbers:

View attachment 185207

Did also run a full sweep of all 3dmarks, but i will post that in one other thread :)

[2021/01/20 16:26:21] Ai Overclock Tuner [Manual] BCLK Frequency [100.0000] Memory Frequency [DDR4-3800MHz] FCLK Frequency [1900MHz] Core Performance Boost [Enabled] CPU Core Ratio [Auto] Core VID [Auto] CCX0 Ratio [Auto] CCX0 Ratio [Auto] TPU [Keep Current Settings] Performance Bias [Auto] PBO Fmax Enhancer [Auto] Precision Boost Overdrive [Manual] PPT Limit [300] TDC Limit [235] EDC Limit [245] Precision Boost Overdrive Scalar [Manual] Customized Precision Boost Overdrive Scalar [4X] Max CPU Boost Clock Override [50] Platform Thermal Throttle Limit [Auto] DRAM CAS# Latency [14] Trcdrd [15] Trcdwr [8] DRAM RAS# PRE Time [12] DRAM RAS# ACT Time [24] Trc [36] TrrdS [4] TrrdL [4] Tfaw [16] TwtrS [4] TwtrL [10] Twr [12] Trcpage [Auto] TrdrdScl [2] TwrwrScl [2] Trfc [252] Trfc2 [187] Trfc4 [115] Tcwl [14] Trtp [6] Trdwr [9] Twrrd [2] TwrwrSc [1] TwrwrSd [6] TwrwrDd [6] TrdrdSc [1] TrdrdSd [5] TrdrdDd [5] Tcke [Auto] ProcODT [40 ohm] Cmd2T [1T] Gear Down Mode [Enabled] Power Down Enable [Disabled] RttNom [RZQ/7] RttWr [RZQ/3] RttPark [RZQ/1] MemAddrCmdSetup [Auto] MemCsOdtSetup [Auto] MemCkeSetup [Auto] MemCadBusClkDrvStren [24.0 Ohm] MemCadBusAddrCmdDrvStren [20.0 Ohm] MemCadBusCsOdtDrvStren [24.0 Ohm] MemCadBusCkeDrvStren [24.0 Ohm] Mem Over Clock Fail Count [Auto] Voltage Monitor [Die Sense] CPU Load-line Calibration [Level 3] CPU Current Capability [140%] CPU VRM Switching Frequency [Manual] Fixed CPU VRM Switching Frequency(KHz) [500] CPU Power Duty Control [T.Probe] CPU Power Phase Control [Extreme] CPU Power Thermal Control [120] VDDSOC Load-line Calibration [Level 3] VDDSOC Switching Frequency [Auto] VDDSOC Phase Control [Extreme] DRAM Current Capability [130%] DRAM Power Phase Control [Extreme] DRAM Switching Frequency [Auto] CPU Core Current Telemetry [Auto] CPU SOC Current Telemetry [Auto] Force OC Mode Disable [Disabled] SB Clock Spread Spectrum [Auto] VTTDDR Voltage [Auto] VPP_MEM Voltage [Auto] DRAM CTRL REF Voltage on CHA [Auto] DRAM CTRL REF Voltage on CHB [Auto] VDDP Voltage [Auto] 1.8V Standby Voltage [Auto] CPU 3.3v AUX [Auto] 1.2V SB Voltage [Auto] DRAM R1 Tune [Auto] DRAM R2 Tune [Auto] DRAM R3 Tune [Auto] DRAM R4 Tune [Auto] PCIE Tune R1 [Auto] PCIE Tune R2 [Auto] PCIE Tune R3 [Auto] PLL Tune R1 [Auto] PLL reference voltage [Auto] T Offset [Auto] Sense MI Skew [Auto] Sense MI Offset [Auto] Promontory presence [Auto] Clock Amplitude [Auto] CPU Core Voltage [Offset mode] - Offset Mode Sign [+] - CPU Core Voltage Offset [0.01250] CPU SOC Voltage [Manual] - VDDSOC Voltage Override [1.11875] DRAM Voltage [1.54500] VDDG CCD Voltage Control [0.890] VDDG IOD Voltage Control [Auto] CLDO VDDP voltage [0.880] 1.00V SB Voltage [Auto] 1.8V PLL Voltage [Auto] TPM Device Selection [Discrete TPM] Erase fTPM NV for factory reset [Enabled] PSS Support [Enabled] PPC Adjustment [PState 0] NX Mode [Enabled] SVM Mode [Disabled] SMT Mode [Auto] Core Leveling Mode [Automatic mode] CCD Control [Auto] SATA Port Enable [Enabled] SATA Mode [AHCI] NVMe RAID mode [Disabled] SMART Self Test [Enabled] Hot Plug [Disabled] Hot Plug [Disabled] Hot Plug [Disabled] Hot Plug [Disabled] Hot Plug [Disabled] Hot Plug [Disabled] Hot Plug [Disabled] Hot Plug [Disabled] HD Audio Controller [Enabled] PCIEX16_1 Bandwidth Bifurcation configuration [Auto Mode] PCIEX16_2 Bandwidth Bifurcation configuration [Auto Mode] When system is in working state [All On] Q-Code LED Function [POST Code Only] When system is in sleep, hibernate or soft off states [All On] Realtek 2.5G LAN Controller [Enabled] Realtek PXE OPROM [Disabled] Intel LAN Controller [Enabled] Intel LAN OPROM [Disabled] ASM1074 Controller [Enabled] Wi-Fi 6 (802.11ax) Controller [Disabled] Bluetooth Controller [Enabled] USB power delivery in Soft Off state (S5) [Enabled] PCIEX16_1 Mode [Auto] PCIEX16_2 Mode [Auto] PCIEX1 Mode [Auto] PCIEX16_3 Mode [Auto] M.2_1 Link Mode [Auto] M.2_2 Link Mode [Auto] SB Link Mode [Auto] ErP Ready [Disabled] Restore AC Power Loss [Power Off] Power On By PCI-E [Disabled] Power On By RTC [Disabled] Above 4G Decoding [Enabled] Re-Size BAR Support [Auto] SR-IOV Support [Disabled] Legacy USB Support [Enabled] XHCI Hand-off [Enabled] Corsair Voyager GTX 0 [Auto] USB Device Enable [Enabled] U32G2_2 [Enabled] U32G2_3 [Enabled] U32G2_4 [Enabled] U32G1_10 [Enabled] U32G1_11 [Enabled] USB12 [Enabled] USB13 [Enabled] U32G2_7 [Enabled] U32G2_8 [Enabled] U32G2_C9 [Enabled] Network Stack [Disabled] Device [SATA6G_7: Samsung SSD 850 PRO 1TB] CPU Temperature [Monitor] CPU Package Temperature [Monitor] MotherBoard Temperature [Monitor] VRM Temperature [Monitor] T_Sensor Temperature [Monitor] Water In T Sensor Temperature [Monitor] Water Out T Sensor Temperature [Monitor] CPU Fan Speed [Monitor] CPU Optional Fan Speed [Monitor] Chassis Fan 1 Speed [Monitor] Chassis Fan 2 Speed [Monitor] Chassis Fan 3 Speed [Monitor] High Amp Fan Speed [Monitor] W_PUMP+ Speed [Monitor] AIO PUMP Speed [Monitor] PCH Fan Speed [Monitor] Flow Rate [Monitor] CPU Core Voltage [Monitor] 3.3V Voltage [Monitor] 5V Voltage [Monitor] 12V Voltage [Monitor] CPU Fan Q-Fan Control [Auto] CPU Fan Step Up [2.1 sec] CPU Fan Step Down [0 sec] CPU Fan Speed Low Limit [600 RPM] CPU Fan Profile [Manual] CPU Fan Upper Temperature [70] CPU Fan Max. Duty Cycle (%) [100] CPU Fan Middle Temperature [50] CPU Fan Middle Duty Cycle (%) [50] CPU Fan Lower Temperature [30] CPU Fan Min Duty Cycle (%) [40] Chassis Fan 1 Q-Fan Control [Auto] Chassis Fan 1 Q-Fan Source [CPU] Chassis Fan 1 Step Up [0 sec] Chassis Fan 1 Step Down [0 sec] Chassis Fan 1 Speed Low Limit [600 RPM] Chassis Fan 1 Profile [Manual] Chassis Fan 1 Upper Temperature [70] Chassis Fan 1 Max. Duty Cycle (%) [100] Chassis Fan 1 Middle Temperature [50] Chassis Fan 1 Middle Duty Cycle (%) [65] Chassis Fan 1 Lower Temperature [20] Chassis Fan 1 Min Duty Cycle (%) [60] Chassis Fan 2 Q-Fan Control [Auto] Chassis Fan 2 Q-Fan Source [CPU] Chassis Fan 2 Step Up [0 sec] Chassis Fan 2 Step Down [0 sec] Chassis Fan 2 Speed Low Limit [600 RPM] Chassis Fan 2 Profile [Manual] Chassis Fan 2 Upper Temperature [65] Chassis Fan 2 Max. Duty Cycle (%) [100] Chassis Fan 2 Middle Temperature [45] Chassis Fan 2 Middle Duty Cycle (%) [60] Chassis Fan 2 Lower Temperature [40] Chassis Fan 2 Min Duty Cycle (%) [60] Chassis Fan 3 Q-Fan Control [Auto] Chassis Fan 3 Q-Fan Source [CPU] Chassis Fan 3 Step Up [0 sec] Chassis Fan 3 Step Down [0 sec] Chassis Fan 3 Speed Low Limit [600 RPM] Chassis Fan 3 Profile [Manual] Chassis Fan 3 Upper Temperature [70] Chassis Fan 3 Max. Duty Cycle (%) [100] Chassis Fan 3 Middle Temperature [45] Chassis Fan 3 Middle Duty Cycle (%) [100] Chassis Fan 3 Lower Temperature [40] Chassis Fan 3 Min Duty Cycle (%) [100] High Amp Fan Q-Fan Control [Auto] High Amp Fan Q-Fan Source [CPU] High Amp Fan Step Up [0 sec] High Amp Fan Step Down [0 sec] High Amp Fan Speed Low Limit [600 RPM] High Amp Fan Profile [Manual] High Amp Fan Upper Temperature [70] High Amp Fan Max. Duty Cycle (%) [100] High Amp Fan Middle Temperature [45] High Amp Fan Middle Duty Cycle (%) [70] High Amp Fan Lower Temperature [30] High Amp Fan Min Duty Cycle (%) [60] Water Pump+ Q-Fan Control [Auto] Water Pump+ Q-Fan Source [CPU] Water Pump+ Upper Temperature [70] Water Pump+ Max. Duty Cycle (%) [100] Water Pump+ Middle Temperature [50] Water Pump+ Middle Duty Cycle (%) [65] Water Pump+ Lower Temperature [30] Water Pump+ Min Duty Cycle (%) [60] AIO Pump Q-Fan Control [Auto] AIO Pump Q-Fan Source [CPU] AIO Pump Upper Temperature [70] AIO Pump Max. Duty Cycle (%) [100] AIO Pump Middle Temperature [50] AIO Pump Middle Duty Cycle (%) [65] AIO Pump Lower Temperature [30] AIO Pump Min Duty Cycle (%) [60] Above 4GB MMIO Limit [39bit (512GB)] Fast Boot [Enabled] Next Boot after AC Power Loss [Fast Boot] Boot Logo Display [Disabled] Bootup NumLock State [On] POST Report [5 sec] Wait For 'F1' If Error [Enabled] Option ROM Messages [Force BIOS] Interrupt 19 Capture [Disabled] Setup Mode [Advanced Mode] Launch CSM [Disabled] OS Type [Other OS] AMI Native NVMe Driver Support [Enabled] Flexkey [Reset] Setup Animator [Disabled] Load from Profile [5] Profile Name [20.01 minus 30] Save to Profile [5] DIMM Slot Number [DIMM_A1] Bus Interface [PCIEX16_1] Download & Install ARMOURY CRATE app [Enabled] CPU Frequency [0] CPU Voltage [0] CCD Control [Auto] Core control [Auto] SMT Control [Auto] Overclock [Enabled ] Memory Clock Speed [Auto] Tcl [Auto] Trcdrd [Auto] Trcdwr [Auto] Trp [Auto] Tras [Auto] Trc Ctrl [Auto] TrrdS [Auto] TrrdL [Auto] Tfaw Ctrl [Auto] TwtrS [Auto] TwtrL [Auto] Twr Ctrl [Auto] Trcpage Ctrl [Auto] TrdrdScL Ctrl [Auto] TwrwrScL Ctrl [Auto] Trfc Ctrl [Auto] Trfc2 Ctrl [Auto] Trfc4 Ctrl [Auto] Tcwl [Auto] Trtp [Auto] Tcke [Auto] Trdwr [Auto] Twrrd [Auto] TwrwrSc [Auto] TwrwrSd [Auto] TwrwrDd [Auto] TrdrdSc [Auto] TrdrdSd [Auto] TrdrdDd [Auto] ProcODT [Auto] Power Down Enable [Auto] Cmd2T [Auto] Gear Down Mode [Auto] CAD Bus Timing User Controls [Auto] CAD Bus Drive Strength User Controls [Auto] Data Bus Configuration User Controls [Auto] Infinity Fabric Frequency and Dividers [Auto] ECO Mode [Disable] Precision Boost Overdrive [Advanced] PBO Limits [Motherboard] Precision Boost Overdrive Scalar [Auto] Curve Optimizer [Per Core] Core 0 Curve Optimizer Sign [Negative] Core 0 Curve Optimizer Magnitude [30] Core 1 Curve Optimizer Sign [Negative] Core 1 Curve Optimizer Magnitude [30] Core 2 Curve Optimizer Sign [Negative] Core 2 Curve Optimizer Magnitude [30] Core 3 Curve Optimizer Sign [Negative] Core 3 Curve Optimizer Magnitude [30] Core 4 Curve Optimizer Sign [Negative] Core 4 Curve Optimizer Magnitude [30] Core 5 Curve Optimizer Sign [Negative] Core 5 Curve Optimizer Magnitude [30] Core 6 Curve Optimizer Sign [Negative] Core 6 Curve Optimizer Magnitude [30] Core 7 Curve Optimizer Sign [Negative] Core 7 Curve Optimizer Magnitude [30] Core 8 Curve Optimizer Sign [Negative] Core 8 Curve Optimizer Magnitude [30] Core 9 Curve Optimizer Sign [Negative] Core 9 Curve Optimizer Magnitude [30] Core 10 Curve Optimizer Sign [Negative] Core 10 Curve Optimizer Magnitude [30] Core 11 Curve Optimizer Sign [Negative] Core 11 Curve Optimizer Magnitude [30] Core 12 Curve Optimizer Sign [Negative] Core 12 Curve Optimizer Magnitude [30] Core 13 Curve Optimizer Sign [Negative] Core 13 Curve Optimizer Magnitude [30] Core 14 Curve Optimizer Sign [Negative] Core 14 Curve Optimizer Magnitude [30] Core 15 Curve Optimizer Sign [Negative] Core 15 Curve Optimizer Magnitude [30] Max CPU Boost Clock Override [0MHz] Platform Thermal Throttle Limit [Auto] LN2 Mode [Auto] SoC/Uncore OC Mode [Disabled] VDDP Voltage Control [Auto] VDDG Voltage Control [Auto] NUMA nodes per socket [Auto] Custom Pstate0 [Auto] L1 Stream HW Prefetcher [Auto] L2 Stream HW Prefetcher [Auto] Core Watchdog Timer Enable [Auto] SMEE [Auto] Core Performance Boost [Auto] Global C-state Control [Disabled] Power Supply Idle Control [Typical Current Idle] SEV ASID Count [Auto] SEV-ES ASID Space Limit Control [Auto] Streaming Stores Control [Auto] Local APIC Mode [Auto] ACPI _CST C1 Declaration [Auto] MCA error thresh enable [Auto] PPIN Opt-in [Auto] Fast Short REP MOVSB [Enabled] Enhanced REP MOVSB/STOSB [Enabled] RdRand Speedup Disable [Enabled] IBS hardware workaround [Auto] DRAM scrub time [Auto] Poison scrubber control [Auto] Redirect scrubber control [Auto] Redirect scrubber limit [Auto] NUMA nodes per socket [Auto] Memory interleaving [Auto] Memory interleaving size [Auto] 1TB remap [Auto] DRAM map inversion [Auto] ACPI SRAT L3 Cache As NUMA Domain [Auto] ACPI SLIT Distance Control [Auto] ACPI SLIT remote relative distance [Auto] GMI encryption control [Auto] xGMI encryption control [Auto] CAKE CRC perf bounds Control [Auto] 4-link xGMI max speed [Auto] 3-link xGMI max speed [Auto] xGMI TXEQ Mode [Auto] PcsCG control [Auto] Disable DF to external downstream IP SyncFloodPropagation [Auto] Disable DF sync flood propagation [Auto] CC6 memory region encryption [Auto] Memory Clear [Auto] Overclock [Enabled] Memory Clock Speed [Auto] Tcl [Auto] Trcdrd [Auto] Trcdwr [Auto] Trp [Auto] Tras [Auto] Trc Ctrl [Auto] TrrdS [Auto] TrrdL [Auto] Tfaw Ctrl [Auto] TwtrS [Auto] TwtrL [Auto] Twr Ctrl [Auto] Trcpage Ctrl [Auto] TrdrdScL Ctrl [Auto] TwrwrScL Ctrl [Auto] Trfc Ctrl [Auto] Trfc2 Ctrl [Auto] Trfc4 Ctrl [Auto] Tcwl [Auto] Trtp [Auto] Tcke [Auto] Trdwr [Auto] Twrrd [Auto] TwrwrSc [Auto] TwrwrSd [Auto] TwrwrDd [Auto] TrdrdSc [Auto] TrdrdSd [Auto] TrdrdDd [Auto] ProcODT [Auto] Power Down Enable [Auto] Disable Burst/Postponed Refresh [Auto] DRAM Maximum Activate Count [Auto] Cmd2T [Auto] Gear Down Mode [Auto] CAD Bus Timing User Controls [Auto] CAD Bus Drive Strength User Controls [Auto] Data Bus Configuration User Controls [Auto] Data Poisoning [Auto] DRAM Post Package Repair [Default] RCD Parity [Auto] DRAM Address Command Parity Retry [Auto] Write CRC Enable [Auto] DRAM Write CRC Enable and Retry Limit [Auto] Disable Memory Error Injection [True] DRAM ECC Symbol Size [Auto] DRAM ECC Enable [Auto] DRAM UECC Retry [Auto] TSME [Auto] Data Scramble [Auto] DFE Read Training [Auto] FFE Write Training [Auto] PMU Pattern Bits Control [Auto] MR6VrefDQ Control [Auto] CPU Vref Training Seed Control [Auto] Chipselect Interleaving [Auto] BankGroupSwap [Auto] BankGroupSwapAlt [Auto] Address Hash Bank [Auto] Address Hash CS [Auto] Address Hash Rm [Auto] SPD Read Optimization [Enabled] MBIST Enable [Disabled] Pattern Select [PRBS] Pattern Length [6] Aggressor Channel [1 Aggressor Channel] Aggressor Static Lane Control [Disabled] Target Static Lane Control [Disabled] Worst Case Margin Granularity [Per Chip Select] Read Voltage Sweep Step Size [1] Read Timing Sweep Step Size [1] Write Voltage Sweep Step Size [1] Write Timing Sweep Step Size [1] IOMMU [Auto] Precision Boost Overdrive [Auto] Precision Boost Overdrive Scalar [Auto] FCLK Frequency [Auto] SOC OVERCLOCK VID [0] UCLK DIV1 MODE [Auto] VDDP Voltage Control [Auto] VDDG Voltage Control [Auto] SoC/Uncore OC Mode [Auto] LN2 Mode [Auto] ACS Enable [Auto] PCIe ARI Support [Auto] PCIe ARI Enumeration [Auto] PCIe Ten Bit Tag Support [Auto] cTDP Control [Auto] EfficiencyModeEn [Auto] Package Power Limit Control [Auto] APBDIS [Auto] DF Cstates [Auto] CPPC [Auto] CPPC Preferred Cores [Auto] NBIO DPM Control [Auto] Early Link Speed [Auto] Presence Detect Select mode [Auto] Preferred IO [Auto] CV test [Auto] Loopback Mode [Auto]

Data Link Feature Exchange [Disabled]

Can you test Geek bench 5

Zach_01

The voltage feed to the CPU cores is 1 value alone. There is no capability of individual voltage regulation per core. What the CPU cores are getting is the “CPU Core Voltage (SVI2 TFN)” as reported by HWiNFO sensors mode.
What you alter by curves is speed per core.

New CPU new luck.. Bios: F33a PPT/TDC/EDC 210/124/155 scaler: 1x boost override: +0 Mhz curve optimizer: -30 on 14 cores, -25 on 1 core, -18 on 1 core vcore: normal with a minus (-) offset of 0.08750V LLT: medium SoC LLT: medium Infinity Fabric: 1600 Mhz (Memory 3200 CL14) I was aiming at good enough score with low temperatures and no temperature spikes, here are my results: (Ambient temperature: 24C, arctic liquid freezer ii 360 with arctic silver 5 applied, setup for "maximum" silence at full load, fans around 1000 rpm on radiator, 450 rpm system fans)

1611966429790.png

NOTE:

I have copy/pasted into the screenshot parts of a second screenshot within hwinfo (just scrolled down to show power draw etc.. and copy/pasted that part into main screenshot)

Last edited: Jan 30, 2021

droopyRO

F33a BIOS on my X570 Elite, got to negative 20 on all cores 5600X. At -25 i get blue screen errors, did not fiddle with per core offset yet.
Screenshot (408).jpg

jesdals

On my master F33a was to unstable - did manage to boot with 2000 infinity - but got WHEAs at even 1900 so vent back to F32

droopyRO

Until this madness with inflated prices and little availabilty of computer parts, especially GPUs i will not touch any kind of overclocking. Stock settings and undevolt for me :) So far F33a is stable for me at stock with undervolt.

Last edited: Jan 30, 2021

On my master F33a was to unstable - did manage to boot with 2000 infinity - but got WHEAs at even 1900 so vent back to F32

@jesdals Hi. I'm impressive with your cpu-z results. Could you provide your finally setting to achieve this result? PBO, voltages, llc and others
Thank you ☺️

jesdals

@jesdals Hi. I'm impressive with your cpu-z results. Could you provide your finally setting to achieve this result? PBO, voltages, llc and others
Thank you ☺️

You can see my settings above, theres pics from bios and thats actually settings - did cold settings with higher settings but without better results

Zach_01

Just a suggestion if anyone is willing to follow as a test. If using AGESA V2 1.1.8.0~1.2.0.0 to stop using HWiNFO (not load at all) and see if WHEA/Cache Hierarchy errors or crashes vanish.

Just a discussion over HWiNFO forums brought this up. Maybe users here can confirm... or not.

harm9963

New ASUS ROG E X570 , no PBO , -23 all cores plus only ASUS app, holds clocks longer and higher , temps are cool , no peaky bench mark, 75 degrees room temp , no throttle at all , this will be my day to day, will run for week , get used everything , then will try to break my ASUS X470 Prime Pro numbers.
30 min loop , MB sure makes a difference !

  • Capturenot28.PNG
  • Capturepassno.PNG

jesdals

Just a suggestion if anyone is willing to follow as a test. If using AGESA V2 1.1.8.0~1.2.0.0 to stop using HWiNFO (not load at all) and see if WHEA/Cache Hierarchy errors or crashes vanish.

Just a discussion over HWiNFO forums brought this up. Maybe users here can confirm... or not.

I did find the F33a unstable regardless if HWinfo was running or not - so my experience with Agesa 1.2.0.0 Gigabyte version is that its not stable yet - even at 1933MHz 1:1:1 it wasnt stable.

Det0x

One last hurrah for bios 3003 before i update to a bios with AMD AM4 AGESA V2 PI 1.2.0.0 and support for Nvidia smart access memory.
Cold air benching with EK custom waterloop+TechN Zen3 waterblock :) Curve optimizer = -30 allcore Stable in everything i throw at it, and no WHEA errors.

View attachment 185204

Cinebench r23 multithread = 32229 points Cinebench r23 singlethread = 1729 points Cinebench r20 multithread = 12441 points Cinebench r20 singlethread = 674 points Cinebench r15 multithread = 5404 points Cinebench r15 multithread = 288 points

CPU-Z validator @ https://valid.x86.fr/dl125q

Some Asus realbench + Passmark performancetest numbers @ https://www.passmark.com/baselines/V10/display.php?id=135921464997 (This machine is ranked #36 out of 156355 results globally)

View attachment 185205
https://www.passmark.com/products/performancetest/index.php
Geekbench 4 @ https://browser.geekbench.com/v4/cpu/16005550 Singlethread = 8215 points Multithread = 74733 points

Geekbench5 @ https://browser.geekbench.com/v5/cpu/6050938

Singlethread = 1844 points Multithread = 20054 points Some heavy IBT high+very high and Y-Cruncher numbers:

View attachment 185207

Did also run a full sweep of all 3dmarks, but i will post that in one other thread :)

[2021/01/20 16:26:21] Ai Overclock Tuner [Manual] BCLK Frequency [100.0000] Memory Frequency [DDR4-3800MHz] FCLK Frequency [1900MHz] Core Performance Boost [Enabled] CPU Core Ratio [Auto] Core VID [Auto] CCX0 Ratio [Auto] CCX0 Ratio [Auto] TPU [Keep Current Settings] Performance Bias [Auto] PBO Fmax Enhancer [Auto] Precision Boost Overdrive [Manual] PPT Limit [300] TDC Limit [235] EDC Limit [245] Precision Boost Overdrive Scalar [Manual] Customized Precision Boost Overdrive Scalar [4X] Max CPU Boost Clock Override [50] Platform Thermal Throttle Limit [Auto] DRAM CAS# Latency [14] Trcdrd [15] Trcdwr [8] DRAM RAS# PRE Time [12] DRAM RAS# ACT Time [24] Trc [36] TrrdS [4] TrrdL [4] Tfaw [16] TwtrS [4] TwtrL [10] Twr [12] Trcpage [Auto] TrdrdScl [2] TwrwrScl [2] Trfc [252] Trfc2 [187] Trfc4 [115] Tcwl [14] Trtp [6] Trdwr [9] Twrrd [2] TwrwrSc [1] TwrwrSd [6] TwrwrDd [6] TrdrdSc [1] TrdrdSd [5] TrdrdDd [5] Tcke [Auto] ProcODT [40 ohm] Cmd2T [1T] Gear Down Mode [Enabled] Power Down Enable [Disabled] RttNom [RZQ/7] RttWr [RZQ/3] RttPark [RZQ/1] MemAddrCmdSetup [Auto] MemCsOdtSetup [Auto] MemCkeSetup [Auto] MemCadBusClkDrvStren [24.0 Ohm] MemCadBusAddrCmdDrvStren [20.0 Ohm] MemCadBusCsOdtDrvStren [24.0 Ohm] MemCadBusCkeDrvStren [24.0 Ohm] Mem Over Clock Fail Count [Auto] Voltage Monitor [Die Sense] CPU Load-line Calibration [Level 3] CPU Current Capability [140%] CPU VRM Switching Frequency [Manual] Fixed CPU VRM Switching Frequency(KHz) [500] CPU Power Duty Control [T.Probe] CPU Power Phase Control [Extreme] CPU Power Thermal Control [120] VDDSOC Load-line Calibration [Level 3] VDDSOC Switching Frequency [Auto] VDDSOC Phase Control [Extreme] DRAM Current Capability [130%] DRAM Power Phase Control [Extreme] DRAM Switching Frequency [Auto] CPU Core Current Telemetry [Auto] CPU SOC Current Telemetry [Auto] Force OC Mode Disable [Disabled] SB Clock Spread Spectrum [Auto] VTTDDR Voltage [Auto] VPP_MEM Voltage [Auto] DRAM CTRL REF Voltage on CHA [Auto] DRAM CTRL REF Voltage on CHB [Auto] VDDP Voltage [Auto] 1.8V Standby Voltage [Auto] CPU 3.3v AUX [Auto] 1.2V SB Voltage [Auto] DRAM R1 Tune [Auto] DRAM R2 Tune [Auto] DRAM R3 Tune [Auto] DRAM R4 Tune [Auto] PCIE Tune R1 [Auto] PCIE Tune R2 [Auto] PCIE Tune R3 [Auto] PLL Tune R1 [Auto] PLL reference voltage [Auto] T Offset [Auto] Sense MI Skew [Auto] Sense MI Offset [Auto] Promontory presence [Auto] Clock Amplitude [Auto] CPU Core Voltage [Offset mode] - Offset Mode Sign [+] - CPU Core Voltage Offset [0.01250] CPU SOC Voltage [Manual] - VDDSOC Voltage Override [1.11875] DRAM Voltage [1.54500] VDDG CCD Voltage Control [0.890] VDDG IOD Voltage Control [Auto] CLDO VDDP voltage [0.880] 1.00V SB Voltage [Auto] 1.8V PLL Voltage [Auto] TPM Device Selection [Discrete TPM] Erase fTPM NV for factory reset [Enabled] PSS Support [Enabled] PPC Adjustment [PState 0] NX Mode [Enabled] SVM Mode [Disabled] SMT Mode [Auto] Core Leveling Mode [Automatic mode] CCD Control [Auto] SATA Port Enable [Enabled] SATA Mode [AHCI] NVMe RAID mode [Disabled] SMART Self Test [Enabled] Hot Plug [Disabled] Hot Plug [Disabled] Hot Plug [Disabled] Hot Plug [Disabled] Hot Plug [Disabled] Hot Plug [Disabled] Hot Plug [Disabled] Hot Plug [Disabled] HD Audio Controller [Enabled] PCIEX16_1 Bandwidth Bifurcation configuration [Auto Mode] PCIEX16_2 Bandwidth Bifurcation configuration [Auto Mode] When system is in working state [All On] Q-Code LED Function [POST Code Only] When system is in sleep, hibernate or soft off states [All On] Realtek 2.5G LAN Controller [Enabled] Realtek PXE OPROM [Disabled] Intel LAN Controller [Enabled] Intel LAN OPROM [Disabled] ASM1074 Controller [Enabled] Wi-Fi 6 (802.11ax) Controller [Disabled] Bluetooth Controller [Enabled] USB power delivery in Soft Off state (S5) [Enabled] PCIEX16_1 Mode [Auto] PCIEX16_2 Mode [Auto] PCIEX1 Mode [Auto] PCIEX16_3 Mode [Auto] M.2_1 Link Mode [Auto] M.2_2 Link Mode [Auto] SB Link Mode [Auto] ErP Ready [Disabled] Restore AC Power Loss [Power Off] Power On By PCI-E [Disabled] Power On By RTC [Disabled] Above 4G Decoding [Enabled] Re-Size BAR Support [Auto] SR-IOV Support [Disabled] Legacy USB Support [Enabled] XHCI Hand-off [Enabled] Corsair Voyager GTX 0 [Auto] USB Device Enable [Enabled] U32G2_2 [Enabled] U32G2_3 [Enabled] U32G2_4 [Enabled] U32G1_10 [Enabled] U32G1_11 [Enabled] USB12 [Enabled] USB13 [Enabled] U32G2_7 [Enabled] U32G2_8 [Enabled] U32G2_C9 [Enabled] Network Stack [Disabled] Device [SATA6G_7: Samsung SSD 850 PRO 1TB] CPU Temperature [Monitor] CPU Package Temperature [Monitor] MotherBoard Temperature [Monitor] VRM Temperature [Monitor] T_Sensor Temperature [Monitor] Water In T Sensor Temperature [Monitor] Water Out T Sensor Temperature [Monitor] CPU Fan Speed [Monitor] CPU Optional Fan Speed [Monitor] Chassis Fan 1 Speed [Monitor] Chassis Fan 2 Speed [Monitor] Chassis Fan 3 Speed [Monitor] High Amp Fan Speed [Monitor] W_PUMP+ Speed [Monitor] AIO PUMP Speed [Monitor] PCH Fan Speed [Monitor] Flow Rate [Monitor] CPU Core Voltage [Monitor] 3.3V Voltage [Monitor] 5V Voltage [Monitor] 12V Voltage [Monitor] CPU Fan Q-Fan Control [Auto] CPU Fan Step Up [2.1 sec] CPU Fan Step Down [0 sec] CPU Fan Speed Low Limit [600 RPM] CPU Fan Profile [Manual] CPU Fan Upper Temperature [70] CPU Fan Max. Duty Cycle (%) [100] CPU Fan Middle Temperature [50] CPU Fan Middle Duty Cycle (%) [50] CPU Fan Lower Temperature [30] CPU Fan Min Duty Cycle (%) [40] Chassis Fan 1 Q-Fan Control [Auto] Chassis Fan 1 Q-Fan Source [CPU] Chassis Fan 1 Step Up [0 sec] Chassis Fan 1 Step Down [0 sec] Chassis Fan 1 Speed Low Limit [600 RPM] Chassis Fan 1 Profile [Manual] Chassis Fan 1 Upper Temperature [70] Chassis Fan 1 Max. Duty Cycle (%) [100] Chassis Fan 1 Middle Temperature [50] Chassis Fan 1 Middle Duty Cycle (%) [65] Chassis Fan 1 Lower Temperature [20] Chassis Fan 1 Min Duty Cycle (%) [60] Chassis Fan 2 Q-Fan Control [Auto] Chassis Fan 2 Q-Fan Source [CPU] Chassis Fan 2 Step Up [0 sec] Chassis Fan 2 Step Down [0 sec] Chassis Fan 2 Speed Low Limit [600 RPM] Chassis Fan 2 Profile [Manual] Chassis Fan 2 Upper Temperature [65] Chassis Fan 2 Max. Duty Cycle (%) [100] Chassis Fan 2 Middle Temperature [45] Chassis Fan 2 Middle Duty Cycle (%) [60] Chassis Fan 2 Lower Temperature [40] Chassis Fan 2 Min Duty Cycle (%) [60] Chassis Fan 3 Q-Fan Control [Auto] Chassis Fan 3 Q-Fan Source [CPU] Chassis Fan 3 Step Up [0 sec] Chassis Fan 3 Step Down [0 sec] Chassis Fan 3 Speed Low Limit [600 RPM] Chassis Fan 3 Profile [Manual] Chassis Fan 3 Upper Temperature [70] Chassis Fan 3 Max. Duty Cycle (%) [100] Chassis Fan 3 Middle Temperature [45] Chassis Fan 3 Middle Duty Cycle (%) [100] Chassis Fan 3 Lower Temperature [40] Chassis Fan 3 Min Duty Cycle (%) [100] High Amp Fan Q-Fan Control [Auto] High Amp Fan Q-Fan Source [CPU] High Amp Fan Step Up [0 sec] High Amp Fan Step Down [0 sec] High Amp Fan Speed Low Limit [600 RPM] High Amp Fan Profile [Manual] High Amp Fan Upper Temperature [70] High Amp Fan Max. Duty Cycle (%) [100] High Amp Fan Middle Temperature [45] High Amp Fan Middle Duty Cycle (%) [70] High Amp Fan Lower Temperature [30] High Amp Fan Min Duty Cycle (%) [60] Water Pump+ Q-Fan Control [Auto] Water Pump+ Q-Fan Source [CPU] Water Pump+ Upper Temperature [70] Water Pump+ Max. Duty Cycle (%) [100] Water Pump+ Middle Temperature [50] Water Pump+ Middle Duty Cycle (%) [65] Water Pump+ Lower Temperature [30] Water Pump+ Min Duty Cycle (%) [60] AIO Pump Q-Fan Control [Auto] AIO Pump Q-Fan Source [CPU] AIO Pump Upper Temperature [70] AIO Pump Max. Duty Cycle (%) [100] AIO Pump Middle Temperature [50] AIO Pump Middle Duty Cycle (%) [65] AIO Pump Lower Temperature [30] AIO Pump Min Duty Cycle (%) [60] Above 4GB MMIO Limit [39bit (512GB)] Fast Boot [Enabled] Next Boot after AC Power Loss [Fast Boot] Boot Logo Display [Disabled] Bootup NumLock State [On] POST Report [5 sec] Wait For 'F1' If Error [Enabled] Option ROM Messages [Force BIOS] Interrupt 19 Capture [Disabled] Setup Mode [Advanced Mode] Launch CSM [Disabled] OS Type [Other OS] AMI Native NVMe Driver Support [Enabled] Flexkey [Reset] Setup Animator [Disabled] Load from Profile [5] Profile Name [20.01 minus 30] Save to Profile [5] DIMM Slot Number [DIMM_A1] Bus Interface [PCIEX16_1] Download & Install ARMOURY CRATE app [Enabled] CPU Frequency [0] CPU Voltage [0] CCD Control [Auto] Core control [Auto] SMT Control [Auto] Overclock [Enabled ] Memory Clock Speed [Auto] Tcl [Auto] Trcdrd [Auto] Trcdwr [Auto] Trp [Auto] Tras [Auto] Trc Ctrl [Auto] TrrdS [Auto] TrrdL [Auto] Tfaw Ctrl [Auto] TwtrS [Auto] TwtrL [Auto] Twr Ctrl [Auto] Trcpage Ctrl [Auto] TrdrdScL Ctrl [Auto] TwrwrScL Ctrl [Auto] Trfc Ctrl [Auto] Trfc2 Ctrl [Auto] Trfc4 Ctrl [Auto] Tcwl [Auto] Trtp [Auto] Tcke [Auto] Trdwr [Auto] Twrrd [Auto] TwrwrSc [Auto] TwrwrSd [Auto] TwrwrDd [Auto] TrdrdSc [Auto] TrdrdSd [Auto] TrdrdDd [Auto] ProcODT [Auto] Power Down Enable [Auto] Cmd2T [Auto] Gear Down Mode [Auto] CAD Bus Timing User Controls [Auto] CAD Bus Drive Strength User Controls [Auto] Data Bus Configuration User Controls [Auto] Infinity Fabric Frequency and Dividers [Auto] ECO Mode [Disable] Precision Boost Overdrive [Advanced] PBO Limits [Motherboard] Precision Boost Overdrive Scalar [Auto] Curve Optimizer [Per Core] Core 0 Curve Optimizer Sign [Negative] Core 0 Curve Optimizer Magnitude [30] Core 1 Curve Optimizer Sign [Negative] Core 1 Curve Optimizer Magnitude [30] Core 2 Curve Optimizer Sign [Negative] Core 2 Curve Optimizer Magnitude [30] Core 3 Curve Optimizer Sign [Negative] Core 3 Curve Optimizer Magnitude [30] Core 4 Curve Optimizer Sign [Negative] Core 4 Curve Optimizer Magnitude [30] Core 5 Curve Optimizer Sign [Negative] Core 5 Curve Optimizer Magnitude [30] Core 6 Curve Optimizer Sign [Negative] Core 6 Curve Optimizer Magnitude [30] Core 7 Curve Optimizer Sign [Negative] Core 7 Curve Optimizer Magnitude [30] Core 8 Curve Optimizer Sign [Negative] Core 8 Curve Optimizer Magnitude [30] Core 9 Curve Optimizer Sign [Negative] Core 9 Curve Optimizer Magnitude [30] Core 10 Curve Optimizer Sign [Negative] Core 10 Curve Optimizer Magnitude [30] Core 11 Curve Optimizer Sign [Negative] Core 11 Curve Optimizer Magnitude [30] Core 12 Curve Optimizer Sign [Negative] Core 12 Curve Optimizer Magnitude [30] Core 13 Curve Optimizer Sign [Negative] Core 13 Curve Optimizer Magnitude [30] Core 14 Curve Optimizer Sign [Negative] Core 14 Curve Optimizer Magnitude [30] Core 15 Curve Optimizer Sign [Negative] Core 15 Curve Optimizer Magnitude [30] Max CPU Boost Clock Override [0MHz] Platform Thermal Throttle Limit [Auto] LN2 Mode [Auto] SoC/Uncore OC Mode [Disabled] VDDP Voltage Control [Auto] VDDG Voltage Control [Auto] NUMA nodes per socket [Auto] Custom Pstate0 [Auto] L1 Stream HW Prefetcher [Auto] L2 Stream HW Prefetcher [Auto] Core Watchdog Timer Enable [Auto] SMEE [Auto] Core Performance Boost [Auto] Global C-state Control [Disabled] Power Supply Idle Control [Typical Current Idle] SEV ASID Count [Auto] SEV-ES ASID Space Limit Control [Auto] Streaming Stores Control [Auto] Local APIC Mode [Auto] ACPI _CST C1 Declaration [Auto] MCA error thresh enable [Auto] PPIN Opt-in [Auto] Fast Short REP MOVSB [Enabled] Enhanced REP MOVSB/STOSB [Enabled] RdRand Speedup Disable [Enabled] IBS hardware workaround [Auto] DRAM scrub time [Auto] Poison scrubber control [Auto] Redirect scrubber control [Auto] Redirect scrubber limit [Auto] NUMA nodes per socket [Auto] Memory interleaving [Auto] Memory interleaving size [Auto] 1TB remap [Auto] DRAM map inversion [Auto] ACPI SRAT L3 Cache As NUMA Domain [Auto] ACPI SLIT Distance Control [Auto] ACPI SLIT remote relative distance [Auto] GMI encryption control [Auto] xGMI encryption control [Auto] CAKE CRC perf bounds Control [Auto] 4-link xGMI max speed [Auto] 3-link xGMI max speed [Auto] xGMI TXEQ Mode [Auto] PcsCG control [Auto] Disable DF to external downstream IP SyncFloodPropagation [Auto] Disable DF sync flood propagation [Auto] CC6 memory region encryption [Auto] Memory Clear [Auto] Overclock [Enabled] Memory Clock Speed [Auto] Tcl [Auto] Trcdrd [Auto] Trcdwr [Auto] Trp [Auto] Tras [Auto] Trc Ctrl [Auto] TrrdS [Auto] TrrdL [Auto] Tfaw Ctrl [Auto] TwtrS [Auto] TwtrL [Auto] Twr Ctrl [Auto] Trcpage Ctrl [Auto] TrdrdScL Ctrl [Auto] TwrwrScL Ctrl [Auto] Trfc Ctrl [Auto] Trfc2 Ctrl [Auto] Trfc4 Ctrl [Auto] Tcwl [Auto] Trtp [Auto] Tcke [Auto] Trdwr [Auto] Twrrd [Auto] TwrwrSc [Auto] TwrwrSd [Auto] TwrwrDd [Auto] TrdrdSc [Auto] TrdrdSd [Auto] TrdrdDd [Auto] ProcODT [Auto] Power Down Enable [Auto] Disable Burst/Postponed Refresh [Auto] DRAM Maximum Activate Count [Auto] Cmd2T [Auto] Gear Down Mode [Auto] CAD Bus Timing User Controls [Auto] CAD Bus Drive Strength User Controls [Auto] Data Bus Configuration User Controls [Auto] Data Poisoning [Auto] DRAM Post Package Repair [Default] RCD Parity [Auto] DRAM Address Command Parity Retry [Auto] Write CRC Enable [Auto] DRAM Write CRC Enable and Retry Limit [Auto] Disable Memory Error Injection [True] DRAM ECC Symbol Size [Auto] DRAM ECC Enable [Auto] DRAM UECC Retry [Auto] TSME [Auto] Data Scramble [Auto] DFE Read Training [Auto] FFE Write Training [Auto] PMU Pattern Bits Control [Auto] MR6VrefDQ Control [Auto] CPU Vref Training Seed Control [Auto] Chipselect Interleaving [Auto] BankGroupSwap [Auto] BankGroupSwapAlt [Auto] Address Hash Bank [Auto] Address Hash CS [Auto] Address Hash Rm [Auto] SPD Read Optimization [Enabled] MBIST Enable [Disabled] Pattern Select [PRBS] Pattern Length [6] Aggressor Channel [1 Aggressor Channel] Aggressor Static Lane Control [Disabled] Target Static Lane Control [Disabled] Worst Case Margin Granularity [Per Chip Select] Read Voltage Sweep Step Size [1] Read Timing Sweep Step Size [1] Write Voltage Sweep Step Size [1] Write Timing Sweep Step Size [1] IOMMU [Auto] Precision Boost Overdrive [Auto] Precision Boost Overdrive Scalar [Auto] FCLK Frequency [Auto] SOC OVERCLOCK VID [0] UCLK DIV1 MODE [Auto] VDDP Voltage Control [Auto] VDDG Voltage Control [Auto] SoC/Uncore OC Mode [Auto] LN2 Mode [Auto] ACS Enable [Auto] PCIe ARI Support [Auto] PCIe ARI Enumeration [Auto] PCIe Ten Bit Tag Support [Auto] cTDP Control [Auto] EfficiencyModeEn [Auto] Package Power Limit Control [Auto] APBDIS [Auto] DF Cstates [Auto] CPPC [Auto] CPPC Preferred Cores [Auto] NBIO DPM Control [Auto] Early Link Speed [Auto] Presence Detect Select mode [Auto] Preferred IO [Auto] CV test [Auto] Loopback Mode [Auto]

Data Link Feature Exchange [Disabled]

So i have finally updated my crosshair viii hero to bios 3204. No major changes compared to bios 3003, other then i don't need to run with cpu +voltage offset anymore and UBS ports are alittle less buggy with the hp reverb g2 vr headset.

Have seen alot of talk about the OCCT "large data set" stresstest in regards to WHEA errors, and how hard it was to run error-free so i decided to give it a 1 hour run :)

screenshot.png

screenshot done.png

Completed without any errors on my 24/7 settings, which are: (fans on auto) -30 allcore curve optimizer 4x8gigabyte samsung b-die @ 1900/3800mhz 1:1 with the infinity fabric.

This is my bloaty gaming windows, so latency are alittle on the high side.

harm9963

NEW ASUS DARK HERO MB, OC switching is amazing ! Warm day to, room is 78f !

  • 21.PNG
  • 1478.PNG
  • Capturenewerec.PNG
  • Capture12000.PNG

harm9963

No throttle , not surprise with a high end MB.
Dark Hero is fanless , stays at a static 40c , compare to my old ROG E X570 - 56C .

  • Captureniceooc.PNG

Mussels

We just need an automated way to tune in the curve optimiser and we'll allllll be happy.

jesdals

I seem to loose about 25 points in single core CPUz benzmark disabling the AMD cool and quiet function in the CPU bios section on my Auros Master X570, can any one verify that it hurts singlecore performance?

harm9963

  • Capturenit2.1.PNG

Page 2

Space Lynx

@harm9963 wow, thank you so much for the settings I think I have a winner, Best CPU-Z score I have ever had, great Cinebench R20 and R23 scores, voltage CPU Core SVI never goes beyond 1.469V) and temperatures are okish, never reaching 90C which is OK for Zen 3 CPU (According to AMD Technicians) and not thermal throttling. I think you might have a better cooling solution and your ambient temperature is lower, I'm thinking of getting the Artic Liquid Freezer II 360mm. Could you tell me what you have? Thanks! 1. Voltage: Auto 2. LLC: Auto 3. 200/200/145 for PPT/TDC/EDC 4. Curve -30 5. Scalar: Auto 6. Max Boost Override: +50Mhz 7. Platform Thermal: Auto

View attachment 183490 View attachment 183491

View attachment 183492 View attachment 183493

Can you pass prime95 with those settings?

@lynx29 at the time it did. The problem with the CO is when leaving the machine idle. I'm no longer running those settings. I found better tools/procedures to test stability while using CO. I'm running the following: X570 Aorus MASTER rev 1.2 with BIOS F32 Voltage: Auto CPU LLC: Auto PPT: 210 TDC: 150 EDC: 170 Scalar: Auto CO: -15, -20, -20, -15, +2, -15, -25, -25, -25, -20, -20, -20, -25, -25, -15, -15 Max Boost: +50Mhz Thermal Platform: 85C Using F33a beta bios with AGESA 1.2.0.0 I get better curve but I rolled back to F32 official bios.

These settings pass prime95, OCCT, AIDA, etc.

Space Lynx

@lynx29 at the time it did. The problem with the CO is when leaving the machine idle. I'm no longer running those settings. I found better tools/procedures to test stability while using CO. I'm running the following: X570 Aorus MASTER rev 1.2 with BIOS F32 Voltage: Auto CPU LLC: Auto PPT: 210 TDC: 150 EDC: 170 Scalar: Auto CO: -15, -20, -20, -15, +2, -15, -25, -25, -25, -20, -20, -20, -25, -25, -15, -15 Max Boost: +50Mhz Thermal Platform: 85C Using F33a beta bios with AGESA 1.2.0.0 I get better curve but I rolled back to F32 official bios.

These settings pass prime95, OCCT, AIDA, etc.

How did you know the CO negatives for each core? Is it that new tool that just came out?

@lynx29 basically you put -5 on all cores CO and test each core and if all of them passes, you add another -5. If one of the cores fails then you switch to per core CO and revert -5 to the core that failed you and then keep going.

In my case my core 4 (0 to 15) wasn't able to run stable using negative values or even 0, so I had to put +2. AGESA 1.2.0.0 fixes this for me but because i reverted to F32 I had to put +2 again.

Would guess it all comes down to the silicon lottery with me being able to run -30 allcore along with better cooling (?) The last time i was lucky was with a newcastle cpu running at above 3ghz on air cooling!

View attachment 187113

Yes, that would certainly help. I cannot run -30 all-core. It will run great during benchmarks, and improves my scores a good bit, but when the computer is idle it will reset / WHEA error on my highest priority cores. Also, I'm assuming you have better cooling than me. I'm running on regular old air (NH-D15) and using a quiet fan profile.

Makaveli

This is a great thread with alot of good info. Question if you set PBO to use motherboard limits. The settings look high compared to what i've seen you guys using.

Ryzen master.png

Precision boost overdrive.png

on CO Core 04 is my fastest and Core 01 second fastest. -17 on Core 04 -15 on Core 01 -10 on the rest of my cores

CO.png

jesdals

I found even with extreme good cooling there was no benefit for settings above 125 on Edc and Tdc, but it differ from motherboard to motherboard model and cpu lottery

Makaveli

I found even with extreme good cooling there was no benefit for settings above 125 on Edc and Tdc, but it differ from motherboard to motherboard model and cpu lottery

Good to know.

With my current settings its stable and performance is good the only thing that i see is high temps on a cinebench test. I will hit 90c on that. if I test with the aida64 stress test with AVX I hit 80c and everything else is well below that.

Question if you set PBO to use motherboard limits.

If you are OK with your temps, then motherboard limits are fine to use and should give you the best multi-core performance in all cases. I would do some single core stability testing with the curve offsets you are running though, and potentially reduce max cpu boost clock override. -17 on your fastest core along with +200 Max cpu boost clock override seems aggressive but this is going to be chip dependent, but if stable that is a pretty sweet result.

Typically people are able to set more negative values on the less favored cores than on their favored (highest performance) cores as those are already more aggressively tuned out of the box.

Makaveli

If you are OK with your temps, then motherboard limits are fine to use and should give you the best multi-core performance in all cases. I would do some single core stability testing with the curve offsets you are running though, and potentially reduce max cpu boost clock override. -17 on your fastest core along with +200 Max cpu boost clock override seems aggressive but this is going to be chip dependent, but if stable that is a pretty sweet result.

Typically people are able to set more negative values on the less favored cores than on their favored (highest performance) cores as those are already more aggressively tuned out of the box.

I need to do more testing on the off sets these are first set of numbers I tried. And pretty much left it at that to check stability so I've been on these for a week now no crashing on idle or Whea errors and its stable. Choose 200Mhz for the Max cpu boast and can hit 5.05Ghz and that seems stable also. I under the impression that you set the higher negative offset on your faster cores. As for silicon quality as well we all know it varies. The CTR app for whats is worth says I have a golden sample if that means anything.

5800X CTR Golden Sample2.png

I need to do more testing on the off sets these are first set of numbers I tried. And pretty much left it at that to check stability so I've been on these for a week now no crashing on idle or Whea errors and its stable. Choose 200Mhz for the Max cpu boast and can hit 5.05Ghz and that seems stable also. I under the impression that you set the higher negative offset on your faster cores. As for silicon quality as well all know it varies. The CTR app for whats is worth says I have a golden sample if that means anything.

View attachment 187705

Nice, mine comes back as silver and with a +200 offset I actually need to set a positive curve offset on 2 cores if I want 100% stability.

harm9963

Dynamic OC switching/ -20 still working and still learning with this feature, ASUS DARK HERO.

plus warm day ,75f room !

  • Capture013835.PNG
  • Capture5301.PNG
  • Capturetemp.PNG
  • Capturedocsw.PNG
  • Capture1818.PNG

I have upgraded to EK AIO 360 and it did improve my scores. However my ambient temp is too high (not using A/C).

CPU-Z 2021-02-11 21.43.14.png

CINEBENCH R20.060 2021-02-11 21.45.17.png

CINEBENCH R23.200 2021-02-11 21.51.59.png

harm9963

Only -20 , DOCS , BIOS auto

  • Capture20a.PNG
  • Capture1823.PNG
  • Capturetf.PNG

Happy to report that CTR 2.1 beta 6 (for early subscribers only, I'm a patreon supporter of @1usmus) is looking pretty good. I was able to get way more performance than using PBO+CO. Take a look at mi Cinebench R23. Using CTR 2.1 beta 6 ===================== PPT: 200 TDC: 150 EDC: 250 Temperature: 85C PX: 5000/4850/4800 at 1480mV (For single boosting) P2: 4650/4600 at 1356mV P1: 4575/4525 at 1256mV

CINEBENCH R23.200 2021-02-21 17.20.10.png

I just wanted to add Cinebench R20

CINEBENCH R20.060 2021-02-21 17.52.57.png

that worked - if its unstable does it identify which core/thread it was on?

Yes, although I need to do something to record a log of where it is up to in case the whole computer crashes, it does monitor the results.txt file that prime95 writes and then reports any error as well as copying that to a new file named per core and loop iteration.

Octopuss

I have my 5800X up and running, and would like to do a little undervolting, however I am fairly confused. 1) How can I tell how much of an offset to set for each core? From what I read, it's during idle when too much can cause instability, so how the hell do I test that, and how can I even know what core is malfunctioning? 2) I learned there is something like quality of cores and that Hwinfo can show that, but I can't make ANY sense out of this:

1614103731862.png

What the heck does all this mean? I've already looked at Ryzen Master and cores 1 and 7 are the best, but these numbers I don't understand. 3) When taking the cores quality into consideration, which ones are to be run with higher lor lower curve offset? I keep reading conflicting information about this. 4) Is it enough to just play with the curve optimizer in my use case? My head is already like a baloon staring into the BIOS.

I have no idea what kind of sample do I have, probably utter shit, because all core boost in Prime95 is 4300MHz on average, and highest frequency frequency shown in Hwinfo over some period of time (not constant load) is 4850MHz.


Page 3

I have my 5800X up and running, and would like to do a little undervolting, however I am fairly confused. 1) How can I tell how much of an offset to set for each core? From what I read, it's during idle when too much can cause instability, so how the hell do I test that, and how can I even know what core is malfunctioning? 2) I learned there is something like quality of cores and that Hwinfo can show that, but I can't make ANY sense out of this:

View attachment 189672

What the heck does all this mean? I've already looked at Ryzen Master and cores 1 and 7 are the best, but these numbers I don't understand. 3) When taking the cores quality into consideration, which ones are to be run with higher lor lower curve offset? I keep reading conflicting information about this. 4) Is it enough to just play with the curve optimizer in my use case? My head is already like a baloon staring into the BIOS.

I have no idea what kind of sample do I have, probably utter shit, because all core boost in Prime95 is 4300MHz on average, and highest frequency frequency shown in Hwinfo over some period of time (not constant load) is 4850MHz.

Part of the confusing thing is that no one can tell you what core offset will be stable for each core on your cpu. Every cpu is different and the only way to figure it out is by testing your cpu to see what is stable and gives good results. One thing that generally seems to be the case is that you can set a higher offset on the cores that are not rated as highly. This is why you see recommendations like set your best cores to X and all others to Y. However, even that might not be true on a single sample and so copying someone else's recommended settings is probably not going to work well. Take a step back and first understand which are the relevant bios settings. Then do some testing of your own to figure out what works well, while keeping an eye on temps and effective clocks in HWinfo.

I posted more details and a script here to test with prime95: https://www.overclock.net/threads/s...script-for-zen-3-curve-offset-tuning.1777112/

"I have no idea what kind of sample do I have, probably utter shit, because all core boost in Prime95 is 4300MHz on average, and highest frequency frequency shown in Hwinfo over some period of time (not constant load) is 4850MHz." That pretty much matches my stock behavior on a 5800x, and to be honest, tuning curve offsets, while keeping things 100% stable, didn't gain very much at all. 4850MHz is the stock max boost. With a power hungry avx all core load 4300MHz is about what you will get. Disabling AVX in p95 will give higher frequencies.

Review frequencies seem to be higher. e.g. https://www.techpowerup.com/review/amd-ryzen-7-5800x/21.html

I have a feeling review samples were very good chips, but "This test uses a custom-coded application that mimics real-life performance—it is not a stress test like Prime95.", so the test used for that graph is most likely using less power (and therefore generating less heat) than prime95.

Octopuss

I really just want to undervolt the thing a bit for temperatures purposes, I'm not interested in hunting for higher frequencies, because I feel like for any meaningful gains I'd have to spend enormous times understanding tons of settings and testing for even longer, and the little I could get with just some minor adjustments would make absolutely zero difference for regular usage.

I really just want to undervolt the thing a bit for temperatures purposes, I'm not interested in hunting for higher frequencies, because I feel like for any meaningful gains I'd have to spend enormous times understanding tons of settings and testing for even longer, and the little I could get with just some minor adjustments would make absolutely zero difference for regular usage.

It's not that simple since precision boost will try to run as high a frequency as it can within the power and temp limits. If you want your cpu to run cooler, you can enable PBO and lower the max temperature limit from the default value of 90C. Then PB will automatically limit to whatever temp you decide. Your other option is to enable PBO, but rather than raise the power limits, lower them below stock settings.

Final option is to manually set frequency and voltage. But then you lose the benefits of increased frequency when not all cores are loaded, and some power saving when cores are idle.

Octopuss

Well, I blindly lowered the curve by -5 and -10 and highest temperatures in Prime95 dropped by ~5°C, so I guess it works.

Well, I blindly lowered the curve by -5 and -10 and highest temperatures in Prime95 dropped by ~5°C, so I guess it works.

Ok, well I guess check that it is stable now and if it is then be happy

Or continue down the slippery slope to find the lowest stable curve offset you can get for each core :-D

Mussels

I have my 5800X up and running, and would like to do a little undervolting, however I am fairly confused. 1) How can I tell how much of an offset to set for each core? From what I read, it's during idle when too much can cause instability, so how the hell do I test that, and how can I even know what core is malfunctioning? 2) I learned there is something like quality of cores and that Hwinfo can show that, but I can't make ANY sense out of this:

View attachment 189672

What the heck does all this mean? I've already looked at Ryzen Master and cores 1 and 7 are the best, but these numbers I don't understand. 3) When taking the cores quality into consideration, which ones are to be run with higher lor lower curve offset? I keep reading conflicting information about this. 4) Is it enough to just play with the curve optimizer in my use case? My head is already like a baloon staring into the BIOS.

I have no idea what kind of sample do I have, probably utter shit, because all core boost in Prime95 is 4300MHz on average, and highest frequency frequency shown in Hwinfo over some period of time (not constant load) is 4850MHz.

Those clocks are about right for stock settings If you enable PBO and set +200, you'll see 5050 for the low load core boosts, ideally randomly each core will reach that over time.

I cant read the HWinfo stats either lol. Next you need to play with the curve optimiser and deal with the chaos of stability testing when you're undervolting one core at a time like the rest of us

Octopuss

Could you add code so the script can be run when prime95.exe is already present in the current folder? To me it makes more sense to download P95 manually and put it there. The script also replies on specific version of the program. Btw why that specific 84k FFT size?

Those clocks are about right for stock settings If you enable PBO and set +200, you'll see 5050 for the low load core boosts, ideally randomly each core will reach that over time.

I cant read the HWinfo stats either lol. Next you need to play with the curve optimiser and deal with the chaos of stability testing when you're undervolting one core at a time like the rest of us

I think I read that if you increase the frequency limit, whatever undervolt you might have will likely not be stable anymore. Is that true?

Also I don't understand how can people suggest to lower the curve by -40 or something like that. I am on very brief unscietific testing of fourth core, and -30 crashed Prime95 on 3 out of 4 of them within minutes. One crashed after 10 hours, one flat out restarted the PC :D

Could you add code so the script can be run when prime95.exe is already present in the current folder? To me it makes more sense to download P95 manually and put it there. The script also replies on specific version of the program.

The latest version will check if there is an existing p95 folder and use that instead of extracting p95 again. So if you want you can manually extract any version of p95 into a p95 subfolder and then run the script. You can also change the name of the p95 zip file to extract by editing the variable near the top of the script. Just keep in mind I only tested it with that specific version of p95.

Btw why that specific 84k FFT size?

Good question. I should add this to the readme or something... when I started testing early on I was using a range of FFT sizes, but noticed that I typically hit an error after reaching 84k FFT size. So I just set it to that as a shortcut. It could be that higher values are better though, or that different values expose different instabilities. Feel free to experiment with different values and if you are shooting for 110% stability, then setting it to a range and running for at least a couple hours on each core is something you could try.

I think I read that if you increase the frequency limit, whatever undervolt you might have will likely not be stable anymore. Is that true?

This is true, since the cpu will reach higher boost clocks and those might not be stable with the undervolt. Actually people are even finding they need to add voltage to some cores (via a positive core offset) to be stable if they set a boost clock override >0. This is the case with my cpu where one core needs +10, another +3, but all others can keep a 0 or negative setting.

Also I don't understand how can people suggest to lower the curve by -40 or something like that. I am on very brief unscietific testing of fourth core, and -30 crashed Prime95 on 3 out of 4 of them within minutes. One crashed after 10 hours, one flat out restarted the PC :D

I feel you... actually part of what motivated me to share this script was to help with all the confusion created by posts that suggest setting core offsets a certain way rather than testing what specifically works with your cpu. I think part of what makes this hard is that it is possible for an unstable core offset to remain hidden in a lot of tests as it will need a single core load on that specific core to expose it.

Everyone's definition of stable is different as well. My testing is more geared to having a 24/7 reliable machine that runs at a good speed, rather than trying to reach the highest single benchmark score I can. Those are different goals as for a high score you just need things stable for long enough to complete that test.

The latest version will check if there is an existing p95 folder and use that instead of extracting p95 again. So if you want you can manually extract any version of p95 into a p95 subfolder and then run the script. You can also change the name of the p95 zip file to extract by editing the variable near the top of the script. Just keep in mind I only tested it with that specific version of p95. Good question. I should add this to the readme or something... when I started testing early on I was using a range of FFT sizes, but noticed that I typically hit an error after reaching 84k FFT size. So I just set it to that as a shortcut. It could be that higher values are better though, or that different values expose different instabilities. Feel free to experiment with different values and if you are shooting for 110% stability, then setting it to a range and running for at least a couple hours on each core is something you could try. This is true, since the cpu will reach higher boost clocks and those might not be stable with the undervolt. Actually people are even finding they need to add voltage to some cores (via a positive core offset) to be stable if they set a boost clock override >0. This is the case with my cpu where one core needs +10, another +3, but all others can keep a 0 or negative setting. I feel you... actually part of what motivated me to share this script was to help with all the confusion created by posts that suggest setting core offsets a certain way rather than testing what specifically works with your cpu. I think part of what makes this hard is that it is possible for an unstable core offset to remain hidden in a lot of tests as it will need a single core load on that specific core to expose it.

Everyone's definition of stable is different as well. My testing is more geared to having a 24/7 reliable machine that runs at a good speed, rather than trying to reach the highest single benchmark score I can. Those are different goals as for a high score you just need things stable for long enough to complete that test.

Great job man

Octopuss

I think the script (version from 24.2.) is wrong. When I run Hwinfo, only one thread per core is being used. I guess the affinity is misconfigured.

I think the script (version from 24.2.) is wrong. When I run Hwinfo, only one thread per core is being used. I guess the affinity is misconfigured.

Did you change any of the variables from their defaults? There is a new setting, $use_smt=$true; If that is set to false then it will only use one thread on each core. Check in task manager to make sure the affinity is set correctly for prime95 while it is running. Let me know if it is not working on your system.

Thanks!

  • p95_affinity.png

Last edited: Feb 27, 2021

New AMD chipset driver - on x570 it seems stable P200T125E115
View attachment 187257
View attachment 187258

Hi mate I'm new here and have been reeding about this overclocking options with the Curve Optimizer. I followed and done it on my ROG Crosshair VIII Dark Hero with Ryzen 5950x. It seams that I am missing somenthing during the process. I only setup PBO via the advanced options with PPT:200 TDC:140 EDC:200 PBO Scalar: 10X Max CPU Boost Clock Override: 200Mhz Is stable but when I run Cinebench R23 multicore I get 28706 and the cores do not go over 4300MHz. Temp is arond 88 Celcius Have a couple of questions. 1. Is it possible for you to share screens of your Bios for relevant overclocking details to see if I am missing somethig. 2. With the Curve Optimizer is there a need to setup anything else like CPU Core Ratio (per ccx) or other settings in the Bios to get a 4500MHz multicore?

Regards,

Hi mate I'm new here and have been reeding about this overclocking options with the Curve Optimizer. I followed and done it on my ROG Crosshair VIII Dark Hero with Ryzen 5950x. It seams that I am missing somenthing during the process. I only setup PBO via the advanced options with PPT:200 TDC:140 EDC:200 PBO Scalar: 10X Max CPU Boost Clock Override: 200Mhz Is stable but when I run Cinebench R23 multicore I get 28706 and the cores do not go over 4300MHz. Temp is arond 88 Celcius Have a couple of questions. 1. Is it possible for you to share screens of your Bios for relevant overclocking details to see if I am missing somethig. 2. With the Curve Optimizer is there a need to setup anything else like CPU Core Ratio (per ccx) or other settings in the Bios to get a 4500MHz multicore?

Regards,

You need to find the curve offset settings, hidden somewhere under advanced options, amd overclocking in your bios. Then lower, and test for stability with both single core and multicore loads.

I would suggest setting Boost Clock Override back to 0. The 5950x will go up to 5050Mhz at stock.

Octopuss

Did you change any of the variables from their defaults? There is a new setting, $use_smt=$true; If that is set to false then it will only use one thread on each core. Check in task manager to make sure the affinity is set correctly for prime95 while it is running. Let me know if it is not working on your system.

Thanks!

This is really weird. the affinity is set correctly, but there is still load only on one of the virtual cores.

1614585037800.png

1614585014278.png

When you run p95 directly, it detects everything correctly, but when you do so from the script, apparently it does not.

jesdals

Hi mate I'm new here and have been reeding about this overclocking options with the Curve Optimizer. I followed and done it on my ROG Crosshair VIII Dark Hero with Ryzen 5950x.

It seams that I am missing somenthing during the process.

Try using PBO via the advanced options with PPT:200 TDC:125 EDC:115 PBO Scalar: 2-4X Max CPU Boost Clock Override: 200Mhz and a curve of negative -10 and then add more negative curve until not stable and then try to change TDC and then EDC

to begin with - do note my previos bios pics - I personally also set Infinity fabric and memory to 1:1 - in my case its stable at 1900MHz - giving 3800MHz memory

Try using PBO via the advanced options with PPT:200 TDC:125 EDC:115 PBO Scalar: 2-4X Max CPU Boost Clock Override: 200Mhz and a curve of negative -10 and then add more negative curve until not stable and then try to change TDC and then EDC

to begin with - do note my previos bios pics - I personally also set Infinity fabric and memory to 1:1 - in my case its stable at 1900MHz - giving 3800MHz memory

I have worked out most of the values for the curve with negative -15 for best 4 cores and -30 for the remaining cores. Is stable with PPT185 TDC140 EDC140 Now what I have been noticing from the beginning is that the temp in full load when I’m compressing videos is around 90 Celcius which is too hot to keep for hours if I have a batch to run. How do I reduce the temp to something like 80s during full load?

Regards

I have worked out most of the values for the curve with negative -15 for best 4 cores and -30 for the remaining cores. Is stable with PPT185 TDC140 EDC140 Now what I have been noticing from the beginning is that the temp in full load when I’m compressing videos is around 90 Celcius which is too hot to keep for hours if I have a batch to run. How do I reduce the temp to something like 80s during full load?

Regards

Set PBO temperature limit to 80, or whatever you want your max temp to be. 90c is the default, and should be fine. I have mine at 85 though. This is assuming you've already checked there is no issue with your cooling (temps ok at lower loads).

This is really weird. the affinity is set correctly, but there is still load only on one of the virtual cores.
View attachment 190457
View attachment 190456

When you run p95 directly, it detects everything correctly, but when you do so from the script, apparently it does not.

By default the script only runs a single actual thread of p95 to try and reach as high a boost clock as possible. It then assigns it to both threads on a single core and let's windows voice between them as it likes. Is this what you are seeing?

Octopuss

I don't think that's what I'm seeing. Whatever I am seeing it's wrong, because when I run prime95 manually, the load is on both threads, which I assume is the correct behaviour with SMT enabled.

I don't think that's what I'm seeing. Whatever I am seeing it's wrong, because when I run prime95 manually, the load is on both threads, which I assume is the correct behaviour with SMT enabled.

When you run it manually, are you limiting the number of threads that p95 will create? By default, the script sets the number of threads to '1'

1614708517374.png

Octopuss

I set it to two, obviously. Two threads per core.

I set PBO frequency to +200MHz and when running p95, not only is the frequency not higher, it seems to be slightly lower than before (I am still testing the cores individually to have stable max boost). Is that even possible?

Zach_01

Latest beta versions of HWiNFO offer a tooltip description when you hover the pointer over a sensor. Anyway the “Core X Clock (perf#x/x)” naming scheme means the following... perf#x/x The second x (/x) is the performance order of the cores hard coded into the CPU logic after its fabrication evaluation. Meaning that the cores with numbers 1, 2 can operate at higher frequency on the same voltage from cores with numbers 7, 8 if we are talking about an 8core SKU. The first x (x/) is the order the Windows scheduler is preferring to load first and keep the loads more in single or reduced threaded apps/games. If the scheduler is “smart” enough, is trying to keep the loads on the same 4/8core CCX (depending if the CPU is 3000 or 5000) to reduce latency by avoiding CCX and CCD cross talks. And also is trying to load the higher perf# cores more. This behavior depends also on the power plan and/or the version of Windows and chipset drivers. The core loading “amount” can be represented by 3 other things on HWiNFO sensors window. 1. The core effective speed. The higher the core effective speed the higher the load on those cores. 2. The core C-states residency. Where C0 is the active state, C1 is the halt state and C6 is the power off (sleep). 3. The core T usage (%) Usually the discrete clock and raw multiplier readings can represent the core loading but at some points it could be misleading. Only effective clock and active/nonactive states can show this accurately. —————————————— Keep in mind that desktop Ryzens (unlike mobile ones) can’t feed cores with individual voltage. Instead the CPU operates cores at different speed on the same voltage. The core feeding voltage is one and only, the “CPU Core Voltage (SVI2 TFN)” as HWiNFO reports. To feed voltage individually to cores you either do it with multi VR rails but this adds high complexity and cost for the boards... or you do it with on die core V regulators the dLDOs. At least ZEN3 has them but they are not put to use on desktop SKUs. On mobile they do use them, where power control and efficiency is essential.

EDIT: by mistake I misspell dLDOs with cLDOs...

Octopuss

Could that mean that undervolting with curve optimizer is poinless to set on per-core basis? Because if I can only undervolt the top one or two cores by -5 and the crappy ones by -30, I presume the controller would only go as low as -5, right? edit: At the same time, how is this possible then?

1614853446824.png

I set it to two, obviously. Two threads per core.

OK, sorry it wasn't clear. I'm intentionally only running a single thread of p95 to try and get as high a boost clock as possible. This method worked well for me, but I have not done a lot of testing with 2 threads running to see how that compares. If you want to experiment with this you can change the p95 config the script uses by editing "local.txt".

I set PBO frequency to +200MHz and when running p95, not only is the frequency not higher, it seems to be slightly lower than before (I am still testing the cores individually to have stable max boost). Is that even possible?

You should run a benchmark or check the effective clock frequencies to compare. Just a theory, but it's possible you will get a slight degradation in performance if the cpu tries to boost higher, uses more power, and then lowers the effective clocks as a result. If that happens it is likely only a small effect though.

or you do it with on die core V regulators the cLDOs. At least ZEN3 has them but they are not put to use on desktop SKUs. On mobile they do use them, where power control and efficiency is essential.

This is something I have been wondering about. There was another post somewhere saying this was actually enabled in Zen3, but I don't know if that was just speculation or not. Assuming there is no documentation or other statement from AMD that they have enabled individual core voltages on desktop then I'll continue to assume all cores run off a single voltage. One thing I realized, if the cpu has voltage regulators on die, then are the seemingly high (1.4-1.5) voltages what is actually seen by the cores? It would explain a lot to me if the cpu voltage we are seeing is actually the supply to onboard voltage regulators and not the actual voltage that the cores run at.

Could that mean that undervolting with curve optimizer is poinless to set on per-core basis? Because if I can only undervolt the top one or two cores by -5 and the crappy ones by -30, I presume the controller would only go as low as -5, right? edit: At the same time, how is this possible then?

View attachment 190906

There's multiple things at play here, but the short answer is that it is not pointless to optimize on a per-core bases. At least if you want to get the last few % and as far as my understanding goes: 1. -5 on your best core might actually result in the same VID as -30 on your worst core, since each core has a different voltage curve set from the factory (and curve offset is setting an offset to the factory default). 2. Core VID in HWInfo is the voltage requested by that core. The actual voltage is going to be different to that, and assuming there is only a single voltage set for all cores it will be based on the highest voltage requested among the cores and possibly other parts of the die. 3. For a multi-core load the voltage will be set based on the highest voltage requested among the active cores. You can focus on those cores when tuning curve offset to save time, but you need to figure out which ones they really are first. Also note that by default HWInfo will report the same VID for all cores, but will read out sensor/register values sequentially, so in your screenshot you might just be seeing the VID changing between the time it read the value for Core 0 and Core 1. There is a setting that was added in HWiNFO v6.40 which seems to fix this. If you enable this I'd expect that you will see a slightly different VID for each core:

  • Added Snapshot Polling mode for AMD Zen-based CPUs.

Octopuss

At this point I don't really know what I'm doing anymore, there are just too many questions and I'm totally sick of prime95 by now :D Should I test the individual curve offsets with all other curves at default or is it perfectly fine to do so with some offsets already set? Does the increased frequency limit even matter when (according to someone on a different forum) it adds "up to" 200MHz, meaning it's more like +100MHz with two cores and probably next to nothing with more than that running under load? Just by having programs opened and browsing the internet, I see activity on four cores most of the time. and Is the frequency limit of any use when I'm not overclocking in the classical way of increasing power limits, voltages etc.? Can/does the increased frequency limit affect how much of an offset can I use?

Re: threads. I am not sure. I've always - for years - read that for HT/SMT enabled CPU, you have to run two threads, otherwise the load won't be 100%. I don't want to dig into reading about any of the technical klingon behind that, so I just do what I've been doing for years.


Page 4

Zach_01

1. You should run a benchmark or check the effective clock frequencies to compare. Just a theory, but it's possible you will get a slight degradation in performance if the cpu tries to boost higher, uses more power, and then lowers the effective clocks as a result. If that happens it is likely only a small effect though.

2. This is something I have been wondering about. There was another post somewhere saying this was actually enabled in Zen3, but I don't know if that was just speculation or not. Assuming there is no documentation or other statement from AMD that they have enabled individual core voltages on desktop then I'll continue to assume all cores run off a single voltage.

3. One thing I realized, if the cpu has voltage regulators on die, then are the seemingly high (1.4-1.5) voltages what is actually seen by the cores? It would explain a lot to me if the cpu voltage we are seeing is actually the supply to onboard voltage regulators and not the actual voltage that the cores run at.

4. There's multiple things at play here, but the short answer is that it is not pointless to optimize on a per-core bases. At least if you want to get the last few % and as far as my understanding goes:

_1. -5 on your best core might actually result in the same VID as -30 on your worst core, since each core has a different voltage curve set from the factory (and curve offset is setting an offset to the factory default). _2. Core VID in HWInfo is the voltage requested by that core. The actual voltage is going to be different to that, and assuming there is only a single voltage set for all cores it will be based on the highest voltage requested among the cores and possibly other parts of the die. _3. For a multi-core load the voltage will be set based on the highest voltage requested among the active cores. You can focus on those cores when tuning curve offset to save time, but you need to figure out which ones they really are first.

5. Also note that by default HWInfo will report the same VID for all cores, but will read out sensor/register values sequentially, so in your screenshot you might just be seeing the VID changing between the time it read the value for Core 0 and Core 1. There is a setting that was added in HWiNFO v6.40 which seems to fix this. If you enable this I'd expect that you will see a slightly different VID for each core:

  • Added Snapshot Polling mode for AMD Zen-based CPUs.

1. Yes the effective clock shows the sustainability of the boost/clock. In other words, the true(er) speed of a core. 2. (post #13) 3. The voltages of 1.4~1.5V is what the board VRM are supplying. At least on my R5 3600/AorusPro X570 system I can verify this via the VR sensor readings (VR VOUT). Anyone with VR VOUT readings and 5000series CPU? ...shed some light please, although I expect it to be the same as 3000series (see 2.) 4. I too don't think it's pointless. The individual offset could just mean different end speed for each core at the same voltage, but that is not something I can verify as I don't own a 5000 CPU. And this is making sense if you think about how the core speeds are acting under stock conditions. Why the curve optimizer would work any different? It is just a voltage/speed curve alteration over stock. 5. As matter of fact, after I enabled it I see identical VIDs (cur/min/max/avg). Prior to this there was a difference in requests. And again, I own a 3000series CPU. --------------------------- Lets not forget that there is no software that can monitor and report speed/voltage 100% accurately when those values are different every 1~20ms (depending power plan), and a software's polling period is 500, 1000, 2000ms.

By the way, dont use polling period under 1000ms on HWiNFO because it will probably keep the CPU more on active state (and on higher voltage) and out of halt/sleep states.

At this point I don't really know what I'm doing anymore, there are just too many questions and I'm totally sick of prime95 by now :D Should I test the individual curve offsets with all other curves at default or is it perfectly fine to do so with some offsets already set? Does the increased frequency limit even matter when (according to someone on a different forum) it adds "up to" 200MHz, meaning it's more like +100MHz with two cores and probably next to nothing with more than that running under load? Just by having programs opened and browsing the internet, I see activity on four cores most of the time. and Is the frequency limit of any use when I'm not overclocking in the classical way of increasing power limits, voltages etc.? Can/does the increased frequency limit affect how much of an offset can I use?

Re: threads. I am not sure. I've always - for years - read that for HT/SMT enabled CPU, you have to run two threads, otherwise the load won't be 100%. I don't want to dig into reading about any of the technical klingon behind that, so I just do what I've been doing for years.

OK, the first thing I want to say is that this is all for fun. If you are tired of testing or not having fun anymore, then take a break :) In the overall scheme of performance of your PC we are really only chasing small gains here and in the end it's probably not worth doing other than for the fun of it as AMD has already tuned the CPU to work very close to its limits out of the box. If you want the simple solution, just enable XMP, enable PBO (but keep all PBO settings at default), run a few stability tests and check temps, and then just enjoy using your PC. With regards to your questions:

1. It's perfectly fine to test with different offsets set on each core. Actually this is how I think you need to do it since you want to end up with some offset set for each core. Take a look at the "Example Scenario" in this post for some idea of how to go about this: https://www.overclock.net/threads/s...script-for-zen-3-curve-offset-tuning.1777112/

2. The increased frequency limit might not make much or any difference in any real load. For a 5600x it is probably worth trying for +200. For a 5800x it might not be worth increasing at all since, 3. Yes, if you increase the frequency limit it can affect how much curve offset you can apply, which will then reduce the maximum boost frequency for that core.

1. Yes the effective clock shows the sustainability of the boost/clock. In other words, the true(er) speed of a core. 2. (post #13) 3. The voltages of 1.4~1.5V is what the board VRM are supplying. At least on my R5 3600/AorusPro X570 system I can verify this via the VR sensor readings (VR VOUT). Anyone with VR VOUT readings and 5000series CPU? ...shed some light please, although I expect it to be the same as 3000series (see 2.) 4. I too don't think it's pointless. The individual offset could just mean different end speed for each core at the same voltage, but that is not something I can verify as I don't own a 5000 CPU. And this is making sense if you think about how the core speeds are acting under stock conditions. Why the curve optimizer would work any different? It is just a voltage/speed curve alteration over stock. 5. As matter of fact, after I enabled it I see identical VIDs (cur/min/max/avg). Prior to this there was a difference in requests. And again, I own a 3000series CPU. --------------------------- Lets not forget that there is no software that can monitor and report speed/voltage 100% accurately when those values are different every 1~20ms (depending power plan), and a software's polling period is 500, 1000, 2000ms.

By the way, dont use polling period under 1000ms on HWiNFO because it will probably keep the CPU more on active state (and on higher voltage) and out of halt/sleep states.

Not sure if I have VR VOUT on my board? Here are a couple of screenshots that show different VIDs on a 5800x. First one is p95, 16 threads, FFT size of 84 in place, avx2. Average voltages best captures the difference between cores.

Second one is captured while running the script, single thread, FFT size of 84 in place, avx disabled. Look at max voltages for this one.

  • p95_all_core.png
  • p95_single_thread.png

Det0x

Today i found out how much latency difference there really is between a 1CCD and 2CCD Zen3 CPU :) (5600x + 5800x VS 5900x + 5950x) I disabled one CCD on my 5950x to simulate a 5800x 1 CCD = 51.7 ns in aida64 2 CCD = 54.2 ns in aida64

1ccd vs 2ccd.png

Running 4x8gigs bdie memory and PBO CO @ standard 24/7 everyday settings

It is not how you do things. Also, 1.538v - I hope it is due to a fault in the sensor. If you are looking for ST scores you clamp PPT, and otherwise for MT you keep PPT just 10% above EDC. Those scores are too high. Better safe than sorry.

Your comprehension of how electricity works is sorely lacking. 1.5 volts is not an issue for CPU. Do high voltage transmission lines burn up carrying 200,000 volts? No, with an exception, when an object gets close enough that electrons jump from transmission line to grounded object, then electrons making that jump can cause a fireworks display and damage from electrons crossing the charged gap. Volts don't create heat, amps create heat. That is why transmission lines can handle 200,000 volts but are transmitting a few amps. Batteries, on the other hand, operate at low voltages, but with high amperage (current). When a battery is dead shorted the 100+ amps can cause the cable to get hot, start glowing red, and/or otherwise burn up. You CPU operates more like a battery than a high voltage transmission line.

The conclusion to this matter is that volts don't cause the CPU to fail, though I suppose that if they could get high enough and there was some less insulated part within the CPU that allowed the electrons to jump (see transmission line example) then volts could be responsible for killing CPU. Amperage is what makes your CPU hot. Find out what amperage your CPU can handle and work with that. Current CPUs have so many sensors and monitoring capabilities that they protect themselves, for the most part, from being burnt up due to overclocking.

Your comprehension of how electricity works is sorely lacking.

The conclusion to this matter is that volts don't cause the CPU to fail, though I suppose that if they could get high enough and there was some less insulated part within the CPU that allowed the electrons to jump (see transmission line example) then volts could be responsible for killing CPU. Amperage is what makes your CPU hot. Find out what amperage your CPU can handle and work with that. Current CPUs have so many sensors and monitoring capabilities that they protect themselves, for the most part, from being burnt up due to overclocking.

You seem awfully insecure for someone without even a trace 'notion' of how electricity travels in the opposite path that electrons travel... The issue is, I 'know' how electrons are even shaped like and they can lose track into unobserved paths since they are not shaped like a 'single' orbital physical ring, but double orbital rings with a second perpendicular one connected at the electron's moment center that can crossect without any discrimination for physical barriers - its quantum orbit does not interact with physical matter, just cuts across it. So much to say, electrons having an extra invisible quantum orbital ring that they can always switch between just as easily as the one they are physically travelling on - that is the very definition of Murphy's Law in electricity "what will go wrong, does" without you even knowing it. Your understanding of electrons is just half of the story, you are trying to solve it as a solid matter. That is not how it intuitively behaves.

Octopuss

Less physics and e-peen, more human speech pls :p

You seem awfully insecure for someone without even a trace 'notion' of how electricity travels in the opposite path that electrons travel... The issue is, I 'know' how electrons are even shaped like and they can lose track into unobserved paths since they are not shaped like a 'single' orbital physical ring, but double orbital rings with a second perpendicular one connected at the electron's moment center that can crossect without any discrimination for physical barriers - its quantum orbit does not interact with physical matter, just cuts across it. So much to say, electrons having an extra invisible quantum orbital ring that they can always switch between just as easily as the one they are physically travelling on - that is the very definition of Murphy's Law in electricity "what will go wrong, does" without you even knowing it. Your understanding of electrons is just half of the story, you are trying to solve it as a solid matter. That is not how it intuitively behaves.

But as the famous physicist Richard Feynman once said, “It doesn’t matter how beautiful your theory is, it doesn’t matter how smart you are. If it doesn’t agree with experiment, it’s wrong.” And he was a theoretician.

I'm sorry for being rude and insulting to you.

I'm not insecure, only allowing for variances with other aspects that I don't understand as fully.

Octopuss

Any idea how long should I keep OCCT tests running on each core? I'm used to prime95 where I won't settle for anything less than 12 hours without errors, but OCCT might do completely different things and this might not apply at all.

Any idea how long should I keep OCCT tests running on each core? I'm used to prime95 where I won't settle for anything less than 12 hours without errors, but OCCT might do completely different things and this might not apply at all.

Just run it until the test reaches your usual operating cpu temperature. Generally, I find no difficulty in setting the temperature just right, instead I find it hard keeping it down. The core will present an error in case its overclock is not stable. You can adjust the temperature using the FFT length gauge(speaking about P95). I set it 16-16 for a near original power draw that isn't inherently too synthetic.

Octopuss

Um, you know these CPUs reach the max temperature within about two seconds, right?

Octopuss

I have no idea what are you talking about.
I guess for you it is, otherwise you wouldn't suggest what you wrote (which didn't answer my question even partially).

Any idea how long should I keep OCCT tests running on each core? I'm used to prime95 where I won't settle for anything less than 12 hours without errors, but OCCT might do completely different things and this might not apply at all.

I think it is similar to prime95 in the way it works. For a final stability test I would do at least one hour per core, but generally I've found prime95 to be better at finding errors than occt.

In my BIOS for the B550-E Gaming you can reach "Precision Boost Overdrive" via two different routes: the "AI Tweaker" tab or or the "Advanced" tab. Both have nearly the same options except that the latter allows you to change the Curve Optimizer function while the former does not. Upon further testing I have noticed that both PBOs behave vastly differently: turning the one under AI Tweaker on, my 5600x's power consumption shoots from 76W (stock draw) to 95W or more (it's also worth noting that turning on "AMD Performance Enhancer" under the same AI Tweaker menu has the same effect, and I've found no practical difference between enabling either that or PBO). Whereas turning on the PBO under the Advanced tab on with the same sub-options (and Curve Optimizer on Auto), power draw stays at 76W, the normal power draw constraint. It is clear that despite being named the same and having very similar sub-options, these two features are completely different. Can someone knowledgeable please explain this mess? Example images just to show the two different paths (the values for the options are not what I'm using):

20210422_152850-12.jpg

20210422_153119-12.jpg

Last edited: Apr 25, 2021

Zach_01

PBO on auto means disabled.
Only on enabled is really enabled. That’s why you saw this difference.

Mussels

In my BIOS for the B550-E Gaming you can reach "Precision Boost Overdrive" via two different routes: the "AI Tweaker" tab or or the "Advanced" tab. Both have nearly the same options except that the latter allows you to change the Curve Optimizer function while the former does not. Upon further testing I have noticed that both PBOs behave vastly differently: turning the one under AI Tweaker on, my 5600x's power consumption shoots from 76W (stock draw) to 95W or more (it's also worth noting that turning on "AMD Performance Enhancer" under the same AI Tweaker menu has the same effect, and I've found no practical difference between enabling either that or PBO). Whereas turning on the PBO under the Advanced tab on with the same sub-options (and Curve Optimizer on Auto), power draw stays at 76W, the normal power draw constraint. It is clear that despite being named the same and having very similar sub-options, these two features are completely different. Can someone knowledgeable please explain this mess? Example images just to show the two different paths (the values for the options are not what I'm using):

View attachment 198116

View attachment 198117

my board has this too, its weird to see two sets of the same settings that dont sync

PBO on auto means disabled.
Only on enabled is really enabled. That’s why you saw this difference.

But see the second screenshot. In order to access all the options you can see, it must be set to "advanced", which must mean it is on, otherwise there would be no point in setting any of the options it reveals. Which of the OTHER options besides "PBO -> Advanced" must be turned on for PBO to work from this menu? PBO limits?

Zach_01

I apologize in advance as for some of the following I'm not entirely sure or cant remember which is which... Its been over a year since I was messing with OC settings and I'm old now... 1. I'm not entirely sure but if you choose "Advanced" mode you have to at least choose a "Max CPU boost clock Override" other than 0MHz. (fact=)Other settings can be on "Auto" or choose your own. If you just put it to Enable it does everything by it self. 2.(fact=) PBO Scalar when on Disabled/Auto is by default X1. Only on "Manual" and other than X1 is doing something. Be aware of it as past the X2-3 the voltage feed of the CPU could be too high, or even stupid high... (X4/5+) 3. Some/a lot settings are placed into 2 different places. AMD CBS AMD Overclocking

4. I think the one overrides the other but cant remember which one. Also settings on "AMD Overclocking" cant be returned to default on auto clear CMOS/default settings after several failed boot attempts that most boards do.

Are rounding errors with SSE CoreCylcer test important?

kiddagoat

Ryzen 5000 Precision Boost Overdrive 2 Guide - YouTube This video helped me out tremendously. He breaks this down and explains just about all of the settings. I found this one night after I got my 5900x. I haven't dialed in my curve yet, but I am currently at -15 on all cores and I boost 4.7ghz all cores and I have seen single cores get up to 5.1-5.2ghz.

Lots of patience and trial and error.

But see the second screenshot. In order to access all the options you can see, it must be set to "advanced", which must mean it is on, otherwise there would be no point in setting any of the options it reveals. Which of the OTHER options besides "PBO -> Advanced" must be turned on for PBO to work from this menu? PBO limits?

It might help to think of it like this: PBO stands for precision boost override. Precision boost is the stock functionality that is used to calculate what clock speed to run the cpu at. It does this based on a number of factors including power, temp, voltage required for stability at a given frequency, max safe voltage, etc. PBO allows you to over-ride some of the stock settings for power, temp, etc. Typically PBO is used to increase the stock power limits. You can also use it to lower them, as well as lower or raise temp limits. So, when you "enable" PBO, you also need to make sure that the settings for power, temp, etc are as you want them. If they are all still at the default setting, then simply enabling PBO as a feature won't change anything. In the ASUS bios, there is an ASUS provided menu for PBO control. This is the one under the "AI Tweaker" tab. When you enable PBO in this tab, the bios automatically increases the stock power limits. As far as I have seen, any settings for power limits, etc, in this tab take priority over ones in the Advanced menu. i.e. whatever power limit is set here will be the one that is used, regardless of what is set in the "Advanced" tab. The "Advanced" tab has the options that AMD exposes as part of the standard AGESA implementation. So, you basically have the ASUS menu (under AI Tweaker) and the AMD menu (under Advanced).

My recommendation is to do all of your settings under the ASUS provided "AI Tweaker" menu, since those settings take priority. Leave everything under Advanced->AMD Overclocking at default/auto or motherboard to make sure there is no conflict.

The one exception to this is when you come to tune curve offsets, since those are not available in the AI Tweaker section. For those, adjust under Advanced->AMD Overclocking->Precision Boost Override->Curve Optimizer.

Be careful that if you disable PBO, the Curve Optimizer settings still take effect (even though the menu disappears in the bios). So, if you want to return to stock, you need to either reset the bios completely, or make sure to go in and zero all off the curve offsets before disabling PBO or returning any of the other settings to stock.

I'm testing my curve optimizer undervolts with CoreCycler, and it turns out my new 5600x's Core 1 returns rounding errors even with no offset (negative offset set to 0). Does this mean this core is defective (and if so, is it worth it to activate my warranty just for this)? Or should I attempt to fix this by applying a positive offset?

@blu3dragon thank you for your info!


Page 5

I'm testing my curve optimizer undervolts with CoreCycler, and it turns out my new 5600x's Core 1 returns rounding errors even with no offset (negative offset set to 0). Does this mean this core is defective (and if so, is it worth it to activate my warranty just for this)? Or should I attempt to fix this by applying a positive offset?

@blu3dragon thank you for your info!

I'm just learning about this CoreCycler and the ins and outs of PBO, etc, but I have to ask, does this core pass with XMP (DOCP) turned off? Also, how much influence does LLC have? I've run prime95 for hours at various settings and found neg30 to all but my last core at neg27 was stable for hours upon hours (Edit: That includes monitoring HWiNFO64 and verifying clock and effective clock are matching). Now I run this CoreCycler, and they're failing within a minute or so, so now I feel like I don't know up from down, left from right. Back to testing!

I'm just learning about this CoreCycler and the ins and outs of PBO, etc, but I have to ask, does this core pass with XMP (DOCP) turned off? Also, how much influence does LLC have? I've run prime95 for hours at various settings and found neg30 to all but my last core at neg27 was stable for hours upon hours (Edit: That includes monitoring HWiNFO64 and verifying clock and effective clock are matching). Now I run this CoreCycler, and they're failing within a minute or so, so now I feel like I don't know up from down, left from right. Back to testing!

That's interesting, because CoreCycler at default settings uses none other than Prime95. What's probably happening is that you were running Prime95 on all cores, which makes the cores reach a lower clock, whereas CoreCycler runs Prime95 on one core at a time, which reaches a high clock. Compare my 5600x running 4050 MHz all cores vs. 4650 MHz on one core on CoreCycler. The higher clock exposes the instability. It's also worth noting that it's unlikely all your cores really are stable at -27 or below. Check out the offsets I reached: -9, +3, -25, -14, -19, 7. They are all over the place. You'd have to have bought a processor hand-picked by Jesus himself for all cores to be stable with the offsets you describe (or perhaps by Beelzebub himself, since I've read that the cores that are unstable with a not-very-negative offset is because they're already receiving their optimal voltage/are good cores).

An update to my issue: In the end I just grit my teeth and put +3 offset on Core 1, it seems to be perfectly stable now.

That's interesting, because CoreCycler at default settings uses none other than Prime95. What's probably happening is that you were running Prime95 on all cores, which makes the cores reach a lower clock, whereas CoreCycler runs Prime95 on one core at a time, which reaches a high clock. Compare my 5600x running 4050 MHz all cores vs. 4650 MHz on one core on CoreCycler. The higher clock exposes the instability. It's also worth noting that it's unlikely all your cores really are stable at -27 or below. Check out the offsets I reached: -9, +3, -25, -14, -19, 7. They are all over the place. You'd have to have bought a processor hand-picked by Jesus himself for all cores to be stable with the offsets you describe (or perhaps by Beelzebub himself, since I've read that the cores that are unstable with a not-very-negative offset is because they're already receiving their optimal voltage/are good cores).

An update to my issue: In the end I just grit my teeth and put +3 offset on Core 1, it seems to be perfectly stable now.

It wasn't too terrible dialing in the offsets. I just watched twitch and youtube while corecycler did its thing. I ended up at -28, -20, -28, -29, -16, -18, -27, -14 with +200Mhz boost. I run some single core benches and older games and wanted to keep that boost. 7's the dreamer wanting to kiss 5050Mhz which is probably why it needs -14 to reach it, but it was a recurring theme during reboots, cursing 4, 5, and 7 for dashing my plans.

I'll throw an Arctic Freezer II 280mm aio on it today or tomorrow, and then I'll finally be able to see the RAM again, hiding underneath a Noctua NH-D14 120mm fan. I'll be able to pull 2 sticks and maybe do some better bench runs.

It wasn't too terrible dialing in the offsets. I just watched twitch and youtube while corecycler did its thing. I ended up at -28, -20, -28, -29, -16, -18, -27, -14 with +200Mhz boost. I run some single core benches and older games and wanted to keep that boost. 7's the dreamer wanting to kiss 5050Mhz which is probably why it needs -14 to reach it, but it was a recurring theme during reboots, cursing 4, 5, and 7 for dashing my plans.

I'll throw an Arctic Freezer II 280mm aio on it today or tomorrow, and then I'll finally be able to see the RAM again, hiding underneath a Noctua NH-D14 120mm fan. I'll be able to pull 2 sticks and maybe do some better bench runs.

Those are still some very impressive offsets.

You might want to try leaving corecycler running when you are not using the pc as those background tasks could be lowering your boost clocks.

Most of you running stable in everything with a CO offset but crash at some point during idle. It's called DF-CSTATES turn it OFF in bios. Also, download zanstates to turn off c6 package all together. I had stable CO on my 5600x values of, -15, -20, -15, +8, -11, -15 and passed core cycler 12 hour runs with all fft sizes, passes prime 95, y-cruncher, etc but upon idle would crash. Someone else mentioned this to me so I take no credit. I'll also add I'm running a 4x8 3200c14 kit of tforce dark pro bdie @4000mhz 1:1 mode. (Tuned) if you go to the ZEN overclocking spread sheet, im on there. (notes to consider, ive lowered my IOD, CCD and SOC voltages ALOT since posting this submission on that spreadsheet. but none the less, my RAM is stable, and now finally my CO values are too. (you can also use an exe called TOOL hard to find, but it will tell you your silicon quality, (if it works right, some have issues as do i) stating i have a 80 sil quality. will attach zenstates download and this "tool" as thats what its called, to my gdrive. (using tool go to DB query, then select amd v/f select number of cores etc, then hit get sil.... (also has a monitor feature but it wont stay open long due to its poll rate being so high.) as for zenstates, simply open it and go to power, then turn off package C6 state, (this will need done upon every reboot) unless you allow it to run at start up.

you can also overclock you processor etc with this. (so be careful).

  • 7GoiJ6N.png
  • Screenshot (135).png

Last edited: Jun 20, 2021

Hi everyone I'm new in this topic and i wanna make sure I'm doing something right for my system since my aim is to keep a more cooler cpu in general without damaging any component of my PC but keeping the performance if possible. Now I found this video and i wanna make sure this is actually accurate and it's safe to do I must add that I'm using a ryzen 5600x with an aorus elite b550m Mobo

20210722_230031.jpg.9d8a079fd371d216e34c8c2e17459fd3.jpg

20210722_230058.jpg.740f5f6cbf3ecaff723a974ebd21e1bd.jpg

Also I know it might be just marketing but why does it says in the warning section that undervolting may damage or shorten the life or the processor or any other system components ? So there's a risk at doing this ? someone in another forum suggested me to put these values in the pictures i added, not sure if they are right for me so i havent applied them yet.

thanks for your time !

Last edited: Jul 23, 2021

Mussels

undervolting wont kill the system, but that -30 is pretty aggressive and might be unstable

Makaveli

undervolting wont kill the system, but that -30 is pretty aggressive and might be unstable

yup if you are going use CO at all cores and not per core I would start at maybe -10 on all cores then slowly test and work my way up.

So doing undervolt won't damage the cpu or any other part at all ? Long-term speaking. And I tested what the person did on the video and it's impressive how the temps lowers very significantly while keeping the cores working as normally would, so at least under some kind of load the system seemed stable at -30 on all cores in curve optimizer, didn't got to test on idle yet so as you guys said that might be the trick to test, also say if I jusr wanna do this unfervolting through curve optimizer and keep my clocks as default without adding any more Mhz , is fine no ? Theres not some kind of tradeoff I need to perform in order to get the clocks I aim to get ?

Also why i notice that some fiddle around with the values in ppt, tdc and EDC? Is it for ocing factor ? Is it find to keep it disabled while doing curve optimizer undervolting ? Or even just leave it at default and just doing the co undervolting?

I would download and run corecycler to test for stability.

Doing this as we speak and i noticed how it's running iterations but the number of iterations are set to 1000 by default, isn't that gonna take a very long time ? I did change the runtimepercore to 60 seconds

jesdals

Hi everyone I'm new in this topic and i wanna make sure I'm doing something right for my system since my aim is to keep a more cooler cpu in general without damaging any component of my PC but keeping the performance if possible. Now I found this video and i wanna make sure this is actually accurate and it's safe to do I must add that I'm using a ryzen 5600x with an aorus elite b550m Mobo
View attachment 209419View attachment 209420 Also I know it might be just marketing but why does it says in the warning section that undervolting may damage or shorten the life or the processor or any other system components ? So there's a risk at doing this ? someone in another forum suggested me to put these values in the pictures i added, not sure if they are right for me so i havent applied them yet.

thanks for your time !

You can make pics by pressing F12 in that bios

Makaveli

Doing this as we speak and i noticed how it's running iterations but the number of iterations are set to 1000 by default, isn't that gonna take a very long time ? I did change the runtimepercore to 60 seconds

You don't have to wait that long if a core is going to fail it will do it fairly quickly.

You don't have to wait that long if a core is going to fail it will do it fairly quickly.

alright im back with results:

6minute corecycle.PNG

cinebench30min.PNG

also added the log file of the cinebench test let me know if everything is working as intended. what i did on bios was just disable pbo limits (not on manual nor motherboard )

and on curve optimizer set all cores at -30

Zach_01

Curve Optimizer is a little more complicated than just an undervolt mechanism. While negative offsets are indeed undervolt the CPU will increase speed and its PBO limits (PPT/EDC/TDC) way beyond its stock limits. This increases temp dramatically. If anyone wants to keep it as safe as possible and wants just undervolt must set the desired limits of PBO or have it completely at stock. Default limits of 5600X are: PPT: 76W TDC: 60A EDC: 90A I would consider the following as safe if cooling device can keep it at reasonable levels (80~85C max, 100% load). PPT: 90~95W TDC: 70~75A EDC: 100~110A

Max operating temp for 5000 is 90C.

Curve Optimizer is a little more complicated than just an undervolt mechanism. While negative offsets are indeed undervolt the CPU will increase speed and its PBO limits (PPT/EDC/TDC) way beyond its stock limits. This increases temp dramatically. If anyone wants to keep it as safe as possible and wants just undervolt must set the desired limits of PBO or have it completely at stock. Default limits of 5600X are: PPT: 76W TDC: 60A EDC: 90A I would consider the following as safe if cooling device can keep it at reasonable levels (80~85C max, 100% load). PPT: 90~95W TDC: 70~75A EDC: 100~110A

Max operating temp for 5000 is 90C.

Isn't disabling pbo limits like having it at stock ? Cause when I ran cinebench i had open Ryzen master and the PPT , EDC and tdc were at the values that you mentioned as stock

freeagent

Core cycler fails for me with my current settings.. I get no hardware errors, no junk files, no bluscreens, passes every single thing, except that program. I can fold, wcg, game, internet, desktop just fine.. no errors. Not a random number generator at all.. but is a kickass calculator.. all core set to -30 +200 with custom ppt tdc edc and she rips. Falls on its face with core cycler, so I suppose it is a random number generator :D The 5900X’s that are faster at superpi-32m than mine at hwbot are running at 5500MHz+

For now :D

It was difficult to tune the 5900X compared to the 5600X. That’s when I realized we aren’t controlling any limits, just manipulating them.. at least on the big CPU’s. Even on the small ones too I suppose, but there is way less going on and you can use pbo to brute force it to top clocks all the time, the big ones you have to cheat a little.

Mussels

Curve Optimizer is a little more complicated than just an undervolt mechanism. While negative offsets are indeed undervolt the CPU will increase speed and its PBO limits (PPT/EDC/TDC) way beyond its stock limits. This increases temp dramatically. If anyone wants to keep it as safe as possible and wants just undervolt must set the desired limits of PBO or have it completely at stock. Default limits of 5600X are: PPT: 76W TDC: 60A EDC: 90A I would consider the following as safe if cooling device can keep it at reasonable levels (80~85C max, 100% load). PPT: 90~95W TDC: 70~75A EDC: 100~110A

Max operating temp for 5000 is 90C.

Throwing that in the PBO link in my sig

freeagent

I would consider the following as safe if cooling device can keep it at reasonable levels (80~85C max, 100% load). PPT: 90~95W TDC: 70~75A

EDC: 100~110A

On my 5600X: PPT: 200W TDC: 140 EDC: 180

+200MHz All core set to -30 = Perfectly safe, runs cool, rips hard.

Been running -30 allcore and +50MHz pbo for 3 months now, still stable and a bit faster and cooler than stock :) Currently running PPT 45W since I run monerominer with nicehash.

Mussels

I'm gunna guess 5600x get all the cherry picked cores, you lucky undervolting bastards


Page 6

Zach_01

I'm gunna guess 5600x get all the cherry picked cores, you lucky undervolting bastards

:D Usually though, low segmented SKUs are getting the worst silicon. At least that was the case with R5 3600. If you think about it, there is no (reasonable) point for AMD to cherry pick silicon for the “low end” part of the whole series. If any CPU “needs” the best cores that would be the 5800X. I’m guessing that since 5000 series the 7nm node has improved much overall. 5600X being the only single CCD 6core(active) CPU on 8core CCD, maybe this is giving it the edge.

Just a thought I’m not going to pretend that I know facts.

Mussels

:D Usually though, low segmented SKUs are getting the worst silicon. At least that was the case with R5 3600. If you think about it, there is no (reasonable) point for AMD to cherry pick silicon for the “low end” part of the whole series. If any CPU “needs” the best cores that would be the 5800X. I’m guessing that since 5000 series the 7nm node has improved much overall. 5600X being the only single CCD 6core(active) CPU on 8core CCD, maybe this is giving it the edge.

Just a thought I’m not going to pretend that I know facts.

well, with the 5600x and 5800x starting as the same chip, the odds on the 5600x having all better cores is actually higher - only needs one weak core to not make the 5800x cut, leaving 6-7 good ones as a possibility

freeagent

Just throwing this out there, but I have a very hard time getting my 5600X stable at more than 4850.. 4900 is good for some light benching. So looking at how cores alone scale I don’t think it’s that good of a bin..

I'm gunna guess 5600x get all the cherry picked cores, you lucky undervolting bastards

It's a bit easier to UV 5600X due to lower SC speed and fewer cores :) Freeagents sample is unusual though. Getting a 5800X with 2 extra cores and 200MHz higher stock speed stable at -30 CO is a bit trickier :)

freeagent

I don’t know.. I didn’t know anything about AM4 when I bought into it.. just noob tunes that’s all.. someone should send me their cpu to play with to see if I can get it to do the same thing :D

Don’t worry you would get it back ;)

Zach_01

well, with the 5600x and 5800x starting as the same chip, the odds on the 5600x having all better cores is actually higher - only needs one weak core to not make the 5800x cut, leaving 6-7 good ones as a possibility

Comparing those two I could agree, yes.

It's a bit easier to UV 5600X due to lower SC speed and fewer cores :) Freeagents sample is unusual though. Getting a 5800X with 2 extra cores and 200MHz higher stock speed stable at -30 CO is a bit trickier :)

5600X(s) are also 8core parts (CCDs). Have 2 cores disabled that didn't make it. Defective from beginning or after through evaluation or just disabled because AMD wanted 6core CPUs. Actually you have to think every 8core CCD as exactly the same on the 7nm node. From 5600X to 5950X.

5950X gets the best CDDs and 5600X the worst(?). AMD has to segment all core performance equivalent to core count as much as possible.

Does the 5950X (2x8core) performs twice (+100%) as a 5800X (1x8core)? No...

5950X performs about +55~85% better than a 5800X but on about the same power consumption (140~142W). No doubt 5950X has the better cores and already a "high" negative V/F curve compared to 5800X by AMD.

Now, comparing the same way the 5900X (2x8core but 2x6c active) with 5600X (1x8core but 1x6c active) is a little different because of their different power consumption. 5900X has 87% higher power draw from 5600X. Does it perform 87% better? Is it by avg higher or lower than 87%? Well if you check benchmarks you'll see that a 5900X performs by avg around 90% better than 5600X. So the cores of 5900X are same or even a little better than 5600X. So far 5950X better than 5800X 5900X slightly better than 5600X 5900X lands in between the 5950X and 5800X in terms of quality (best cores) because all 3 have the same power draw 140~142W. So far 1. 5950X 2. 5900X 3. 5800X Where does 5600X goes though? Easy... 5900X has 50% more cores, and around 30~40% more performance, but the same power compared to 5800X. You can say its a big difference in quality. 5950X has 33% more cores, and around 25~30% more performance, but the same power compared to 5900X. You can say its a substantial difference in quality. So if 5600X has cores with same or slightly worst quality from 5900X then we have it all. In terms of core quality the order is this: 1. 5950X 2. 5900X 3. 5600X 4. 5800X Its not hard to understand that 5600X(s) have the best headroom for negative V/F curve from stock settings because of their low stock power draw compared to all other 5000 SKUs.

Its product segmentation.

Early on the rumor was that 5800x was unavailable and priced relatively close to the 5900x because it needed a completely working die and AMD could equally use those for a 5950x. 5900x max stock boost clock is 4950MHz I think, vs 4850MHz for the 5800x. 5800x needs 8 cores that good, while 5900x only needs 6 per die. It depends on the yield AMD is getting. Are there more chips with 6 good cores that can make 4950MHz, or more with all 8 cores working that can't quite make the 4950MHz cut? My guess would be the former. Either way, I think it is reasonable to say that a 5800x will have a better die than a 5600x. In the worst case, you could disable the two worst cores on a 5800x to make higher clock speeds :-D (Although I don't know a way to pick which actual cores are turned off)

The other factor here (as mentioned already) is that a CO of e.g. -5 on a 5800x is not equivalent to a CO of -5 on a 5600x, since the curve set from the factory is not the same. As I understand it, the curve is actually set per cpu, so even from one 5800x to another, -5 does not mean the core will be getting the same voltage. -5 on two cores on the same chip doesn't even mean that.

With OCCT v9, how do you setup for testing CO stability?

Nordic

We just need an automated way to tune in the curve optimiser and we'll allllll be happy.

I just started messing with core optimizer and isn't this the truth. It would be really nice if a script could change the core optimizer value and stress test from the bios. It would only require simple if then logic. 1) Set Core X to -5 2) Stress test Core X for Y minutes 3) If the test completed successfully, set Core X an additional -5, and repeat step 2. If the test failed, set Core X +1, and repeat Step 2. Do not test the same value twice. Blue Dragons testing script is helpful but not the whole package. It seems like it would be trivially easy to run some sort of simple stress test at the bios level and let the motherboard figure this out automatically. Or in the OS if we could access the curve optimizer settings.

I do have a question. Is it worth trying to tune curve optimizer if I am on air cooling and thermally limited?

tabascosauz

I just started messing with core optimizer and isn't this the truth. It would be really nice if a script could change the core optimizer value and stress test from the bios. It would only require simple if then logic. 1) Set Core X to -5 2) Stress test Core X for Y minutes 3) If the test completed successfully, set Core X an additional -5, and repeat step 2. If the test failed, set Core X +1, and repeat Step 2. Do not test the same value twice. Blue Dragons testing script is helpful but not the whole package. It seems like it would be trivially easy to run some sort of simple stress test at the bios level and let the motherboard figure this out automatically. Or in the OS if we could access the curve optimizer settings.

I do have a question. Is it worth trying to tune curve optimizer if I am on air cooling and thermally limited?

Technically AMD probably has access to CO through Ryzen Master since it's all under the same submenu, but we all know that it's not the most polished/reliable software around, frankly wouldn't trust it even if it did... People entrusted their Renoir APUs to 1usmus CTR and it ran 1.5V all core through them to verify an "overclock" soooo I would just stick to BIOS and forget about this software convenience BS. If you want to be reasonably stable then the default corecycler settings script is fine, so 6 minutes per core. Run through it 2-3 times and you're good, all in a day's work. But I still stick to the 68 minute All FFT config, insane amount of time to test but at least I know it's 110% stable. Been running my 5900X curve and it still holds after switching to the B550 Unify-X. Still haven't proven it wrong.

Then there's also the OCCT test that does basically the same thing but is a bit better structured/polished. I don't use it yet but @freeagent likes it

Might explore some different testing methodologies for my 5600G. Currently at -15 no problems but fine tuning still needed.

freeagent

I use OCCT, Linpack Xtreme, TM5, Superpi 32m, y-cruncher, and a few others.. If I were to run core cycler my system would fail in seconds lol.

So its absolutely stable, but its not..

Nordic

I am running core cycler 1 minutes per core starting from a value of negative 10. If it passes the test, I increase the value by 1. If it fails I decrease it by 1. When I find the highest value a core can use, I have core cycler ignore that core. Usually the cores that fail, fail immediately. This method is painfully slow. It would be really nice if I could automate it. So far after 5 runs I have only found the final value for Core 9. Oddly enough, the cores I thought were my best are the ones that keep failing. After I have completed this process, I should have a fairly good idea of what my curve looks like. I intend to let core cycler run for an hour per core and use other stress tests to ensure stability after.

I am unclear on a few things though. I don't know if I am going to have to go through this process again when I install water cooling. I am thermally limited right now. I was also planning on trying out x2 scaling to see if that makes a difference, but that isn't worth trying until I have water cooling anyway.

tabascosauz

Oddly enough, the cores I thought were my best are the ones that keep failing.

Welcome to 2CCD haha, sample size still too small but I suspect average 5900X/5950X don't usually have much real undervolt headroom on their 2 preferred cores Mine looks like this:

5900x curve optimizer.png

For preliminary testing I'd say default 6min is worth it, given that it cycles every few seconds. Anything stable on default config already shouldn't give you major problems in daily usage, just give it a couple loops as stability may not be consistent. Then if you REALLY want to ensure stability do the All FFT hour or long OCCT testing, but I'm just OCD :D

Nordic

Cores 0,1, and 4 are down to CO -7 and going down. Core 9 is settled at CO -11. The rest of the cores up to -16 and just keep going. It would be really nice if this was automated. Not sure how far this will go.

It seemed stable at all cores -10 too. I was considering leaving it there because it worked fine and a 15% multi threaded improvement is not bad at all. Good thing I am stability testing.

tabascosauz

Cores 0,1, and 4 are down to CO -7 and going down. Core 9 is settled at CO -11. The rest of the cores up to -16 and just keep going. It would be really nice if this was automated. Not sure how far this will go.

It seemed stable at all cores -10 too. I was considering leaving it there because it worked fine and a 15% multi threaded improvement is not bad at all. Good thing I am stability testing.

On the 6min config I was "stable" at -7 and -10 on 0 and 1. No real instability symptoms from any of the cores so clearly usable, honestly, but once I went to the exhaustive 68min config it was quickly apparent that those two cores could go no further than -2 and -7. But yeah, unless you get really lucky and can push more than -15 on those preferred cores only, little ST benefit since the algorithm cannot be forced to use any other cores. As in the picture Core 2 easily becomes the best core after -30 CO, but I cannot ever use it. MT gets a big boost but might still be hotter at iso-voltage

Optimumtech said that CO benefits ST temps greatly, but he was testing at -15 to -30. Zero difference here at -2 and -7.

freeagent

I should take the time to play with it like that.. its just so tedious lol..

Nordic

I should take the time to play with it like that.. its just so tedious lol..

Extremely tedious. I have a timer going so that I know when to come back, see what failed, input new values into the bios, and start all over. In the mean time I have made tremendous progress on a yard project. I am very happy that I have Core Cycler to automate some of the steps.

freeagent

Honestly.. I just used SuperPi 32M to tune my single core boost, every other program I use will just bring all the cores down to one speed and float around by a few MHz, that's why I didn't try to exploit each core because I didn't think windows was smart enough to use the cores like that.

It was not that easy to do lol.. those guys have some skill for sure :toast:

Nordic

It took several hours of the smallest amount of stability testing but I finally have rough numbers.

Core 0

Core 1

Core 2

Core 3

Core 4

Core 5

Core 6

Core 7

Core 8

Core 9

Core 10

Core 11

Core 12

Core 13

Core 14

Core 15

What is the quickest way to test CO stability? Get a rough indication? Considering trying to get +200 pbo stable. Everything at -30 allcore and +50 is rock stable since May, but I get rare reboots if I try +100 or more. +200 was stable for 24H, then suddenly pc restartet at medium load.

tabascosauz

It took several hours of the smallest amount of stability testing but I finally have rough numbers.

Default config test? That's some commitment right there :toast: took me the better part of a month to work mine out and I was still cutting corners by sticking to multiples of 5 on ten of the cores Positive offset is some kind of mystical beast on these chips lol, this is probably the 2nd time I've ever seen it

What is the quickest way to test CO stability? Get a rough indication? Considering trying to get +200 pbo stable. Everything at -30 allcore and +50 is rock stable since May, but I get rare reboots if I try +100 or more. +200 was stable for 24H, then suddenly pc restartet at medium load.

Corecycler script, don't change the config from default, close all possible background apps and disconnect from internet. Run the script, it tests each core for 6 minutes then cycles to the next one. Let it keep going and observe every once in a while until it's gone through every core 2-3 times. Get thru 3 iterations without errors and I'd say that's good enough for most ppl. Shouldn't take long at all on a 6-core.

I'll probably be doing the same tonight for my 5600G, sitting on lazy -15 all core right now but I need to test.

Default config test? That's some commitment right there :toast: took me the better part of a month to work mine out and I was still cutting corners by sticking to multiples of 5 on ten of the cores Positive offset is some kind of mystical beast on these chips lol, this is probably the 2nd time I've ever seen it Corecycler script, don't change the config from default, close all possible background apps and disconnect from internet. Run the script, it tests each core for 6 minutes then cycles to the next one. Let it keep going and observe every once in a while until it's gone through every core 2-3 times. Get thru 3 iterations without errors and I'd say that's good enough for most ppl. Shouldn't take long at all on a 6-core.

I'll probably be doing the same tonight for my 5600G, sitting on lazy -15 all core right now but I need to test.

Went through 2 full cycles now, all at -30, except one core at -28, had error in 1 cycle at one core, but -28 fixed it. Satisfied with performance :)

freeagent

With the stock clock range I get errors @ -30 CO on my 5900. After that I just didn't care too much. I don't think windows uses the cores the way core cycler does, not even remotely..

Nordic

Default config test? That's some commitment right there :toast: took me the better part of a month to work mine out and I was still cutting corners by sticking to multiples of 5 on ten of the cores

Positive offset is some kind of mystical beast on these chips lol, this is probably the 2nd time I've ever seen it

As per my earlier post, I ran core cycler at 1 minute per core until it didn't fail. When it did fail, it almost always failed within seconds. This gave me the very rough numbers I shared above. While I slept last night, I ran Core Cycler 20 minutes per core and it only found an error on Core 0. I intend to Run Core Cycler until I get no errors anymore then try OCCT.

To my knowledge, I am the third invididual in this long thread who had a positive offset. I suspect I needed a positive offset because I have LLC set to 3.


Page 7

freeagent

You should only need LLC when running manual all core clocks no? Like 1 clock 1 voltage?

Nordic

You should only need LLC when running manual all core clocks no? Like 1 clock 1 voltage?

Before I started down this path, I read several 5000's series overclocking guides, watched several overclocking guide videos, and even read this WHOLE thread. Several of the guides recommended LLC + Curve Optimizer both lower voltage. Given that I now have one core in the positive, I think the LLC and PBO might work better for a golden chip. Mine is not that.

freeagent

I should get on that.. I haven't watched any videos, just a little on reddit and some scraps of info that google gave me.

Nordic

I should get on that.. I haven't watched any videos, just a little on reddit and some scraps of info that google gave me.

This curve optimizer stuff is not inutitive. I was mainly searching for a way to test each core efficiently. To be honest, I found the best information in scattered throughout this thread. I didn't see anything about Core Cycler elsewhere.

freeagent

This curve optimizer stuff is not inutitive.

Well.. its just tedious. Once I started getting closer to the 0 mark in CO I gave up lol. It does what I need it to do.. no errors anywhere, no bad files, no bad logs.. seems ok to me! The heaviest of loads like Linpack run at 4500, a little lighter with OCCT at about 4750, TM5 runs at 4850, SuperPi runs at 5100-5150, Cinebench runs at about 4700, F@H runs at 4550-4650. If my limits aren't where they should be I don't see that 5150 and if I do it crashes. If it can hold that 5150 it should be good everywhere else.. maybe.

Nordic

I almost left curve optimizer alone for much better than stock performance. I really wanted the pleasure of having tuned the chip. Heavy loads aren't helpful for testing your CO. By everything I have read, you want a light load that minimizes heat encouraging your cpu to hit the highest core clocks. Heavier loads may run hotter, preventing it from reaching higher core clocks where edge cases might be found. Oddly enough, one of the best stability tests is windows recovery. Core Cycler makes it easy by automating the testing of each core.

If it is too tedious for you, that's fine. Enjoy the performance and go play games rather than tinker and tune.

freeagent

Heavy loads aren't helpful for testing your CO. By everything I have read, you want a light load that minimizes heat encouraging your cpu to hit the highest core clocks. Heavier loads may run hotter, preventing it from reaching higher core clocks where edge cases might be found.

Normally that would be true. And you would see that while setting your PPT TDC and EDC. Its about balance with the bigger chips. With my 5600 I set it to my max all core clock and let er rip with Linpack. I took note of what it was pulling, and added those values in bios. I was then able to set -30 and she boosts and holds 4850 for just about everything. Heavy stuff still dips to 4650-4750, but not under.. shit now I have to plug it in again to make sure I am not full of shit. Pretty sure it doesn't go under 4650.. my 5900 was way different, the way I did my 5600 was not working at all for the 5900. I had to take to google for that one :D

Nordic

Normally that would be true. And you would see that while setting your PPT TDC and EDC. Its about balance with the bigger chips. With my 5600 I set it to my max all core clock and let er rip with Linpack. I took note of what it was pulling, and added those values in bios. I was then able to set -30 and she boosts and holds 4850 for just about everything. Heavy stuff still dips to 4650-4750, but not under.. shit now I have to plug it in again to make sure I am not full of shit. Pretty sure it doesn't go under 4650.. my 5900 was way different, the way I did my 5600 was not working at all for the 5900. I had to take to google for that one :D

With curve optimizer you are telling each core to run at a lower voltage allowing that core to boost higher. At a certain point, you won't have enough voltage to run the peak boost clocks stable. That is why when messing with CO, you want a light load. You want a work load that can push the cpu to peak boost clocks because you may have instability there. That is why CO is so tedious. You need to test each core at its peak boost clock which is a moving target. To make things worse for me, I will probably need to do all this stability testing again when I begin watercooling. Cooler temperatures will allow PBO to push the clocks further and more consistently. Settings that may have been stable on air because I couldn't sustain high enough boost clocks to trigger an error. EDIT:

You should only need LLC when running manual all core clocks no? Like 1 clock 1 voltage?

I tried LLC level 3 on and off. It didn't matter for benchmarks or stability. I still have a positive offset with and without it on. I am going to leave it off going forward.

With the stock clock range I get errors @ -30 CO on my 5900. After that I just didn't care too much. I don't think windows uses the cores the way core cycler does, not even remotely..

The second best core according to windows is actually my worst and the only one that can't run -30 at +200. Dunno where win get their info from...

Nordic

One thing I have learned is if one of the cores fails to pass a stability check, go down by at least two. Going up or down by one is a pointless endeavor. Having gone through this process, I think I might prefer the strategy I read earlier on in the thread to only move in increments of 5. If - 15 fails go down to - 10.

My two best cores according to Ryzen master both have positive offsets. Why can't I ever win the silicone lottery?

One thing I have learned is if one of the cores fails to pass a stability check, go down by at least two. Going up or down by one is a pointless endeavor. Having gone through this process, I think I might prefer the strategy I read earlier on in the thread to only move in increments of 5. If - 15 fails go down to - 10.

My two best cores according to Ryzen master both have positive offsets. Why can't I ever win the silicone lottery?

I apparently won the silicon lottery on my 5600X which does +200 at -30 CO (except one core at -28) and IF does 2066, BUT my ram is shit and barely does 4000cl16 at 1.47V at 2T. Guess you can`t win all ;) My Ryzen 3600 was horrible, best allcore it could do was [email protected], I never saw it boosting to 4.2 and all core stock it ran 3.95 :p

tabascosauz

One thing I have learned is if one of the cores fails to pass a stability check, go down by at least two. Going up or down by one is a pointless endeavor. Having gone through this process, I think I might prefer the strategy I read earlier on in the thread to only move in increments of 5. If - 15 fails go down to - 10.

My two best cores according to Ryzen master both have positive offsets. Why can't I ever win the silicone lottery?

Hence why 10 of my cores are multiples of 5 :D My 3700X was bronze (if bronze = soggy cardboard), my 4650G was bronze, my 5900X is bronze, and I can't be arsed to find out where the all-core for my 5600G is.

Side note, I am a little bit crazy with requiring 4x or 5x iterations of 68 minute All FFTs for every core. However, I'm not entirely crazy. When Cezanne is unstable, it starts rattling off Cache Hierarchy event 19. And I DO MEAN rattling off. I never saw this level of error reporting on the 5900X, must be an APU thing.

You don't have to be visibly unstable in Corecycler to start collecting these - I'm still testing the default 6 minute config of Huge FFTs, and this is on like Iteration #6 with no errors in the test. So be very discerning as to the configuration you use in Corecycler, that's why the config.ini is there.

cache hierarchy.png

cache hierarchy 2.png

There's probably 200-300 WHEAs in there from the past 2 days alone. Every single one of them is Cache Hierarchy, so it's pretty obvious what's going on.

When I ran overnight, I think Core 1 survived something like 24 iterations before it started erroring out in the test. But we all know that's bullshit, because Event Viewer looks like a money printer. That's why I'm going back to the 68 minute config after the default testing.

outpt

given up on core cycler. the true test is youtube/SoTTR. when playing with CO i use the SWAG principle aka scientific wild ass guess. still get all kinds of errors and so forth. if it doesn't bloo screen i ain't going to check event viewer.

Nordic

My Core 0 and Core 1 are both at +15 now. Core Cycler keeps flagging them as having errors. This is getting absurd. Either I have the worst CPU ever or the testing methodology is flawed.

tabascosauz

My Core 0 and Core 1 are both at +15 now. Core Cycler keeps flagging them as having errors. This is getting absurd. Either I have the worst CPU ever or the testing methodology is flawed.

What does OCCT say? If you're running PBO with +200MHz override I suppose you might have less impressive offsets.

Overclocking my 5600x don't work, leave PBO on auto and forget it, same as my RAM, utter shite RAM I have.

Jumping into this thread since I recently got a 5900x. I went through the (painful) process of going one core at a time, starting at -30 and walking it back by 5 until each one could go through an hour of single core OCCT without errors. Took forever, but found those limits. Despite breezing through stress tests, my system was nearly guaranteed to crash within 90 seconds of sitting idle. The actual WHEA errors aren't much help, they tend to point to the same core regardless of if I set that core to -20, 0, or +5, so it seems to be a misleading manifestation of instability elsewhere. Without a better indication of which core is causing the trouble, I decided I would just blindly add +1 to every core's offset whenever it crashed. I've had to do that 3 times so far but I think I'm just about stable now.

Is there a better way to do this? Stress testing doesn't seem useful because if power consumption is on a curve, my issues seem to be toward the bottom of the curve so there's not enough voltage when the core is doing very little.

Nordic

What does OCCT say? If you're running PBO with +200MHz override I suppose you might have less impressive offsets.

I haven't tried OCCT. I am not even running an offset. It is only boosting upto about 4900mhz.When I get back home I am going to turn everything to stock and see if I still have errors. Some cores can boost upto 5050mhz at -10 and lower in curve optimizer.

Is there a better way to do this?

We have Core Cycler, OCCT, and the wait and see approach.

tabascosauz

Overclocking my 5600x don't work, leave PBO on auto and forget it, same as my RAM, utter shite RAM I have.

Twas the story with my 3700X and 4650G, effects of [very] poor silicon quality.

Jumping into this thread since I recently got a 5900x. I went through the (painful) process of going one core at a time, starting at -30 and walking it back by 5 until each one could go through an hour of single core OCCT without errors. Took forever, but found those limits. Despite breezing through stress tests, my system was nearly guaranteed to crash within 90 seconds of sitting idle. The actual WHEA errors aren't much help, they tend to point to the same core regardless of if I set that core to -20, 0, or +5, so it seems to be a misleading manifestation of instability elsewhere. Without a better indication of which core is causing the trouble, I decided I would just blindly add +1 to every core's offset whenever it crashed. I've had to do that 3 times so far but I think I'm just about stable now.

Is there a better way to do this? Stress testing doesn't seem useful because if power consumption is on a curve, my issues seem to be toward the bottom of the curve so there's not enough voltage when the core is doing very little.

Welcome to TPU :lovetpu: What are the actual WHEA errors though? Cache Hierarchy/L1 error should be the ones that result from unstable Curve Optimizer. Bus/Interconnect and Unknown are usually symptoms of Infinity Fabric, so it'd help to know what IF/RAM setup you're running to rule that out first. But if it is a cores issue, from what you're describing it sounds like an idle voltage issue. Laziest fix is to disable Global C-states and Power Supply Low Current Idle (set to Typical), but really just a band-aid remedy. Have you tried testing stability at no PBO and no offset?

I haven't tried OCCT. I am not even running an offset. It is only boosting upto about 4900mhz.When I get back home I am going to turn everything to stock and see if I still have errors. Some cores can boost upto 5050mhz at -10 and lower in curve optimizer.

We have Core Cycler, OCCT, and the wait and see approach.

Perhaps run a quick diagnosis test in CTR and see what it says on quality? Dunno if all-core quality correlates with ST quality, but it's worth a shot. If it comes back with a Bronze sample and pretty low recommendations for all-core OC, might need to lower expectations a bit.

Actually now that I think about it, CTR should be a decent test. All-core stability is usually dictated and limited by your worst core(s), those are also the ones you will need to pay attention to in CO.

you need to be at stock for CTR, CO will skew your results

Last edited: Aug 10, 2021

droopyRO

For my 5600X it was simple. Set it to all cores -25, fail. -20 fail again, -15 success, it has been running like this since Gigabyte released the AGESA update.

Occt is quick. In less than one hour I had all dialled in, been stable since.

What are the actual WHEA errors though? Cache Hierarchy/L1 error should be the ones that result from unstable Curve Optimizer. Bus/Interconnect and Unknown are usually symptoms of Infinity Fabric, so it'd help to know what IF/RAM setup you're running to rule that out first.

Thanks for explaining that because I've never seen this stated anywhere before but it matches my experience in trying to tweak this. The cache hierarchy errors clear up when reducing the undervolt for the stated core but the bus/interconnect ones stick around and the listed core doesn't seem to matter. I'm at 1800/3600 IF/RAM which doesn't seem all that aggressive. Things seem stable at those settings with PBO and CO disabled, or even with PBO set to mobo and CO disabled, so I think its a matter of "wait and see"-ing until I get the right offset values dialed in. Disabling the C states/idle power feels like a bad solution and I think I would just give up on CO before going that route.

tabascosauz

Thanks for explaining that because I've never seen this stated anywhere before but it matches my experience in trying to tweak this. The cache hierarchy errors clear up when reducing the undervolt for the stated core but the bus/interconnect ones stick around and the listed core doesn't seem to matter. I'm at 1800/3600 IF/RAM which doesn't seem all that aggressive. Things seem stable at those settings with PBO and CO disabled, or even with PBO set to mobo and CO disabled, so I think its a matter of "wait and see"-ing until I get the right offset values dialed in. Disabling the C states/idle power feels like a bad solution and I think I would just give up on CO before going that route.

Disabling Cstates certainly isn't ideal but unless a new BIOS can magically solve the problem or you're willing to RMA there aren't many other options. My former 3700X was an absolute turd that required Cstates disabled on both the cores and IF front. Worked without a hitch as long as that was done. You should leave Curve Optimizer to the end, work out your RAM and IF first, you need to be assured that any instability you encounter will be cores-related. 3600 isn't that aggressive (I can run 3600 at just 1.0375V VSOC), but I remember Bus/Interconnect being common on 2020 production 5900X/5950X. What's the batch number on yours?

On the IF part, you can head over to the Ryzen thread and we can see if you post up a Zentimings screenshot

Nordic

Perhaps run a quick diagnosis test in CTR and see what it says on quality? Dunno if all-core quality correlates with ST quality, but it's worth a shot. If it comes back with a Bronze sample and pretty low recommendations for all-core OC, might need to lower expectations a bit

Before messing with any bios overclocking, I ran CTR. It says silver sample. Hence why I think there is something wrong with my testing methodology.

tabascosauz

Before messing with any bios overclocking, I ran CTR. It says silver sample. Hence why I think there is something wrong with my testing methodology.

Probably best to use PBO but manually set default power limits (142/95/140), should have the same effect as PBO off. But leave boost override at +200MHz. You'll still be able to push ST boost clocks to their max with default power limits. Then if you want higher MT perf you can uncap the power limits later when you know that ST is for sure stable and all you have to test is all-core (which is trivially easy). Sometimes high PBO power limits can tank ST performance. Still not sure why this is.

Are you still running only 1 minute iterations? It's not really long enough of a window to do much.


Page 8

Nordic

Probably best to use PBO but manually set default power limits (142/95/140), should have the same effect as PBO off. But leave boost override at +200MHz. You'll still be able to push ST boost clocks to their max with default power limits. Then if you want higher MT perf you can uncap the power limits later when you know that ST is for sure stable and all you have to test is all-core (which is trivially easy). Sometimes high PBO power limits can tank ST performance. Still not sure why this is.

Are you still running only 1 minute iterations? It's not really long enough of a window to do much.

I haven't even been increasing the boost override because I have not even been hitting the limit.

I only ran 1 minute tests to get rough numbers. I have been doing 20 minute runs since to really fine tune the numbers.

tabascosauz

I haven't even been increasing the boost override because I have not even been hitting the limit.

I only ran 1 minute tests to get rough numbers. I have been doing 20 minute runs since to really fine tune the numbers.

The config file has estimates for the time to get through a full cycle of different FFT sizes:

corecycler fft sizes.png

Nordic

I haven't tried OCCT. I am not even running an offset. It is only boosting upto about 4900mhz.When I get back home I am going to turn everything to stock and see if I still have errors. Some cores can boost upto 5050mhz at -10 and lower in curve optimizer.

I set my bios back to their defaults. Everything is at default. I then set up core cycler to run over night at 20 minutes a core. Core Cycler is still finding errors are Core 0 and Core 1. These are the same cores that needed positive offsets earlier. It seems I either have some A) Inherent instability somewhere or B) Core Cycler is a faulty testing mechanism.

When I get a chance I am going to try OCCT and see if it finds errors too.

freeagent

Ok.. so.. this whole time that I have been saying I run 200 130 130 was a mistake. I had Asus performance enhancement enabled, and that brought tdc to 140 and edc to 180. Looking at my notes these are the same ppt tdc edc requirements to run the shit out of my 5600x. Soo.. I ran pi 32m a few times at 200 140 180 -30 +200 and it worked just fine, hit top boost better than I had it before, at least for single core stuff.. Not sure about multicore yet but it should be good I think..

pi.PNG

outpt

200,140,180 on a 5600x??????????????!

jesdals

That is some fantastic tuning or an awesome chip. Are those clocks seen during all-core loads, and is clock stretching not an issue on Zen3?

During all core benchmark like CPU-z you will se all core speeds up to 4600MHz and single up to 5150Mhz

Last edited: Aug 23, 2021

jesdals

New AMD chipset driver may take the top of the boost clocks - went from 5250MHz on best core to 5189MHz and the rest locked at 5015MHz

jesdals

New AMD chipset driver may take the top of the boost clocks - went from 5250MHz on best core to 5189MHz and the rest locked at 5015MHz

After a coupple of days I am seeing the same boost behavior - do not know if there is some training in the newest chipset driver?
Have seen up to 5250MHz again on core 0

  • 1630608560558.png

jesdals

Any one else using both AMD CPU and GPU haveing issues with the latest driver for GPU and CPU. I am getting hard resets - especially in idle situations

Makaveli

There seems to be a new chipset drivers that enables support for windows 11 have you tried it?

droopyRO

Any one else using both AMD CPU and GPU haveing issues with the latest driver for GPU and CPU. I am getting hard resets - especially in idle situations

I do. I even RMA'ed my 5600X yesterday. I was on my old 2700X for two days, i decided that the CPU was broken somehow. But then last night i walked away for about 10 minutes from my PC. It was at idle, with Firefox opened with a few tabs. I returned to see it just restarted, WHEA error in EventViewer just like my 5600X did. Bios is at stock, no settings were made, especialy no curve optimization since there is none for 2700X. I now suspect the AMD drivers or the motherboard. Only two things that are making sense.

@Makaveli i update two days ago. The WHEA error i got was with them installed.

jesdals

I do. I even RMA'ed my 5600X yesterday. I was on my old 2700X for two days, i decided that the CPU was broken somehow. But then last night i walked away for about 10 minutes from my PC. It was at idle, with Firefox opened with a few tabs. I returned to see it just restarted, WHEA error in EventViewer just like my 5600X did. Bios is at stock, no settings were made, especialy no curve optimization since there is none for 2700X. I now suspect the AMD drivers or the motherboard. Only two things that are making sense.

@Makaveli i update two days ago. The WHEA error i got was with them installed.

Well we may have to reinstall windows - hope for a better fix to the bug

There seems to be a new chipset drivers that enables support for windows 11 have you tried it?

Still on windows 10

Makaveli

Well we may have to reinstall windows - hope for a better fix to the bug

Still on windows 10

Its a windows 10 driver just has support for Win 11. i'm still on the previous version

1631732586824.png

New version is

1631732609628.png

I'm currently using a bus speed bump to get a little more out this 5600x however most programs says 100.5-100.9 when I sent 101 in bios :/

I usually have spread spectrum enable just because I have a digital Tv tuner in this and with out it on a lose channels.

Makaveli

I would be careful with pushing that bus speed

And are you using PBO and CO?

jesdals

I'm currently using a bus speed bump to get a little more out this 5600x however most programs says 100.5-100.9 when I sent 101 in bios :/

I usually have spread spectrum enable just because I have a digital Tv tuner in this and with out it on a lose channels.

I have had some performance in 100.01 to 100.1 settings but different bios versions is more or less tolerant - stability issues do often relate to these settings

All I do is set a -15 all core and bump up my PBO limits a little to 165 PPT, 115 TDC and 150 EDC. 5900X does not scale past 170W. Went from 4.1 all core to ~4.5-4.6 GHz. Single core now sitting nicely at 5.05-5.1 GHz.

Octopuss

So you completely disregard the differences between cores and just blindly punch in some numbers. That's not how overclocking is done.

Makaveli

So you completely disregard the differences between cores and just blindly punch in some numbers. That's not how overclocking is done.

I would say doing all core -15 is a good starting point. However you will get the most benefit with per core tuning. This is just abit more time consuming but worth it.

outpt

So you completely disregard the differences between cores and just blindly punch in some numbers. That's not how overclocking is done.

Not bad for being blind. What happens with one or two eyes open?


Page 9

Mussels

So you completely disregard the differences between cores and just blindly punch in some numbers. That's not how overclocking is done.

It's how i do it, i dont have the time for per core testing, sadly.

outpt

Guy on YouTube called BOSMANG has a great guide to pbo tuning. Check it out
Have pen and paper

Octopuss

My 5800X has very different cores, and the offsets I can use without the usual stress testing programs throwing out errors vary wildly between just -5 and -25, just saying (and I still have no idea whether I tested this enough, mostly sticking to Prime95).
Sometimes you get lucky and get no bad side effects, and sometimes you get random weird crashes or program errors that aren't clearly traceable to anything. Just saying.

Makaveli

My 5800X has very different cores, and the offsets I can use without the usual stress testing programs throwing out errors vary wildly between just -5 and -25, just saying (and I still have no idea whether I tested this enough, mostly sticking to Prime95).
Sometimes you get lucky and get no bad side effects, and sometimes you get random weird crashes or program errors that aren't clearly traceable to anything. Just saying.

I recommend using corecycler for per core testing with CO.

Mussels

Just for relevance here: All my core testing was thrown out the window because my stability was erratic due to 1. Melting PCI-E cables 2. My android TV spamming the network during sleep, giving windows seizures Both of those gave false positives for instability, and all that testing time was wasted...

So now i just lower my all core by 1 every week, and see if anything crashes

I would say doing all core -15 is a good starting point. However you will get the most benefit with per core tuning. This is just abit more time consuming but worth it.

that's basically where I started so far too.

I would be careful with pushing that bus speed

And are you using PBO and CO?

Yes for now only, because I have 5600X and not a 5800X. Max Pbo is limited on 5600X

Not bad for being blind. What happens with one or two eyes open?

be one with ry-zen

It's how i do it, i dont have the time for per core testing, sadly.

^ I haven't had time my self either I work nights

Guy on YouTube called BOSMANG has a great guide to pbo tuning. Check it out
Have pen and paper

can you post the video please ? Mussel what is android TV ?, if it's what I think it is it's, probably not as good as hardware tuner like I have.

I'm using old pcix 1x hauppauge TV tuner

droopyRO

Well we may have to reinstall windows - hope for a better fix to the bug

Still on windows 10

Did you reinstall, are you still having the issue ? I did a fresh install on a formated SSD. Same issue i have. Only now a new error appeared, it gives two errors at the same time. All BIOS settings are stock, except to enable XMP. I tried to disable CPB and PBO, also i tried to set the LLC to medium. I will now try another motherboard. And if that fails too then i am out of ideas and i will switch back to my old i5 8600K :/

Reported by component: Processor Core Error Source: Machine Check Exception Error Type: Cache Hierarchy Error

Processor APIC ID: 3

Reported by component: Processor Core Error Source: Machine Check Exception Error Type: Bus/Interconnect Error

Processor APIC ID: 0

Makaveli

Did you reinstall, are you still having the issue ?

I did a fresh install on a formated SSD. Same issue i have. Only now a new error appeared, it gives two errors at the same time. All BIOS settings are stock, except to enable XMP. I tried to disable CPB and PBO, also i tried to set the LLC to medium. I will now try another motherboard. And if that fails too then i am out of ideas and i will switch back to my old i5 8600K :/

Looks like its Core 3 and 0. If you are getting these error at stock and your bios is current., I would still run corecycler.

You may have to rma.

droopyRO

This happens with my 2700X. I RMA'ed my 5600X a few days ago. Could it be the motherboard ? it only happens at idle while i am away from the PC and no programs/tasks are running except opened FIrefox tabs. It all started about a month ago.

Makaveli

This happens with my 2700X. I RMA'ed my 5600X a few days ago. Could it be the motherboard ? it only happens at idle while i am away from the PC and no programs/tasks are running except opened FIrefox tabs. It all started about a month ago.

What bios version are you running on this board?

Ferrum Master

This happens with my 2700X. I RMA'ed my 5600X a few days ago. Could it be the motherboard ? it only happens at idle while i am away from the PC and no programs/tasks are running except opened FIrefox tabs. It all started about a month ago.

Post your real RAM timings and modes. What IC modules? Then simply start rising up voltages. Start with CPU voltage. It could be the LLC is acting weird. Use the offset. If it doesn't work you should up the uncore parts.

Last time is used GB was on X99 and it also was not capable of working stable at stock settings... so think about it... only maker with whom I had these kind of issues was DFI Lanparties, but those were not ever designed to run at stock.

droopyRO

What bios version are you running on this board?

F34, F35, and now F36c. All have the same issue with both the 5600x and the 2700x. See the video i made for detail. All of those errors were n the 5600X chip that is now RMA.

@Ferrum Master all stock, no BIOS settings as i said before. And it started on the 19th of August according to my WHEA error logs. I have this mobo and CPU combo since November 2020. And if i have to manually set something in the BIOS then it is clearly something wrong with it. IMHO no user or poweruser should have to do it.

Makaveli

This is really random to all of a sudden be happening from August 19th.

Did you install any KB windows updates on this date?

Ferrum Master

F34, F35, and now F36c. All have the same issue with both the 5600x and the 2700x. See the video i made for detail. All of those errors were n the 5600X chip that is now RMA.

@Ferrum Master all stock, no BIOS settings as i said before. And it started on the 19th of August according to my WHEA error logs. I have this mobo and CPU combo since November 2020. And if i have to manually set something in the BIOS then it is clearly something wrong with it. IMHO no user or poweruser should have to do it.

My RAM stock timings are absolutely unusable on AMD as they were made for an Intel platform years ago for example. They boot, but they are totally not PRIME stable.

droopyRO

This is really random to all of a sudden be happening from August 19th.

Did you install any KB windows updates on this date?

I don't remember. But that install is gone. Formated the SSD and installed the latest Windows 10 version and updated it.

My RAM stock timings are absolutely unusable on AMD

They worked fine on this mobo since August 2020 with the 2700X and since November with the 5600X. Also tried with a different RAM kit, same issue. Also, i just finished a CoreCycler run, it looks ok to me, i bumped the RAM speed a bit as the Gigabyte mobo already overvolts the RAM to 1.38V even if i set it to 1.35 manualy in the BIOS. I will let it get in to idle and standby and see if it is stable.

EDIT: i swaped the X570 AORUS ELITE with my Tomahawk MAX B450. Same everyting, let's see if i still have the issue or i have found the culprit. The other and only thing i could think of are the AMD drivers or the PSU, but the latter is bought in November 2020 and is a Seasonic Core 650W.

Last edited: Sep 19, 2021

Mussels

Yeah you've got unstable RAM just cause it worked once doesnt mean its not dead now... hardware can just die

WHEA errors are usually CPU voltage related, or unstable RAM... hell i had a faulty corsair RAM stick that i kept around for like 3 years, as i didnt realise it was the cause of the problems (it only crashed at idle/low load, and only after it had warmed up - eventually i ran memtest overnight and caught which stick it was and smashed it to bits)

droopyRO

Two RAM kits that worked fine decided to die at the same time. One in my system and the other that was in a box in a drawer. Two RAM kits from two vendors and that they started to throw WHEA errors on the 19th of August and were running fine until then ? There are more chances to be hit by outerspace particles :) than two RAM kits dying on the same mobo in the span of a couple of days without overclock.
Yesterday i replaced the X570 Elite with a Tomahawk MAX. So far so good, i will run this MSI board for a few days and if it is stable i will RMA the Gigabyte one. If not, i will fiddle with the RAM. The other kit i will test on a Intel machine when i have the time.

Two RAM kits that worked fine decided to die at the same time. One in my system and the other that was in a box in a drawer. Two RAM kits from two vendors and that they started to throw WHEA errors on the 19th of August and were running fine until then ? There are more chances to be hit by outerspace particles :) than two RAM kits dying on the same mobo in the span of a couple of days without overclock.
Yesterday i replaced the X570 Elite with a Tomahawk MAX. So far so good, i will run this MSI board for a few days and if it is stable i will RMA the Gigabyte one. If not, i will fiddle with the RAM. The other kit i will test on a Intel machine when i have the time.

PS: while diggin on the net i found other people with the same issue as mine on AMD hardware https://community.amd.com/t5/proces...get-a-quot-silent-fix-quot/td-p/471392/page/5 https://community.amd.com/t5/proces...restarting-whea-logger-id/td-p/423321/page/94

Updated bios or chipsetdrivers? Wheas is usually not a symptom of bad ram, but rather too high infibity fabric, wrong cpu voltage etc. Bad ram is more of a bsod thing in my experience.

droopyRO

Yesterday i replaced the X570 Elite with a Tomahawk MAX. So far so good,

So, 4 days later and it is stable. No more WHEA or errors, or lockups at idle or sleep. I suspect it is a combination of motherboard and/or the RX 6600 XT. As others have found out. You would think that AMD tests their products on their own hardware:
Reddit

Last edited: Sep 23, 2021

outpt

Does core optimizer work with a all core oc?

Soooo, this is annoying. I have run +200 pbo and - 30 on all core except one that does - 29. Corecycler prime and ycruncher is long term stable, but I get rare reboots 2-3 times a month. - 30 CO allcore was stable for 3 months with +50 so I know the problem isn't low idle voltage, but atleast one core must be slightly unstable during low loads that spike voltage. Any idea how to identify which core? Other tests I can run?

outpt

i just gave up on curve optimizer and started running all core at 4.7ghz using mb voltage settings. which are around 1.3v using gooseberry benchmark to check temps and voltages. at idle about 1.1v or lower gets about 80-81C fill tilt. Or run it pbo enabled and it run 4.85ghz. have not had blue screen error in a week.
there is one thing i have noticed is that i can set it to -30 all cores and it runs fine until i put it to sleep. next morning coming out of sleep it's blue screens to the max. somebody explain that.

Makaveli

Soooo, this is annoying. I have run +200 pbo and - 30 on all core except one that does - 29. Corecycler prime and ycruncher is long term stable, but I get rare reboots 2-3 times a month. - 30 CO allcore was stable for 3 months with +50 so I know the problem isn't low idle voltage, but atleast one core must be slightly unstable during low loads that spike voltage. Any idea how to identify which core? Other tests I can run?

Try minus -15 on all cores and -30 on your best core then -29 on the other one then check your long term stability.


Page 10

outpt

jesdals

i just gave up on curve optimizer and started running all core at 4.7ghz using mb voltage settings. which are around 1.3v using gooseberry benchmark to check temps and voltages. at idle about 1.1v or lower gets about 80-81C fill tilt. Or run it pbo enabled and it run 4.85ghz. have not had blue screen error in a week.
there is one thing i have noticed is that i can set it to -30 all cores and it runs fine until i put it to sleep. next morning coming out of sleep it's blue screens to the max. somebody explain that.

I am haveing some issues with cold boots after disconnection the power to my system - do switch of the power as well or only sleep mode?

tabascosauz

Problem is that the second best core is the - 29 core, the ranking doesnt add up.

Obviously the core quality ranking doesn't mean anything once you've already changed the V-F curves by applying CO offsets. Makes no sense comparing between cores, because every core comes binned very differently and per-core CO changes each core only relative to its own stock V-F. On mine Core 1 only does -5, it takes Core 2 @ -30 just to reach similar 4.925-4.95GHz clocks to Core 1 @ -5. But if you were comparing the two you'd think that Core 2 was significantly "better" for having a much bigger undervolt right? If anything, it should be the other way around. I'd assume the top rated cores are already performing close to their full potential (since they are always preferred by CPPC) - so minimal room for undervolting. Whereas the other cores (esp. on CCD2) can be relaxed and won't matter whether AMD pushes them hard or not out of the box, since they'll never see more than all-core speeds. I've been 3 out of 3 on this so far, 5600G, 5700G and 5900X all have lower achievable offsets on the 2 best cores.

Which is why while a broad, say, -15 undervolt across the board is understandable if you don't have time to test, it also logically makes no sense because -15 on Core 0 is not going to mean the same thing as -15 on Core 5 for example. It's a good place to start before you fine-tune, though, because it gives you a good sense of where every core stands currently.

Obviously the core quality ranking doesn't mean anything once you've already changed the V-F curves by applying CO offsets. Makes no sense comparing between cores, because every core comes binned very differently and per-core CO changes each core only relative to its own stock V-F. On mine Core 1 only does -5, it takes Core 2 @ -30 just to reach similar 4.925-4.95GHz clocks to Core 1 @ -5. But if you were comparing the two you'd think that Core 2 was significantly "better" for having a much bigger undervolt right? If anything, it should be the other way around. I'd assume the top rated cores are already performing close to their full potential (since they are always preferred by CPPC) - so minimal room for undervolting. Whereas the other cores (esp. on CCD2) can be relaxed and won't matter whether AMD pushes them hard or not out of the box, since they'll never see more than all-core speeds. I've been 3 out of 3 on this so far, 5600G, 5700G and 5900X all have lower achievable offsets on the 2 best cores.

Which is why while a broad, say, -15 undervolt across the board is understandable if you don't have time to test, it also logically makes no sense because -15 on Core 0 is not going to mean the same thing as -15 on Core 5 for example. It's a good place to start before you fine-tune, though, because it gives you a good sense of where every core stands currently.

I have the time to test and - 15 all core would reduce performance vs - 30 and +50 which I know us stable :p i get about 200MHz more allcore vs stock using - 30 on all. Stock AC lies at 4.4, - 30 gives me 4.6 avg. I can try reducing a bit on my best core and see if this resolves the issue :) If anyone has a tip for a low liad pr core test which invokes maximum voltage/core clock I bet that could find out which core is the slightly unstable one.

HD64G

Different type of workloads allows more voltage and frequency for Ryzen CPUs. Needs some experimenting (time and patience) to find the best balance. Don't try for the ultimate. Go for the stable first and work from there a bit more. Settle after a few tests and have fun with your PC. Don't waste many hours for practically nothing. Friendly advice.

Different type of workloads allows more voltage and frequency for Ryzen CPUs. Needs some experimenting (time and patience) to find the best balance. Don't try for the ultimate. Go for the stable first and work from there a bit more. Settle after a few tests and have fun with your PC. Don't waste many hours for practically nothing. Friendly advice.

I tweak for fun, and have a lot of patience, only things misding is the tools to identify which core is the problem ;) I have a safe profile to fall back on ;)

Got a tip on another forum, eventviewer from reboots. That worked! Found out both core 0 and 1 had caused a couple of reboots, but I only adjusted core 0. Also adjusted core 1 now, hopefully that fixes it.

outpt

I am haveing some issues with cold boots after disconnection the power to my system - do switch of the power as well or only sleep mode?

coming out of sleep mode or complete shut down. it's runs for hrs. or days till i do any kind of restart/shut down. at this point i have thrown in the towel. besides i only have a xfx 5600xt so its not really being pushed that hard. at this point pbo or ac overclock works just fine. to be honest i haven't seen it do anything at all. O'Well

purecain

What a thread, amazing work overclocking guys. I want the higher single core performance now , and that isnt possible on air.imo Lucky for me I have decent silicon. Since AMD released a (beta)bios that literally turned the pump off, on two of my water-cooling kits before I realised the culprit ive been on air. But after reading only half of this thread I've decided I need that single core performance.

I'll check out the rest of the thread and see which cooling has been the most successful. Arctic cooler 2 AIO stuck in my head, and I'll probably give that company a try.

Ibizadr

Hey guys I have a r7 5800x and play with some bios settings to try to achieve the best of this cpu. In my CO I'm with Max boost: +200mhz PPT:200 TDC:145 EDC:145 CO: - 15 all core except 2 best one with - 10 Try these settings today on cinebench and got this any way to improve since and don't get full pbo in all cores? I have an artic aio 240mm and got in multi thread almost 80ºc. Its possible to achieve 5.0ghz in a single thread with 5800x? Some advice to improve?

IMG_20210927_231800.jpg

Last edited: Sep 28, 2021

Mussels

Hey guys I have a r7 5800x and play with some bios settings to try to achieve the best of this cpu. In my CO I'm with Max boost: +200mhz PPT:200 TDC:145 EDC:145 CO: - 15 all core except 2 best one with - 10 Try these settings today on cinebench and got this any way to improve since and don't get full pbo in all cores? I have an artic aio 240mm and got in multi thread almost 80ºc. Its possible to achieve 5.0ghz in a single thread with 5800x? Some advice to improve?

View attachment 218442

I can get 5050Mhz on mine, remember that the curve undervolt going too far can actually lower the clocks, as the chip doesnt get enough voltage to ramp up

your numbers may simply be too high, for your cooling and VRMs. lower them.

tabascosauz

Hey guys I have a r7 5800x and play with some bios settings to try to achieve the best of this cpu. In my CO I'm with Max boost: +200mhz PPT:200 TDC:145 EDC:145 CO: - 15 all core except 2 best one with - 10 Try these settings today on cinebench and got this any way to improve since and don't get full pbo in all cores? I have an artic aio 240mm and got in multi thread almost 80ºc. Its possible to achieve 5.0ghz in a single thread with 5800x?

Some advice to improve?

Your EDC is like, stock, but what's up with 200W? ~180W is a metric ton of power to be putting through the cores on even a 5950X, let alone a single chiplet, don't matter if you have an AIO..... I'm betting a 5800X can achieve peak performance with a smart combination of PBO settings at just ~120W. Stock 142W is honestly already too much per-core power for a 5800X, that's why you can usually lop off a whole bunch of PPT and be no slower (or even faster). Having a high PPT does nothing for your single threaded performance. Higher PPT enables higher multi thread perf if you have the cooling for it, but scalar, curve optimizer on the relevant 1 or 2 cores, and EDC makes a bigger difference to ST perf. PPT you only need to make sure is "high enough" not to hold back your ST (which is like 75W for single chiplet, easy peasy) PBO +200 should cap out at 5050 for 5800X, so if your chip can handle it then there's no reason why it can't be done.

get @mtcn77 in here, not sure what the rule is, is it like EDC = PPT x 0.9?

Mussels

the post in my sig has some info from GerKNG, he suggested using: PPT: 105W TDC:70A EDC: 95A stock performance, but doesnt let the chiplet blow out into the thermal throttling ranges which can actually let it boost higher

clearly you can go higher, those are very efficient settings not the max performance ones (add 20 to each, for that)

Anyone here has any experience with overclocking a 5950x and testing stability on Linux? I tried PBO+CO settings and tested stability on a Windows install, and everything seemed fine, till I went to Linux and started seeing crashes immediately. So, all my CO values ended up being thrown out of the window on Linux. FYI, I have a Gigabyte X570 AORUS PRO WIFI.

Hey guys I have a r7 5800x and play with some bios settings to try to achieve the best of this cpu. In my CO I'm with Max boost: +200mhz PPT:200 TDC:145 EDC:145 CO: - 15 all core except 2 best one with - 10 Try these settings today on cinebench and got this any way to improve since and don't get full pbo in all cores? I have an artic aio 240mm and got in multi thread almost 80ºc. Its possible to achieve 5.0ghz in a single thread with 5800x? Some advice to improve?

View attachment 218442

Try the limits on auto (142W should be more than enough) and rather see if you can improve the UV. Try core cycler and find lowest values, if you get intermittent crashes during low loads even with 3 passes in corecycler, check the event viewer for whea 18 and find out which lower, lower the UV by 1.

purecain

This is my starting point. I dont how good these scores are as i havnt compared them yet but this is on air as

I have it set up now.

cinebenchScores1.jpg

cpuzBench677.jpg

I'm looking forwards to seeing how close I can get to higher single core performance on a Noctua NH-U12a.

Water-cooling should give me 200-400mhz hopefully at across all cores. My single core run had the cpu running at 4.69 to 4.72ghz through out on air.

outpt

Hey guys I have a r7 5800x and play with some bios settings to try to achieve the best of this cpu. In my CO I'm with Max boost: +200mhz PPT:200 TDC:145 EDC:145 CO: - 15 all core except 2 best one with - 10 Try these settings today on cinebench and got this any way to improve since and don't get full pbo in all cores? I have an artic aio 240mm and got in multi thread almost 80ºc. Its possible to achieve 5.0ghz in a single thread with 5800x? Some advice to improve?

View attachment 218442

pushing those kind of numbers means(probably) nothing. you need to scroll down further until you find ppt,edc,tdc and see what is actually being used. bring those numbers down.
PPT-125W TDC-90A EDC-115A are good starting points. EDC should be about 90% of PPT.

Ibizadr

So after some advice of you guys i change some settings. Now I'm at: Max boost : 200mhz PPT: 180 TDC: 95 EDC: 140 Only have time to do a test with cinebench in single core. I'm happy with the temp it reaches in single core I think 60 was not too hot and in game RTSS never show temps high. For now I want to improve Single core since this pc its most for gaming. I post a ss of hwinfo that I run with cinebench. Do you guys have some advices for my single core UV+Pbo? I see now some more advices and later I will go to turn down my settings to see if I got improvments

Captura_de_Ecra_26.png

jesdals

Hey guys I have a r7 5800x and play with some bios settings to try to achieve the best of this cpu. In my CO I'm with Max boost: +200mhz PPT:200 TDC:145 EDC:145 CO: - 15 all core except 2 best one with - 10 Try these settings today on cinebench and got this any way to improve since and don't get full pbo in all cores? I have an artic aio 240mm and got in multi thread almost 80ºc. Its possible to achieve 5.0ghz in a single thread with 5800x?

Some advice to improve?

Remember Alt+Printscreen thats easier than the phone. And please fill in your system specs - perhaps some one have the same motherboard etc.

I used to ad to one core at the time ( a pain on the 5950x) and wrote down how the changes affected in intervals of 5 to begin with and then 2.

Ibizadr

Remember Alt+Printscreen thats easier than the phone. And please fill in your system specs - perhaps some one have the same motherboard etc.

I used to ad to one core at the time ( a pain on the 5950x) and wrote down how the changes affected in intervals of 5 to begin with and then 2.

If you see my last post before your reply you see a screenshot. After that I change my settings after a couple of people telling me 180w was too much for my 5800x. I change the settings for the screenshot in this post but not tested yet with these settings

MSI_SnapShot_00.jpg

outpt

have a msi mag b550 tomahawk and my bios is very similar.

jesdals

@Ibizadr
I would suggest that you try to look at your Infinity settings - if your memory allow for 3600MHz settings it would be a large boost - with the right memory 1900/3800MHz Infinity settings should be possible on that board.

Ibizadr

@Ibizadr
I would suggest that you try to look at your Infinity settings - if your memory allow for 3600MHz settings it would be a large boost - with the right memory 1900/3800MHz Infinity settings should be possible on that board.

Yeah I think it's possible since it's a Samsung b die. Ryzen dram calculator only have zen 2, the settings for zen 3 works? I will try to overclock my ram since my system it's stable at practically stock settings.
Best way to overclock ram its disabling xmp or use xmp and only change my Infinity fabric dram speed with dram calculator?

outpt

I think I may have stumbled on to may bsod problems. I think my gpu is on the way out. Today I started getting the GSOD it is my understanding that this may be gpu and or gpu memory related. I have turned down the power limits to 80% and memory timings down also. I have 3 cores running between -20 to -28. All other cores are at 0. I haven’t had any death screens in 5 hrs knock on wood. This at the moment means nothing but I have this gpu go belly up once before and took it apart and cleaned it up the thermal paste was as hard as a brick and it worked. After trying everything with the cpu and no whea errors no memory errors and so forth. If this gpu fails I am going to walk out into 5pm Dallas traffic

sumbitch my gpu is failing. all i have is a gtx660 backup

Mussels

Windows key + shift + s uses the new snipping tool in W10 and W11, you can then paste that immediately into TPU with no effort (it's awesome)

answering all the mess of posts above: a 5800x just cant use some of those numbers that high, so letting it max out for no gain isn't helping (0.01% faster for 10C, when the power and heat savings could give a bigger gain) - and black/grey screen crashes.... yeah core undervolting can be like that.


Page 11

I did a bit of testing yesterday with stock 76W limit. In cinebench 23 I get 11043p stock and 11647p (5.5%) with +200 pbo and -29x2/-30x4 cores. Clockspeed stock is 4300 avg stock (1162mv avg) and 4550 using CO (1131mv avg). In SOTTR I get 236fps avg cpu in 1080p low with the above CO-setting and 225fps stock so about 5% performance. Idlevoltage sits around 950mv stock abd 850mv CO. Overall a fair 'free' improvement that produces similar heat.

My allcore 4.8GHz got 2% more fps in SOTTR, but got thermal throttling in CB23 (using 115W+) and requires 1.32V so I would say CO is superior. If I up PPT-limit a bit (90-100W) I bet I could see close to 4.7GHz allcore at around 1.2-1.25V.

droopyRO

So, 4 days later and it is stable. No more WHEA or errors, or lockups at idle or sleep. I suspect it is a combination of motherboard and/or the RX 6600 XT. As others have found out. You would think that AMD tests their products on their own hardware:
Reddit

The "drama" has ended. They said that they can't be fixed. And i had to choose between a new 5600X or my money back. I took the money. I am now on my 8600K and pondering what to do next. Stay on Ryzen 5000 series, go on Intel 11 or wait untill 2022. Thanks again to all who offered help.

iirc you had an early production Ryzen 5000 right? I figure with the way 5000 prices are coming down now, you could just go for a newer production 5600X. I'm not even joking, 99% of the problem threads I've ever seen regarding Vermeer CPUs concerned chips made before January 2021. Or if you want to wait for the new generation of V-cache you could do that too, 8600K is still solid. If you're going to the store to pick one up, look at the CPU through the box's side window to make sure the batch code starts with 21. I don't see how it could possibly be anything else, but to be sure for August or September production the number should be 2131 or greater.

Alder Lake is coming, could wait for further price adjustments since you have a temporary solution. As for Intel, I would just take either a 10900K or wait for Alder Lake if you're air cooling.

Yeah it was bought at launch in November. I delided my 8600K last year, but did not play with it since i bought the 5600X. I am now at 5.1Ghz with it and it is R23 stable for 20 minutes. It gets to 75ºC, crazy how much that liquid metal lowers the temps. I think i will wait to see what Alder Lake delivers and the decide. Thanks.

Last edited: Sep 30, 2021

tabascosauz

The "drama" has ended. They said that they can't be fixed. And i had to choose between a new 5600X or my money back. I took the money. I am now on my 8600K and pondering what to do next. Stay on Ryzen 5000 series, go on Intel 11 or wait untill 2022. Thanks again to all who offered help.

iirc you had an early production Ryzen 5000 right? I figure with the way 5000 prices are coming down now, you could just go for a newer production 5600X. I'm not even joking, 99% of the problem threads I've ever seen regarding Vermeer CPUs concerned chips made before January 2021. Or if you want to wait for the new generation of V-cache you could do that too, 8600K is still solid. If you're going to the store to pick one up, look at the CPU through the box's side window to make sure the batch code starts with 21. I don't see how it could possibly be anything else, but to be sure for August or September production the number should be 2131 or greater.

Alder Lake is coming, could wait for further price adjustments since you have a temporary solution. As for Intel, I would just take either a 10900K or wait for Alder Lake if you're air cooling.

Mussels

I'll also add in that i've never even heard of these dud CPU's outside of tech forums, and maybe seen 5 different users with the issue

the odds of getting another faulty chip are basically nil (there was a lot of people blaming chips with incompatible RAM, mostly odd numbered CAS latency, dual rank corsair LPX)

Ibizadr

I was dealing with CO with the last days everything went fine till the moment I have random reboots. This reboots its related to a higher value in my prefered core? Today it happens when I open bf4 but happens this morning after 8h of pc turned on with mining program open.

I was dealing with CO with the last days everything went fine till the moment I have random reboots. This reboots its related to a higher value in my prefered core? Today it happens when I open bf4 but happens this morning after 8h of pc turned on with mining program open.

Have you run corecycler? Check windows event viewer, see if you have whea 18 when reboot occured, lower negative value by 1 or 2 on affected core.

freeagent

I will tinker with core cycler today.. I had a couple of reboots last night doing 11 stuff.. ahh well.

tabascosauz

I will tinker with core cycler today.. I had a couple of reboots last night doing 11 stuff.. ahh well.

Aren't you running the max +200 5150MHz global limit? I had to scale back my curves a bit on some cores going past 4950MHz. Otherwise they worked just fine at the stock global limit.

outpt

I was dealing with CO with the last days everything went fine till the moment I have random reboots. This reboots its related to a higher value in my prefered core? Today it happens when I open bf4 but happens this morning after 8h of pc turned on with mining program open.

Also look for event id 41. These can point to ram/power supply problems as a general rule

purecain

I have one of the original 5000chips made. I'm definitely considering getting a 2021 model after reading this thread.

That or wait for the next production update or new cpu's. Cool thread! :toast:

Ibizadr

Have you run corecycler? Check windows event viewer, see if you have whea 18 when reboot occured, lower negative value by 1 or 2 on affected core.

Yes in last 24h I have whea 18 - 8 erros. I lower the value in SC this should be enough?

Also look for event id 41. These can point to ram/power supply problems as a general rule

I have on id 41 in last 24h 8, but my supply it's enough for my system and ram runs at original speeds no overclock till now

freeagent

Aren't you running the max +200 5150MHz global limit? I had to scale back my curves a bit on some cores going past 4950MHz. Otherwise they worked just fine at the stock global limit.

Yes I was running at the tippy top and all was well until 11 crossed my path lol.. I knocked 50MHz off and it seems ok, I know it won’t pass core cycler as is.. or maybe it will :D I will have to play with it.. hopefully the cpu and I can come to some sort of understanding :)

Ibizadr

I don't use core cycle what I use it's some games and daily use and cinebench r23 and cpu z combined with hwinfo

tabascosauz

I don't use core cycle what I use it's some games and daily use and cinebench r23 and cpu z combined with hwinfo

Corecycler is not an application. It's a testing script specifically geared towards testing single-core PBO and Curve Optimizer stability.

Without corecycler or OCCT or a similarly appropriate test, you're left without any way of properly verifying CO stability. Cinebench R23 and CPU-Z are next to useless for the fact that they will only ever stress your CPPC preferred boost cores, and mildly at that.

Ibizadr

Corecycler is not an application. It's a testing script specifically geared towards testing single-core PBO and Curve Optimizer stability.

Without corecycler or OCCT or a similarly appropriate test, you're left without any way of properly verifying CO stability. Cinebench R23 and CPU-Z are next to useless for the fact that they will only ever stress your CPPC preferred boost cores, and mildly at that.

I downloaded corecycler and opened that. It's running now, what I need to take a look at event viewer or the log corecycler makes?

tabascosauz

I downloaded corecycler and opened that. It's running now, what I need to take a look at event viewer or the log corecycler makes?

The script is governed by the config script in the folder. By default it runs 6 minute cycles per core, running Prime95 Large FFT, going through all the cores. This is fine for general testing and seeing if anything is blatantly unstable. It's good to run with the default config for at least 3 iterations (each iteration means it has gone through all the cores once). The script output will tell you. Later on you can fine tune your testing by editing config:

  • Number of iterations to run
  • Which cores to run (helpful for narrowing down cores that matter)
  • Which order to run the cores
  • How long to run each core
  • What to run (FFT size for Prime95, AVX or SSE, other tests)
  • What to do upon error (stop or continue)
  • Take breaks between cores

I run the default config first. Later I move on to my final config which is 1h8m per core, All FFT, SSE.

Ibizadr

The script is governed by the config script in the folder. By default it runs 6 minute cycles per core, running Prime95 Large FFT, going through all the cores. This is fine for general testing and seeing if anything is blatantly unstable. It's good to run with the default config for at least 3 iterations (each iteration means it has gone through all the cores once). The script output will tell you. Later on you can fine tune your testing by editing config:

  • Number of iterations to run
  • Which cores to run (helpful for narrowing down cores that matter)
  • Which order to run the cores
  • How long to run each core
  • What to run (FFT size for Prime95, AVX or SSE, other tests)
  • What to do upon error (stop or continue)
  • Take breaks between cores

I run the default config first. Later I move on to my final config which is 1h8m per core, All FFT, SSE.

Yeah I'm on default config for now I watch a video on YouTube and see when errors happen it will appear on corecycler cmd window so for now 4 cores tested and no errors. If my cpu runs without errors that means I can push more in CO like 15 to 16?

tabascosauz

Yeah I'm on default config for now I watch a video on YouTube and see when errors happen it will appear on corecycler cmd window so for now 4 cores tested and no errors. If my cpu runs without errors that means I can push more in CO like 15 to 16?

Stop everything you're doing while the test is running. In order for it to work properly it needs to be able to achieve the highest clocks possible, meaning minimal load on the system. Close all background apps you don't need and let it do its thing, go do something else for a while.

Ibizadr

Stop everything you're doing while the test is running. In order for it to work properly it needs to be able to achieve the highest clocks possible, meaning minimal load on the system. Close all background apps you don't need and let it do its thing, go do something else for a while.

For now only have hwinfo running in background but I have it opened when I start corecycler so for this run I will let it open and in next adjustments I will close everything and try

Yes I was running at the tippy top and all was well until 11 crossed my path lol.. I knocked 50MHz off and it seems ok, I know it won’t pass core cycler as is.. or maybe it will :D I will have to play with it.. hopefully the cpu and I can come to some sort of understanding :)

What you do it's what I do, I try to do it without corecycler but after that I think corecycler it's a better way to see if core have errors. I only do one turn at 6min per core and all went fine tomorrow will do it some more tests

freeagent

I went to run it but it wont run because RM is not installed.. :laugh: Maybe I will quit being soft, I am pretty sure 5125 is good.. but then again I thought 5150 was good. i will just leave it +150 for now.

I quit using Asus Performance enhancement, maybe that's why I have an issue.. that setting is full on cheater mode :D

Edit:

She is boosty :)

boosty.png

Yes in last 24h I have whea 18 - 8 erros. I lower the value in SC this should be enough?

I have on id 41 in last 24h 8, but my supply it's enough for my system and ram runs at original speeds no overclock till now

Yeah, lower value on affected core. Check core number in the whea. You should run core cycler first though, much faster to eliminate errors, use event viewer to adjust the cores core cycler eliminate.

So you completely disregard the differences between cores and just blindly punch in some numbers. That's not how overclocking is done.

My guy. I spent time doing per core and the resulting performance gain compared to -15 all core was fuck all anything, 25 MHz more on the best core at best. -15 all core is good enough for me, because it makes my 5900X reach 5.2 GHz single thread (on almost every core - not just the best ones) and 4.65 GHz all core, on Linux at least, using Htop which shows effective clocks - compared to the stock 5 GHz single thread & 4.15 GHz all core. Don't know about Winblows. If all core Curve Optimizer wasn't a passable choice then why would they even implement it as an option? I can hardly get people to set a quick negative all core Curve Optimizer offset and increase PBO limits a bit on their Zen 3 CPUs so they don't leave performance on the table, let alone convince them to go through all that rigorous testing for no real discernible benefit in the end.

Though I stopped caring about OC in general, it's about time I start using this machine instead of tuning and upgrading it ever since I built it. The ""performance"" ""differences"" were not worth the time. If that's your thing, go ahead. But don't act like doing a quick overclock by setting an all core number isn't a passable thing to do either, some people just don't have the time or care enough to go through all that crap with no real benefit. It's why I stopped caring about RAM OC as well and just halved my tRFC, fixed tRC and called it a day. Literally the same latency improvements as manually optimizing each timing with the added bonus of being stable and not requiring 24h worth of testing after setting each timing. Same goes for the all core option in the Curve Optimizer.

Last edited by a moderator: Oct 6, 2021

My guy. I spent time doing per core and the resulting performance gain compared to -15 all core was fuck all anything, 25 MHz more on the best core at best. -15 all core is good enough for me, because it makes my 5900X reach 5.2 GHz single thread (on almost every core - not just the best ones) and 4.65 GHz all core, on Linux at least, using Htop which shows effective clocks - compared to the stock 5 GHz single thread & 4.15 GHz all core. Don't know about Winblows. If all core Curve Optimizer wasn't a passable choice then why would they even implement it as an option? I can hardly get people to set a quick negative all core Curve Optimizer offset and increase PBO limits a bit on their Zen 3 CPUs so they don't leave performance on the table, let alone convince them to go through all that rigorous testing for no real discernible benefit in the end.

Though I stopped caring about OC in general, it's about time I start using this machine instead of tuning and upgrading it ever since I built it. The ""performance"" ""differences"" were not worth the time. If that's your thing, go ahead. But don't act like doing a quick overclock by setting an all core number isn't a passable thing to do either, some people just don't have the time or care enough to go through all that crap with no real benefit. It's why I stopped caring about RAM OC as well and just halved my tRFC, fixed tRC and called it a day. Literally the same latency improvements as manually optimizing each timing with the added bonus of being stable and not requiring 24h worth of testing after setting each timing. Same goes for the all core option in the Curve Optimizer.

Although I understand your reasoning, I think the most gain can be had by doing little. Curve optimizer is generally tuned in after an hour of core cycler. If you are lucky you can run single core 200MHz higher and all core 200-300MHz higher, if unlucky you still get 200 single core and 100-200 allcore.

As for ram timings, just find your highest working frequency and do som basic tuning. It doesn`t take that long. If you are on B-die just set 1.45V, flat first 3, tras first + second and tRC tras + tRP. tRFC = tRC x 6 on B-die, tRC x 8 on Hynix and tRC x 10 on Micron. The other ones that matter is tFAW which always work at 24, trrds at 6, trrdl at 8, tWR at same as CL, tRTP at half tWR, WTRS 4, WTRL 12. Rest can stay on auto. You gain 80% of the performance potential by doing that and it doesn`t take long. If you use Micron or Hynix things are a bit more complex, but setting tRCD and tRP +4 above tCL and using the rest of the rules tends to work fine there aswell :)

If you are okay with 5-10% better peformance your approch is good, if you want 15-20% try mine, if you want 25% you need to spend a lot of time :)

bubbleawsome

Started getting random BSODs way more often than before. Starting to wonder if my RAM OC is not as stable as previously thought and it's messing with my tuner results.

freeagent

I thought it was my ram too, but its @ stock :D

I haven't installed Ryzen Master, and probably wont.. but I did back it off that 50MHz and I haven't had a problem so far, my usual arsenal seems to be good.. but is it? :D

maybe.. probably.. possibly.. better be..


Page 12

Ibizadr

In corecycler I'm with original config only change time in each core but I see in some posts related to corecycler the creator attend to test it first with small fft to give more boost in light thread to the cores and it gives errors more frequently. In original config it uses huge fft, what the best way to to some tests with small or huge ffts?

tabascosauz

In corecycler I'm with original config only change time in each core but I see in some posts related to corecycler the creator attend to test it first with small fft to give more boost in light thread to the cores and it gives errors more frequently. In original config it uses huge fft, what the best way to to some tests with small or huge ffts?

Small FFT should reduce clocks - in P95 the bigger the FFT size the more it goes into memory/memory controller and you'll often see lower temps and higher clocks. But in reality, testing between Small FFT, Large FFT and Huge FFT, there's not much of a difference in clocks...................as long as you don't enable AVX. So stay on SSE like default. If you have the time (it's easier for you because you only have 8 cores), you could just test All FFTs like I do. But that takes much longer and you have to extend the time period significantly to cover all FFT sizes in that range - I run just over an hour.

More importantly is that you let it run for at least 3 iterations on every core (I let it cycle overnight through at least 5 iterations). Once you're on the edge, often an unstable core will pass one or two iterations, seemingly stable, only to fail on the next one - so if you only let it run for one or two cycles, you might assume that it was "stable" when it actually wasn't.

In corecycler I'm with original config only change time in each core but I see in some posts related to corecycler the creator attend to test it first with small fft to give more boost in light thread to the cores and it gives errors more frequently. In original config it uses huge fft, what the best way to to some tests with small or huge ffts?

I run it at default setting and used event viewer to find the last slghtly unstable core. I would start with - 30 on all and +200 pbo, if negative is way of you get error quick, if it's slightly iff it fails after a few minutes or in test 2.

Ibizadr

When pc reboots when doing corecycle I need to lower the value on the core that reboot my pc? Like to 21 to 19?

When pc reboots when doing corecycle I need to lower the value on the core that reboot my pc? Like to 21 to 19?

Yes, try that. Do you see ehich core it worked on when reboot occured? Only lower that one.

Ibizadr

Yes, try that. Do you see ehich core it worked on when reboot occured? Only lower that one.

I check in event viewer and on log created in corecycle, it's the same core on event viewer and corecycler last core tested so I assumed that core need more voltage

I have my cpu llc on auto in my msi b450 tomahawk max II, it's better to put it in mode 4 like 1usmus advice in CTR 2.1or leave it on auto?

tabascosauz

I check in event viewer and on log created in corecycle, it's the same core on event viewer and corecycler last core tested so I assumed that core need more voltage

I have my cpu llc on auto in my msi b450 tomahawk max II, it's better to put it in mode 4 like 1usmus advice in CTR 2.1or leave it on auto?

The recommendation in CTR is because CTR either wants you to use a static all-core OC according to its recommendation, or use its own hybrid OC mechanism (half of which is still basically static all-core).

Messing around in CO is still under PBO, which is still just stock boost. When on PB stock boost, leave it on auto, because iirc droopy and loose LLC is actually what the boost algorithm wants. Aggressive LLC can make static OC easier, but for stock boost aggressive LLC doesn't help/doesn't change anything/hurts boost.

Ferrum Master

The recommendation in CTR is because CTR either wants you to use a static all-core OC according to its recommendation, or use its own hybrid OC mechanism (half of which is still basically static all-core).

Messing around in CO is still under PBO, which is still just stock boost. When on PB stock boost, leave it on auto, because iirc droopy and loose LLC is actually what the boost algorithm wants. Aggressive LLC can make static OC easier, but for stock boost aggressive LLC doesn't help/doesn't change anything/hurts boost.

I was supporting CRT on the Patreon. Well it turned out ugly. CTR is an ugly mess, I will not ever recommend it to anyone, not mentioning the fact the coder is a whining bitch, his emails for supporters were like from a demanding ex and he didn't realize that his bug driven tool is really so much full of shit it is barely usable to ask money for it and blaming anyone having problems with it, it ain't the tool, but you. Very unprofessional. Well it all depends on your sample and motherboard... it very wild setting wise as far I experienced. The LLC does react different in between vendors, it is normal as VRMS differ a lot and their speed and performance. In the end it is just experiment. Example for my 5950X. 1. Make your RAM solid stable. 2. Do not touch anything CPU voltage wise, adjust the RAM domain and VSOC, CCD, IOD voltages depending on your RAM amount(like how many sticks) gen and speed. It is very crucial not to overdo there, but leaving it stock is not an also an option often. Depends on the board. AUTO often gives worst results. 3. PBO. I do it the old fashioned way, in a notebook(I mean the one having paper). I wrote down my core statistics. I have two CCD's, each two best core pairs, then second then third. So I gradate them. Best for may CCX0 have -2m then second best pair has -6. Then other -16. For Second CCX I have the same except the best have -4 then next pair -10 and then -16 for other. 4. Then do Power limit play depending if you wish more multicore or single core performance. Yes, that's decided there.

Any suggestions for my maybe flawed reasoning?

outpt

when looking for the correct core to adjust go to event viewer log for the processor APIC ID:#. this tells you what core crashed. Go back into CPUZ and look at tool and save report as text. you can save this to desktop if you want to.
some times core numbering can start as core-0 or core-1. CPU-Z helps making sure you are adjusting the correct core.

Ibizadr

I was supporting CRT on the Patreon. Well it turned out ugly. CTR is an ugly mess, I will not ever recommend it to anyone, not mentioning the fact the coder is a whining bitch, his emails for supporters were like from a demanding ex and he didn't realize that his bug driven tool is really so much full of shit it is barely usable to ask money for it and blaming anyone having problems with it, it ain't the tool, but you. Very unprofessional. Well it all depends on your sample and motherboard... it very wild setting wise as far I experienced. The LLC does react different in between vendors, it is normal as VRMS differ a lot and their speed and performance. In the end it is just experiment. Example for my 5950X. 1. Make your RAM solid stable. 2. Do not touch anything CPU voltage wise, adjust the RAM domain and VSOC, CCD, IOD voltages depending on your RAM amount(like how many sticks) gen and speed. It is very crucial not to overdo there, but leaving it stock is not an also an option often. Depends on the board. AUTO often gives worst results. 3. PBO. I do it the old fashioned way, in a notebook(I mean the one having paper). I wrote down my core statistics. I have two CCD's, each two best core pairs, then second then third. So I gradate them. Best for may CCX0 have -2m then second best pair has -6. Then other -16. For Second CCX I have the same except the best have -4 then next pair -10 and then -16 for other. 4. Then do Power limit play depending if you wish more multicore or single core performance. Yes, that's decided there.

Any suggestions for my maybe flawed reasoning?

I don't touch in my ram for now only load xmp. My advise is too run corecycler it's a really useful tool to do the work for CO. I never experienced an error in corecycler only reboots but I think my best cores was a bit high negative (-18,-14) if you don't got errors or reboots try again but edit config file and add more time per core and try to do it at least 3 iteration. For example I pass 3 iterations of 6min per core yesterday and today try 15min per core and it reboots. So always give a second try to corecycler with more time per core.

when looking for the correct core to adjust go to event viewer log for the processor APIC ID:#. this tells you what core crashed. Go back into CPUZ and look at tool and save report as text. you can save this to desktop if you want to.
some times core numbering can start as core-0 or core-1. CPU-Z helps making sure you are adjusting the correct core.

Yes in event viewer appears APIC ID and i think after its the thread number dont know this way to know the numbering via cpu z. Thanks for that tip I use corecycler log to see what core corresponding to APIC ID error

Ferrum Master

I don't touch in my ram for now only load xmp. My advise is too run corecycler it's a really useful tool to do the work for CO. I never experienced an error in corecycler only reboots but I think my best cores was a bit high negative (-18,-14) if you don't got errors or reboots try again but edit config file and add more time per core and try to do it at least 3 iteration. For example I pass 3 iterations of 6min per core yesterday and today try 15min per core and it reboots. So always give a second try to corecycler with more time per core.

Yes in event viewer appears APIC ID and i think after its the thread number dont know this way to know the numbering via cpu z. Thanks for that tip I use corecycler log to see what core corresponding to APIC ID error

XMP is not a thing you can trust on. It is often made for intel platforms, it does not work on AMD the same. Corecyler is for those random reboots on standby. The real prime test or any relatives is for all core stability. For gaming test I use some Final Fantasy benches, I have them found good triggering random BSODS on CPU instability as those are not very GPU dependent.

3 iterations is not enough... I leave usually each test of my config overnight.

AVATARAT

I don't like anymore corecycler because it's overvolting cores. Now I prefer OCCT with the following settings:

1633731880161.png

1633731923402.png

Just run this core by core and lower or upper the Curve optimizer on the core that gets an error. But be careful because sometimes too much or too lower voltage on one core can affect the other core next to it.

Normally you can get an error in 20-30 seconds.

Ibizadr

I don't like anymore corecycler because it's overvolting cores. Now I prefer OCCT with the following settings:

View attachment 220056 View attachment 220057

Just run this core by core and lower or upper the Curve optimizer on the core that gets an error. But be careful because sometimes too much or too lower voltage on one core can affect the other core next to it.

Normally you can get an error in 20-30 seconds.

From my experience in the last days I think you are right when you say sometimes voltage on one core affect other core next to it. I will give a try with occt and your settings too since it gives errors more faster than corecycler in your words.

Anyone know why I can do linpack extreme but fail to run geekbench 5. What is the cause?

jesdals

New AMD chipset driver out

freeagent

Do people install those? What do they actually do that the windows drivers don’t?

Do people install those? What do they actually do that the windows drivers don’t?

The latest one does this:

1634838948718.png

Strangely, it's only listed when you choose the X570 chipset on the driver page, but the release notes say it's for B550 as well.

outpt

dl for my b550 and installed . No problems. Will test

Mussels

L3 Cache results went from ~625 to ~655 Driver good.

1634856816784.png

Nordic

Do people install those? What do they actually do that the windows drivers don’t?

They actually have a very strong impact on performance. It made a huge difference back when ryzen 3000 first came out. This driver will probably be really important for windows 11.

Mussels

Do people install those? What do they actually do that the windows drivers don’t?

For zen 1/2, they add a ryzen power plan that helps performance out.

For Zen 3, they add these sort of performance fixes. They also include chipset, USB drivers, sata, etc etc - minor fixes and such that you dont really notice.

RJARRRPCGP

linpack isn't exposing your instability.

That is common, if there's bus instability. (Usually CPU-to-bus communication issues, most likely CPU-bus termination issues)

I had that with a poopy Core 2 Quad Q6600 G0 at only 367 Mhz FSB, I had to crank the FSB termination voltage for Prime95 in blend mode to not fail with "STOP: 0x00000124" BSOD with "Bus/Interconnect Error" being the reason. That error reason is possibly the most common on 65nm Core 2 Quads.


Page 13

Anyone here has any experience with overclocking a 5950x and testing stability on Linux? I tried PBO+CO settings and tested stability on a Windows install, and everything seemed fine, till I went to Linux and started seeing crashes immediately. So, all my CO values ended up being thrown out of the window on Linux. FYI, I have a Gigabyte X570 AORUS PRO WIFI.

Hi @RandomJerk, I've only ever had linux for the last decade in all my PCs, Workstations, servers, and IoT SoCs. I haven't got a 5950x but I do have a 5800x watercooled all core overclock in my main workstation server; it is coupled with a Powercolor Liquid Devil Ultimate 6900xt.

For CPU stability testing I use prime95 called "mprime" in linux: https://www.mersenne.org/download/

I typically run it for about 2 hours on all threads on the torture setting option 2 small FFTs maximum power test. Then I run the blend test for 12 hours. I'm stable on all core OC at 4.65 under load. I do all core OC as my workstation does 24/7 multithreaded compute so my results are better compared to PBO2.

For GPU stability testing I use Furmark or the Piano Test from Geeks3D. https://www.geeks3d.com/20140304/gp...w-fp64-opengl-4-test-and-online-gpu-database/

It is an older version but does the job natively. If you want the newer version of Furmark then you can run it in Wine I suppose, but I don't bother and the 0.7.0 version does the job.

I modify the various .sh scripts to fit my screen resolution. Just open it up in Geany or the Gnome Text Editor and change the config. For stability testing it is the Windowed scripts that you want to modify since it will run indefinitely until you terminate it.

Mussels

Just because i've got the info scattered in half a dozen threads now, but its relevant to this topic: Temps are key to the clocks when using PBO. I'm at a mere 98W PPT on my 5800x, but because its under/at 60C i'm seeing 4.85GHz all core while gaming.

Performance is absurd.

ViperXTR

Just because i've got the info scattered in half a dozen threads now, but its relevant to this topic: Temps are key to the clocks when using PBO. I'm at a mere 98W PPT on my 5800x, but because its under/at 60C i'm seeing 4.85GHz all core while gaming.

Performance is absurd.

98PPT damn, hope my new cooler improves it as well, but not as absurd as yours though, now im curious how your system handles the likes of heavy avx2 RPCS3 (and avx512 instructions, without needing avx512 itself)

Mussels

98PPT damn, hope my new cooler improves it as well, but not as absurd as yours though, now im curious how your system handles the likes of heavy avx2 RPCS3 (and avx512 instructions, without needing avx512 itself)

I actually am not sure if i'm on 108 or 98 right now, as i was testing a few different options tbh AVX stuff runs lower clocks than gaming, cause its more power hungry i assume Used OCCT and its AVX2 setting - 4.65Ghz 108W, 62C

1636696969195.png

ViperXTR

I actually am not sure if i'm on 108 or 98 right now, as i was testing a few different options tbh AVX stuff runs lower clocks than gaming, cause its more power hungry i assume Used OCCT and its AVX2 setting - 4.65Ghz 108W, 62C

View attachment 224887

Yeah, AVX Workloads tends to downclock the frequency, even more so for AVX2 running AVX512 instructions in which RCPS3 is doing one of the areas where the 11th gen intel can overtake zen3 because of AVX512

1636704530053.png

1636704695961.png

Yeah, AVX Workloads tends to downclock the frequency, even more so for AVX2 running AVX512 instructions in which RCPS3 is doing one of the areas where the 11th gen intel can overtake zen3 because of AVX512

View attachment 224892

View attachment 224893

You can enable AVX-512 on Alder lake in Linux. But the catch is you will need to disable the E-cores...
Here are the benchmarks and how to do it... https://www.phoronix.com/scan.php?page=article&item=alder-lake-avx512&num=1 There is a story within a story here... Intel always had AVX-512 earmarked for Alder-lake but the 10nm design and the sheer density of stuff all working within the 10nm design lead to too much power usage and not enough heat dissipation. It is still a problem and it will be interesting in future architecture on how they will get around that problem and solve it.

The easy path is to do less in the volumetric area they are playing with to have less heat, which means less featured CPUs; however with AMD, Nvidia and Apple going even smaller on TSMC's and Samsungs node-tech; the only way Intel can compete is to do more in the volumetric area they are playing with, which for now is 10nm. They are basically trailing the other two so have to compensate by stuffing more in, in the volumetric area that they have. (Volumetric as architectures are now 3D and no longer planar 2D).

Last edited: Nov 13, 2021

iiNNeX

Hey all, just found this thread as I was looking for help with my 5950x PBO OC settings. Here is where I'm at... I have tried to follow guidance on this thread and have researched what each setting does separately to try better understand what I am doing. Unfortunately I am a little stuck because certain settings seem stable until they are not, for no apparent reason. Let me explain, first off here is the hardware: 5950x MEG X570S ACE (latest bios) 3600 CL16 (Samsung Bdie) running XMP beQuiet Titanium 1000W PSU (connected to an Eaton UPS so power delivery is as clean and stable as it can be) Cooled by Arctic 360 AIO with Corsair ML Pro fans in a Meshify 2 with ML Pro Fans all around. Win 11 Pro Right then, so the bios settings I have are all stock apart from: XMP on PBO Advanced

PTT 220 TDC 135 EDC 180 (tried lower EDC but performance was worse)

Scalar Auto +50mhz boost clock temp target 90 Curve started out at -10 all core, which then passed CPU Z bench, 10 minutes of Cinebench R23 and 30 mins of Linpack OCCT + OCCT bench Temps were solid, I was happy with the performance so I moved on to -12 CO, all good, -15 CO all good, -18 CO still fine, this time I even let it stress for longer and again no WHEA errors and temps were solid (max 73C in an ambient 24.5C).

Best scores at this point were:

C23 - 29000 CPUZ - 688 single 13320 multi Then I decided to try out -20 CO and the machine did boot into windows and ran CPU Z bench, it also ran OCCT linpack however it failed on OCCT bench (pc just shut down). I thought, ok I will just back down to 18. Yeah well I did that and Windows wont boot (BSOD straight after bios loading screen). Went back, tried -15 and this time it boot but once on desktop it crashed again with WHEA error. Now I am at -10 again and it seems to work. I need help before I go mad trying to figure out how something is stable for so long and then once it crashes, I have to go back several steps just to get a working system? Also how do I figure out my best cores so I can start to optimise CO on a per core basis?

Thanks in advance, this is a nightmare but I hope to see myself through it.

Ryzen Master shows your best cores in each CCD. No idea how accurate it is though.

jesdals

Hey all, just found this thread as I was looking for help with my 5950x PBO OC settings. Here is where I'm at... I have tried to follow guidance on this thread and have researched what each setting does separately to try better understand what I am doing. Unfortunately I am a little stuck because certain settings seem stable until they are not, for no apparent reason. Let me explain, first off here is the hardware: 5950x MEG X570S ACE (latest bios) 3600 CL16 (Samsung Bdie) running XMP beQuiet Titanium 1000W PSU (connected to an Eaton UPS so power delivery is as clean and stable as it can be) Cooled by Arctic 360 AIO with Corsair ML Pro fans in a Meshify 2 with ML Pro Fans all around. Win 11 Pro Right then, so the bios settings I have are all stock apart from: XMP on PBO Advanced

PTT 220 TDC 135 EDC 180 (tried lower EDC but performance was worse)

Scalar Auto +50mhz boost clock temp target 90 Curve started out at -10 all core, which then passed CPU Z bench, 10 minutes of Cinebench R23 and 30 mins of Linpack OCCT + OCCT bench Temps were solid, I was happy with the performance so I moved on to -12 CO, all good, -15 CO all good, -18 CO still fine, this time I even let it stress for longer and again no WHEA errors and temps were solid (max 73C in an ambient 24.5C).

Best scores at this point were:

C23 - 29000 CPUZ - 688 single 13320 multi Then I decided to try out -20 CO and the machine did boot into windows and ran CPU Z bench, it also ran OCCT linpack however it failed on OCCT bench (pc just shut down). I thought, ok I will just back down to 18. Yeah well I did that and Windows wont boot (BSOD straight after bios loading screen). Went back, tried -15 and this time it boot but once on desktop it crashed again with WHEA error. Now I am at -10 again and it seems to work. I need help before I go mad trying to figure out how something is stable for so long and then once it crashes, I have to go back several steps just to get a working system? Also how do I figure out my best cores so I can start to optimise CO on a per core basis?

Thanks in advance, this is a nightmare but I hope to see myself through it.

Welcome - when you got a all core setting stable, then you can progress by changeing one core at the time.

iiNNeX

Ryzen Master shows your best cores in each CCD. No idea how accurate it is though.

Ok I ll give those a go.

Welcome - when you got a all core setting stable, then you can progress by changeing one core at the time.

Thank you for all your useful posts earlier in the thread, they definitely helped determine a starting point for me! And yes that is the aim, however as per my opening post I had (or at least assumed) an all core stable CO which turned out to not be so stable. On another note I just discovered CoreCycler so have been running that for an hour which has really helped determine which of my cores couldn't deal with the offset, that explains the odd behaviour before. I will see how far I can get with CO, but other than that is there any other settings I should tweak within BIOS to help with a more stable OC? Voltage, LLC, SOC ect ?

Thanks!

Mussels

Definitely use the single threaded tester (i forgot its name, ugh) - it caught something i'd missed where my top performing core needed a tad higher setting than the rest Go for all core that boots and is roughly stable, then run that tester through allowing it to stop on cores that error. Then you can raise those back up, and go from there.

Early guess: higher perormance cores will want more, slowest cores want less.

iiNNeX

So after 6 hours of testing on CoreCycler I thought I actually had a stable system! Sadly, I was wrong once again... 2 hours of Battlefield 2042 and blue screen with WHEA errors... ffs So now I have dialed things down even more and at this point, unless I am clearly missing a setting here (for stability), I give up and will just get a 12900k later in the year when DDR5 is readily available. Pretty sure I have a dud for a CPU when it comes to CO: Core 5, 6 and 7 seem to only do -5, the rest can do -15 (although were previously stable at -20 but now it wont even boot with that).

Shame.

outpt

this is only a 5800x it might be of some use for you. ppt:120w,tdc:85A,edc:108A max boost +50mhz Cores set to C0-6, C1-6, C2-16 ,C3-16, C4-16, C5-16 ,C6-5 ,C7-5. GSS-Disabled not a engineer but it seems that core 2,3,4,5 are "grouped" together and i was abled to more negative offset. still working on this idea. Cores 0 and 1 are by there selves as are core 6and 7. A-XMP is enabled for 3600mhz memory in testing i do not get whea errors but event id:41which cause bsod or it simple restarts if this happens 2 times in a row windows will go into a repair state and ask if i want boot back into windows or use a usb device to repair. boot back into windows and use Tweaking.com-windows repair and follow the steps for the program this works like a charm and i am able to put the last working curve offsets as stated above. core 6and 7 will not tolerate no more than -5 this known. the other cores i am still working on as i am in no hurry. my 5900x dose not have these problems so far still working on this R9 cpu. i think that most of my problems revolve setting up CO absolutely correct. now i can get both CO and pbo to work together. i have looked thought the forums here and have not the core clusters like i have.

power plan set to best performance and idle voltage does not drop.

Last edited: Nov 30, 2021

iiNNeX

So I raised EDC to 200 and TDC to 145, I then re-did the CO with the strongest 3 cores getting -6 and everything else -16. This is the best scores I can get after an hour or so of messing around with stress tests, at XMP profile ram (3600 CL16)

Going to run a few more hours of various stress tests to make sure its 100% rock solid, before I start tuning the memory, I am going to go for a 3800 CL16 Fast preset using DRAM calculator and see how I get on. Should help push me over the 30k in cinebench and if it's stable, that is all the performance I need for now. I'd like a bit better single-core but it is what it is.

outpt

for me gaming is the acid test. CC dose not equal stability or any other stress test

iiNNeX

for me gaming is the acid test. CC dose not equal stability or any other stress test

Yeah I agree, once stable in benches and stress programs, I will run various games for a few hours as I would normally and confirm stability.

freeagent

For me it’s test mem 5 lol..

it runs those cores so fast, I usually get a thread exception error. I am just running an all core clock right now at 4500. It’s pretty boring.

Haytham

I registered here specifically in order to present my problem with my processor and I'm not good with OC I just try for the first time to OC my CPU First of all I just need the best temps and best low voltage to my CPU keeping the performance at the best as possible I don't need crazy numbers. The first Issue when I set PBO to manual and add PPT and TDC and EDC and enable Curve and making It -16 and enable XMP My PC It work but Restart randomly. The Second Issue i found the bench in CPU Z give me the same result with XMP or without XMP so by default my ram run at 2666 and with XMP 3600 no change In CPU Z score but I can see the bench give me little high score with 2666 !!! Also my PC Restart randomly more with XMP Enabled. It restarts in both but more with XMP. I have latest Bios Update already Installed.

Any advise ?

Ibizadr

I registered here specifically in order to present my problem with my processor and I'm not good with OC I just try for the first time to OC my CPU First of all I just need the best temps and best low voltage to my CPU keeping the performance at the best as possible I don't need crazy numbers. The first Issue when I set PBO to manual and add PPT and TDC and EDC and enable Curve and making It -16 and enable XMP My PC It work but Restart randomly. The Second Issue i found the bench in CPU Z give me the same result with XMP or without XMP so by default my ram run at 2666 and with XMP 3600 no change In CPU Z score but I can see the bench give me little high score with 2666 !!! Also my PC Restart randomly more with XMP Enabled. It restarts in both but more with XMP. I have latest Bios Update already Installed.

Any advise ?

Maybe -16 was to much in some cores of your processor. You need to see in event viewer if you got whea error, after that open the whea error and see what apic ID it was, like APIC ID 0 OR 7 OR 3. See it and told us if you have any whea error. It's not normal with xmp restarts so I assume it was to aggressive co in some core

Haytham

Maybe -16 was to much in some cores of your processor. You need to see in event viewer if you got whea error, after that open the whea error and see what apic ID it was, like APIC ID 0 OR 7 OR 3. See it and told us if you have any whea error. It's not normal with xmp restarts so I assume it was to aggressive co in some core

Thank you for your reply I attached the pic from Event Viewer

I will try to lower the Curve and will try to make It per core

  • kp.png
  • whea.png

Ibizadr

Thank you for your reply I attached the pic from Event Viewer

I will try to lower the Curve and will try to make It per core

Its not that you need to find. You need to find something like my attach pic to see what core fail. Pic it's not mine its from Google but it's what you need to find in event viewer.

Try per core its the best way but it consumes time. You can run now with your settings corecycler and see what core cause problems.

  • 90a4ff47-3c8d-4256-9ee0-c1120c6e81a2.png

Haytham

Its not that you need to find. You need to find something like my attach pic to see what core fail. Pic it's not mine its from Google but it's what you need to find in event viewer.

Try per core its the best way but it consumes time. You can run now with your settings corecycler and see what core cause problems.

I attached the pic from my PC like what you said, I hope this can make us fix the issue

  • wweha.png

Ibizadr

I attached the pic from my PC like what you said, I hope this can make us fix the issue

But this is whea 46 you need to find some whea 19 error, related to apic ID. Or your problem that's not a curve optimizer but other thing

I only use Radeon software to OC my CPU.

Screenshot 2022-02-04 144211.png

Haytham

But this is whea 46 you need to find some whea 19 error, related to apic ID. Or your problem that's not a curve optimizer but other thing

You are right the Issue from the fast cores and slow one each one need to be optimized alone. I found the faster two cores from Ryzen master now The problem has been solved.

Thank you


Page 14

jesdals

I registered here specifically in order to present my problem with my processor and I'm not good with OC I just try for the first time to OC my CPU First of all I just need the best temps and best low voltage to my CPU keeping the performance at the best as possible I don't need crazy numbers. The first Issue when I set PBO to manual and add PPT and TDC and EDC and enable Curve and making It -16 and enable XMP My PC It work but Restart randomly. The Second Issue i found the bench in CPU Z give me the same result with XMP or without XMP so by default my ram run at 2666 and with XMP 3600 no change In CPU Z score but I can see the bench give me little high score with 2666 !!! Also my PC Restart randomly more with XMP Enabled. It restarts in both but more with XMP. I have latest Bios Update already Installed.

Any advise ?

Its proberly "safer" or more correct to start with tuning the memory settings before you start with the curve settings - theres guides arround 1usmus in here https://www.techpowerup.com/download/ryzen-dram-calculator/ B450 chipset will be less flexible with 2x16 gb modules than others using 8gb e.g. with Samsung B-die so expect results to be lesser than those with these

You can read about AMD memory settings here as well https://www.techpowerup.com/review/neo-forza-faye-ddr4-5000-2x-8gb/

When memory and Infinity fabric is stable - then its time for PBO and Curve optimization - would actually play around with the PBO first and then ad the Curve settings last

jesdals

Did some testing and found that raising EDC above TDC gave lesser single and multicore performance

EDC.jpg

running with EDC at 115 gave 10 points more in single thread performance and about 100 points in multiperformance - so with my setup it did not bring better performance

jesdals

New AMD Chipset driver - seems to down step all core boost clocks - done testing for 2 days and it seems that PBO boost is down with 100-150MHz compared with earlier versions Article Number RN-RYZEN-CHIPSET-4-03-03-431

Release Highlights​

  • Fixed PSP driver downgrade issues
  • Fixed some text on Russian OS language pack

Known Issues​

  • Sometimes custom install fails to upgrade to latest drivers.
  • Text alignment issues may be seen on Russian language.
  • Manual system restart required on Non-English OS after the installation is complete.
  • Windows® Installer pop-up message may appear during the installation.
  • Uninstall summary log may incorrectly show uninstall status as fail on non-English OS.
  • May observe a pop-up message "AMD Chipset Software is not responding" when the installer is launched and UI screen is clicked.

Ferrum Master

Don't put EDC past 140A

It will disable boost ie voltages past 1.4V. It is a known issue.

jesdals

Don't put EDC past 140A

It will disable boost ie voltages past 1.4V. It is a known issue.

Hmm have been testing and running with these settings for a while - but setting 245/125/140 did give better results single and multi core! Will test for stability and results with current bios and settings - thanks Ferrum Master!

:lovetpu:

Ibizadr

Hmm have been testing and running with these settings for a while - but setting 245/125/140 did give better results single and multi core! Will test for stability and results with current bios and settings - thanks Ferrum Master!

:lovetpu:

I don't believe in that since the problem with Edc 140 was related to a bios update and not to a chipset driver update. You got better results because maybe you keep your cpu not to hot compared to other settings you use in past.

jesdals

Well results seems better

1647437667342.png

  • 1647437491613.png

Makaveli

I don't believe in that since the problem with Edc 140 was related to a bios update and not to a chipset driver update. You got better results because maybe you keep your cpu not to hot compared to other settings you use in past.

This EDC 140 issue is due to the new bios not chipset drivers. I have downloaded the chipset drivers and will be installing over the weekend to see what it does.

Ferrum Master

How long has this been the case? And how did you find out about it?

Saw it somewhere on Reddit and overclockers forum also.

It applies only to the latest AGESA V2 PI 1.2.0.6b, I was using betas, so I saw something weird for few weeks already.

Ibizadr

Saw it somewhere on Reddit and overclockers forum also.

It applies only to the latest AGESA V2 PI 1.2.0.6b, I was using betas, so I saw something weird for few weeks already.

After 1.2.0.3c Edc past 140 don't boost your cpu higher than before.

After I give a try to the last chipset update I'm rolling it back since I experience yesterday a weird shutdown and on boot ask me to go to bios like I change something and today I got an hard crash (first one this build with 10months) need to shutdown on power and after that no boot. Need to change ram slots (ram led bug on motherboard) and clear cmos. Tomorrow I will rolling it back. AMD messing a lot with latests bios and chipset updates.

Dia01

I can't seem to get too much more past 670 single core score CPUZ. Raising the PBO limits have no effect, thinking just a combination of my chip and the X570M shitty VRM's... PBO Settings: PPT - 142W TDC - 95A EDC - 100A 1st four best cores: -10 Next best 4 cores: -15 All remaining cores: -20

1650271150268.png

jesdals

@Dia01 Well considering that your Infinity fabric is 1800 and compared with the rest I think your scores is pretty fine. One thing your curve setting at -10 and -15. I found that the best cores could take the lowest value, so in your case I would try setting the "bad cores" at -15 and the god cores at -20 and the best at -25.

1650274111981.png

Dia01

"bad cores" at -15 and the god cores at -20 and the best at -25 not stable for me. Tried "bad cores" at -10 and the god cores at -15 and the best at -20, slightly worse score than before. I think I'm where it is likely to be in all honesty.

1650275574439.png

do we have any programs available outside corecycler for testing max boost freq for curve optimizer ? corecycler is good up to around ~4925, outside of that there's still up to 5100 boost for background apps or games yesterday had CO-24 pass corecycler 8h but of course just 2mins of Destiny2 5.1 boost for few seconds and got a nice bsod.

1650474183715.png

1650474472624.png

*never mind, heaven benchmark pushes those high freq randomly I'll loop that overnight

Last edited: Apr 20, 2022

"bad cores" at -15 and the god cores at -20 and the best at -25 not stable for me. Tried "bad cores" at -10 and the god cores at -15 and the best at -20, slightly worse score than before. I think I'm where it is likely to be in all honesty.

View attachment 243982

i found that setting the power limit to "motherboard" in the bios works best for max cpu performance.

Dia01

i found that setting the power limit to "motherboard" in the bios works best for max cpu performance.

My motherboard pre-configured limits are: PPT - 142W TDC - 95A EDC - 140A

I've set the limits to 'Motherboard' and also set manually, I've found no difference raising EDC. Lowering EDC to 100A seems to be the best balance for some reason.

freeagent

Honestly, I used Linpck Xtreme to find the max the CPU will put out, so for me running that program with PBO and APE enabled, I noticed the CPU will do 225w PPT. It will do 153 TDC and 180 EDC, so thats what I set my limits to, and played with CO to get a decent tune going... its worked for me except in Core Cycler. But in absolutely every benchmark in Benchmate, every stress test, daily use, games, time.. It has been cock solid. But core cycler fails pretty hard.. To pass CC I have to dial things back quite a bit.. the only program. Grrr.. gets me fired up every time :laugh:

tabascosauz

:laugh: I dunno if I would daily something that fails default corecycler, but passing default config is good enough 99% of the time. Won't be crashing on you. Thoroughly testing is just for the extra peace of mind since boost is inherently unpredictable. For default config my preferred cores are -5/-8 and most of the 10 others up to -25 or -30. For truly stable, it's -2/-5 and most at -20 or so.

For non-occasional crashing you'd need like -30 on cores that have no business being over -10

How do my settings look, I'm completely new to PBO/CO overclocking though after messing around in the bios tonight, these settings look to be the best I have gotten in CPU-Z bench so far, temps all seem to be in check, RAM at 4000mhz and 1:1 IF clock, Ryzen 5500 CPU and some TeamGroup RAMK which I believe is Samsung B-die, though inconclusive

  • image_2022-04-22_000520075.png

Dia01

How do my settings look, I'm completely new to PBO/CO overclocking though after messing around in the bios tonight, these settings look to be the best I have gotten in CPU-Z bench so far, temps all seem to be in check, RAM at 4000mhz and 1:1 IF clock, Ryzen 5500 CPU and some TeamGroup RAMK which I believe is Samsung B-die, though inconclusive

Have you played around with the Boost Clock Override?

freeagent

I dunno if I would daily something that fails default corecycler

My 5600X was easy :D

I got 4800 CC stable, default anyways.. 5900X is gonna take a bit.. I did start, have a few hours in.. didn't really like what I saw lol.. might as well just run it at stock :D


Page 15

tabascosauz

How do my settings look, I'm completely new to PBO/CO overclocking though after messing around in the bios tonight, these settings look to be the best I have gotten in CPU-Z bench so far, temps all seem to be in check, RAM at 4000mhz and 1:1 IF clock, Ryzen 5500 CPU and some TeamGroup RAMK which I believe is Samsung B-die, though inconclusive

Go to BIOS > AMD CBS and disable TSME. You are penalizing yourself by 7-10ns Boost Clock Override tops out at 200 and has been for a long time TSME aside, latency is high for 4000CL16 but Im assuming that's because you haven't touched tRFC yet

Use zentimings if you want to get into mem stuff

Last edited: Apr 22, 2022

Would bios settings for AMD overclock/CBS etc

Go to BIOS > AMD CBS and disable TSME. You are penalizing yourself by 7-10ns Boost Clock Override tops out at 200 and has been for a long time

TSME aside, latency is high for 4000CL16 but Im assuming that's because you haven't touched tRFC yet

I can't find any TSME setting in my bios at all, I knocked the TRFC down a tad as it was about 560 using default settings, though wanted to be careful as have had issues lowering it too far too soon in the past with these sticks, though that was on a Ryzen 1600AF when I could only run max 3400mhz stable

tabascosauz

Would bios settings for AMD overclock/CBS etc

I can't find any TSME setting in my bios at all, I knocked the TRFC down a tad as it was about 560 using default settings, though wanted to be careful as have had issues lowering it too far too soon in the past with these sticks, though that was on a Ryzen 1600AF when I could only run max 3400mhz stable

tRFC gets kinda easier to run "tighter" ns-wise the higher speed you go. Because you need more VDIMM to push lower tRFC, but each kit has a hard lower limit to tRFC (in number not ns) where it won't boot anymore. If it doesn't do sub 200ns tRFC, it's not B-die. No matter how horrible the bin. Even legacy E-die does 160ns. 140ns can be a little tough for bad B-die but 160ns should be no prob Some vendors now put TSME on the main page under CBS. For others it should still be under UMC settings or something under CBS. And then under Security I think. MSI should be fine, don't recall it being that hard to find on my Unify-X.

Reous tRFC list v21.png

I have MSI B450m Pro-A Max, it's a turd of a board.. still can't see any resemblance of that setting.. Interesting thing though with the TRFC, tried to run low 300's on my 1600AF and it didn't like it at all though that was likely due to the 1600AF IMC and not the RAM, I'm at 400 now, I shall go deeper down the rabbit hole lol

image_2022-04-22_012743891.png

tRFC gets kinda easier to run "tighter" ns-wise the higher speed you go. Because you need more VDIMM to push lower tRFC, but each kit has a hard lower limit to tRFC (in number not ns) where it won't boot anymore. If it doesn't do sub 200ns tRFC, it's not B-die. No matter how horrible the bin. Even legacy E-die does 160ns. 140ns can be a little tough for bad B-die but 160ns should be no prob Some vendors now put TSME on the main page under CBS. For others it should still be under UMC settings or something under CBS. And then under Security I think. MSI should be fine, don't recall it being that hard to find on my Unify-X.

View attachment 244473

b-die is the way to go for lowest tRFC, low tRFC works very well on amd, every -10trfc nets you a good .1ns better latency here's my patriot 4400cl19 kits

1650589394637.png

tabascosauz

b-die is the way to go for lowest tRFC, low tRFC works very well on amd, every -10trfc nets you a good .1ns better latency

here's my patriot 4400cl19 kits

Yes, but at the same time AIDA is also disproportionately a sucker for tRFC, doesn't quite translate to real performance. Big jumps on tRFC make big gains, but smaller differences below 140ns is not quite as noticeable. Other timings have real impact as well but AIDA ignores them. Tradeoff between VDIMM and tRFC. 1.56V set on Gigabyte boards is closer to 1.6V real, since all Gigabyte boards overvolt VDIMM between 0.03-0.06V.

3800_1.png

I have MSI B450m Pro-A Max, it's a turd of a board.. still can't see any resemblance of that setting.. Interesting thing though with the TRFC, tried to run low 300's on my 1600AF and it didn't like it at all though that was likely due to the 1600AF IMC and not the RAM, I'm at 400 now, I shall go deeper down the rabbit hole lol

View attachment 244474

I'll take a look at my Unify-X BIOS later tonight and send some screenies of CBS menu. Can't help you much if you don't provide your zentimings. I'm confident you can do 145ns or better, but you should expect to bump VDIMM for it. tRFC isn't free.

Would bios settings for AMD overclock/CBS etc

I can't find any TSME setting in my bios at all, I knocked the TRFC down a tad as it was about 560 using default settings, though wanted to be careful as have had issues lowering it too far too soon in the past with these sticks, though that was on a Ryzen 1600AF when I could only run max 3400mhz stable

I'll take a look at my Unify-X BIOS later tonight and send some screenies of CBS menu. Can't help you much if you don't provide your zentimings. I'm confident you can do 145ns or better, but you should expect to bump VDIMM for it. tRFC isn't free.

On my Unify it's under Overclocking/Advanced DRAM Configuration

TSME.jpg

Dia01

Can't locate TSME on the main page or in Advance under AMD CBS which the only sub item is DRAM Memory Mapping

Screenshot 2022-04-22 122507.png

Screenshot 2022-04-22 122530.png

Screenshot 2022-04-22 122548.png

jesdals

I had to dig deep under my bios AMD CBS settings to find the TSME setting

b-die is the way to go for lowest tRFC, low tRFC works very well on amd, every -10trfc nets you a good .1ns better latency here's my patriot 4400cl19 kits

View attachment 244475

Heres mine same type of Patriot - like your settings though

1650621130840.png

Last edited: Apr 22, 2022

Octopuss

tabascosauz

A memory encryption method part that used to be part of AMD's "PRO" suite. Renoir and Cezanne APUs have it but AFAIK zero difference on regular desktop chiplet Ryzen bc they don't support it. The 5600G/5700G aren't PRO but I guess AMD just considers all APUs now to have the security features. Non-X Matisse used to be PRO (3700/3900) but Vermeer doesn't seem to be.

Theoretically good for security, but instant 7ns ish latency penalty, so if pushing the APU UMC, it should be off

Last edited: Apr 26, 2022

Octopuss

I am confused. Is it relevant to 5000 series or not?

If you do have the option, then yeah. In my case once out of the box is hard to return it unless there are real problems with it (remember PBO is actually overclocking). My 5950X was unstable on core 4, (I'm running that one in -15, core 2 and 3 are running -20 and everything else at -25).

What power supply do you have? Could it be the PSU not delivering appropriate power.

I just saw this and I gotta give my advice, because old threads do get views. AMD are easy to get an RMA from. I have to send back a couple 5000 series CPU's because of low performance. If I buy an $800 CPU I want an $800 chip. I love AMD but there is a wide performance margin between a dud 5950X and a good one. They only guarantee 4900mhz, if I don't get 5000+ then I'm not gonna be happy plain and simple. Maybe someone else will be content with that, but I spend wayy too much money building systems. If they didn't take em back, I would end up paying to replace them. My point tho is AMD are pretty easy to return to as long as you have the invoice. They never asked me if I overclocked the CPU, and I doubt they care. They use language like that to deter people. And it works to some degree I'm sure. But it's impossible for them to refuse a return for that reason. Any AIB board would nullify the warranty because they all have pbo+ enabled by default. Here's what they ask for. First, you request an rma and you give them the serial number and describe the problem. They may try to talk you through some troubleshooting steps. I let em know that I have done all the troubleshooting to avoid that. Then they send you an email asking what motherboard and bios you were using. They have you send a picture of the cpu installed in the socket with a piece of paper with the reference number in the shot. And again ask you what troubleshooting steps were taken. An issue I did have was with a GPU I bought off Craigslist last year. I needed the kid I bought it from to do the RMA and they would only mail it to the address where he ordered it from. It was a 6800xt with a pad thermal pad so it died right away and I still had the guys number. He was a good kid luckily and It wasn't an issue. But that policy sucks when so many people are having to buy off scalpers So of course you can't say the chip isn't fast enough, you need to say it's defective in some way, IE it reboots when you load a game.

And I don't advise sending back a chip that is fairly decent to try and hit the lottery. Honestly tho, if you get a 5950X that doesn't wanna boost to 5000mhz you hit the silicon lottery backwards and if it's gonna bother you then send it back.

Octopuss

Anyone using CoreCycler here? I keep getting this when I try to run it:

1651437241430.png

I just replaced the included outdated Prime95 version with 30.7.

What's going on?

Octopuss

Yea well. The author encourages people to do so, so I was perplexed for a while until I found an an acknowledged issue on GitHub from november about it. Guess it will take some time.

Oh and I've just noticed the script defaults FFT size to huge, which is a nonsense for this purpose, isn't it? Shouldn't smallest or small be used instead?

tabascosauz

Yea well. The author encourages people to do so, so I was perplexed for a while until I found an an acknowledged issue on GitHub from november about it. Guess it will take some time.

Oh and I've just noticed the script defaults FFT size to huge, which is a nonsense for this purpose, isn't it? Shouldn't smallest or small be used instead?

No. Maybe if you had a static OC on 12900K. We have boost algorithm here. The point is not to hammer the core. The lighter the load the higher boost clock the PB algorithm allows - the point is to run a workload that is reasonably heavy but allows boost clocks to push as high as they possibly can, to gauge whether the current undervolt settings are stable at those theoretical max clocks. In any case even on Huge FFT, it's heavy enough. I don't think I've seen any game run per-core power up past 20W.

That said, in my testing with 1h15m per core, FFT size selection generally doesn't make a huge difference to clocks in corecycler. I leave it on Huge for a good spread. Possibly because default config and the most effective configs all run SSE instructions so maybe without AVX the algorithm treats FFT sizes pretty equally. AVX testing is not very useful in CC, at least for initial tests - you can test AVX after you've worked out some settings with SSE.

Octopuss

But the config file also says # Moderate: 1344K to 4096K - special preset, recommended in the "Curve Optimizer Guide Ryzen 5000" # Heavy: 4K to 1344K - special preset, recommended in the "Curve Optimizer Guide Ryzen 5000" # HeavyShort: 4K to 160K - special preset, recommended in the "Curve Optimizer Guide Ryzen 5000"

So I am really confused here.

tabascosauz

But the config file also says # Moderate: 1344K to 4096K - special preset, recommended in the "Curve Optimizer Guide Ryzen 5000" # Heavy: 4K to 1344K - special preset, recommended in the "Curve Optimizer Guide Ryzen 5000" # HeavyShort: 4K to 160K - special preset, recommended in the "Curve Optimizer Guide Ryzen 5000"

So I am really confused here.

Apologies, I said Huge FFT - my config is 1h15m All FFT. What's there to be confused about? Moderate, Heavy and HeavyShort are just 3 extra profiles to choose from with different FFT ranges. I have no idea about this "Ryzen 5000 guide" but they look just fine, just remember to set the appropriate runtime. Don't get bogged down in names - look at the FFT size you want to test, then look at the recommended test lengths for said FFT profile up the page. Set something that'll test the entirety of the range you want to test, and set it to run at least 3 iterations per core. Boost is unpredictable so 1 or 2 cycles is not enough to prove stability.

Default 6min Huge FFT is generally enough to eliminating crashing/BSOD in day-to-day. After running 3-5 iterations per core with 0 errors, I'd be pretty confident about running it daily. But if you are a stickler for stability, don't be afraid of longer custom testing, just do two cores at a time overnight.

Octopuss

Yea, extra profiles, different sizes, not sure what to choose and why :) Up until now I always chose the smallest size when running p95 manually, but now I'm not sure after what you told me in relationship with the Huge preser.

I left everything on default and set it to only cycle through one core. If it passes 12 hours, I'd move onto the next. I guess that does enough iterations.


Page 16

tabascosauz

Yea, extra profiles, different sizes, not sure what to choose and why :) Up until now I always chose the smallest size when running p95 manually, but now I'm not sure after what you told me in relationship with the Huge preser.

I left everything on default and set it to only cycle through one core. If it passes 12 hours, I'd move onto the next. I guess that does enough iterations.

Still use P95 Small as a torture test for Intel OCs. It's just not the right tool for the job here. Like if you were stressing memory controllers, you would run P95 Large FFT. If you're still doing 6 minute config and doing 1 core at a time, just run it a couple times and go to the next. Save the long tests for a comprehensive config. Huge FFT takes what, 30-45 min to actually test the entire range? 6 minutes only gets around to testing a small snippet, it's a quick and dirty test.

I would still test at least 2 cores at a time. That way they alternate and can take a break from the heat. Preferably cores on different CCDs as well, unfortunately not possible for 5800X

Octopuss

What exactly does an iteration mean in the context of this script btw? Like one pass of a specific FFS size?

It looks like the defaults might be a little misconfigured though, if default size is Huge and time 6 mins...

tabascosauz

What exactly does an iteration mean in the context of this script btw? Like one pass of a specific FFS size?

It looks like the defaults might be a little misconfigured though, if default size is Huge and time 6 mins...

1 iteration is just running through the length of the test once on 1 core. You can set the # of rounds in config. It'll run in the core order specified in config, then repeat. It isn't 100%, but 6 min Huge is a good balance of speed and effectiveness. Strictly speaking it's probably not "stable", as you'll quickly find if you run eg. 1 hour All afterwards, but if you can pass 6 min Huge you *shouldn't* really have any big issues day-to-day with gaming and regular use. It's a bit different than mem stability where no number of errors is acceptable, because chances are you won't be able to match corecycler clocks in daily use (unless your CPU is so good it easily tops out at +200MHz)

Properly testing is best, but it takes a long time. Better for 6-8 core parts, but it took me like 2.5 weeks to run the entire CPU through 1h15m All.

Octopuss

I am still pretty confused by all this. It looked dead simple at first... How did it take 2,5 weeks though? If I take a look at the config and add up the numbers there, even if I double them, that's a few hours at most. I only had one coffee today though! Maybe I am missing something essential.

The script is pretty complicated.

tabascosauz

I am still pretty confused by all this. It looked dead simple at first... How did it take 2,5 weeks though? If I take a look at the config and add up the numbers there, even if I double them, that's a few hours at most. I only had one coffee today though! Maybe I am missing something essential.

The script is pretty complicated.

4-5 iterations per core, 1h15m per iteration, 2 cores testing per night. Some cores required re-adjusting so more nights retesting new settings. And even then 10 of the cores are set in multiples of 5 (-10, -15, -20 etc) because I just couldn't be bothered at that point. With default config, you can easily be done in just 1 day. Just do that, and if you find that it's not stable enough, then go for the long tests.

I never really had any real issues with settings that pass default config, on either 5900X, 5600G or 5700G. Just wanted to be doubly sure for my 5900X daily.

Octopuss

So anyway, if I want to test it really thoroughly, several hours on each core with Huge size (or maybe All) should be pretty safe?

timings and aida bench attached, any advice on my timings?

  • aida.png
  • zentiming.png

Should be fine yeah if you have the time for it.

This is the Curve Optimizer thread, nothing to do with timings or AIDA performance.

So my posts on the previous page regarding TSME you replied to shouldn't be here either?

tabascosauz

So my posts on the previous page regarding TSME you replied to shouldn't be here either?

I won't delete those, that started off with CPU performance and CPU-Z results and eventually meandered into the other stuff. Ryzen discussion tends to become a bit of a melting pot anyway. But let's try to keep purely mem-related stuff in the other threads.

I won't delete those, that started off with CPU performance and CPU-Z results and eventually meandered into the other stuff. Ryzen discussion tends to become a bit of a melting pot anyway. But let's try to keep purely mem-related stuff in the other threads.

My point was, I started off asking about curve optimiser and PBO settings and then it got on to memory timings and YOU started the TSME talk that lasted for nearly a page... so now I should just start a new thread and disregard the previous posts even though there are literallly tons of the same?

Go to BIOS > AMD CBS and disable TSME. You are penalizing yourself by 7-10ns Boost Clock Override tops out at 200 and has been for a long time TSME aside, latency is high for 4000CL16 but Im assuming that's because you haven't touched tRFC yet

Use zentimings if you want to get into mem stuff

Would bios settings for AMD overclock/CBS etc

I can't find any TSME setting in my bios at all, I knocked the TRFC down a tad as it was about 560 using default settings, though wanted to be careful as have had issues lowering it too far too soon in the past with these sticks, though that was on a Ryzen 1600AF when I could only run max 3400mhz stable

tRFC gets kinda easier to run "tighter" ns-wise the higher speed you go. Because you need more VDIMM to push lower tRFC, but each kit has a hard lower limit to tRFC (in number not ns) where it won't boot anymore. If it doesn't do sub 200ns tRFC, it's not B-die. No matter how horrible the bin. Even legacy E-die does 160ns. 140ns can be a little tough for bad B-die but 160ns should be no prob Some vendors now put TSME on the main page under CBS. For others it should still be under UMC settings or something under CBS. And then under Security I think. MSI should be fine, don't recall it being that hard to find on my Unify-X.

View attachment 244473

I have MSI B450m Pro-A Max, it's a turd of a board.. still can't see any resemblance of that setting.. Interesting thing though with the TRFC, tried to run low 300's on my 1600AF and it didn't like it at all though that was likely due to the 1600AF IMC and not the RAM, I'm at 400 now, I shall go deeper down the rabbit hole lol

View attachment 244474

b-die is the way to go for lowest tRFC, low tRFC works very well on amd, every -10trfc nets you a good .1ns better latency here's my patriot 4400cl19 kits

View attachment 244475

Yes, but at the same time AIDA is also disproportionately a sucker for tRFC, doesn't quite translate to real performance. Big jumps on tRFC make big gains, but smaller differences below 140ns is not quite as noticeable. Other timings have real impact as well but AIDA ignores them. Tradeoff between VDIMM and tRFC. 1.56V set on Gigabyte boards is closer to 1.6V real, since all Gigabyte boards overvolt VDIMM between 0.03-0.06V.

View attachment 244477

I'll take a look at my Unify-X BIOS later tonight and send some screenies of CBS menu. Can't help you much if you don't provide your zentimings. I'm confident you can do 145ns or better, but you should expect to bump VDIMM for it. tRFC isn't free.

Ya, cba with that, I'll figure it out myself, thanks.

GoldenX

Tested both the new Ryzen Master and HYDRA-PRO, both are unstable. Good think I did it manually.

tabascosauz

My point was, I started off asking about curve optimiser and PBO settings and then it got on to memory timings and YOU started the TSME talk that lasted for nearly a page... so now I should just start a new thread and disregard the previous posts even though there are literallly tons of the same?

Ya, cba with that, I'll figure it out myself, thanks.

I didn't reprimand you, just saying that we have a number of threads that will generate more discussion for pure mem stuff. You are welcome to create a new thread if you don't feel any existing ones are appropriate. We haven't done a great job of keeping things strictly on the original topic here, and I did have a hand in that. But that was also two weeks ago, and you expressed an initial interest in tweaking CO. On topic has to start somewhere.

Ibizadr

timings and aida bench attached, any advice on my timings?

First advice, focus only one thing and the other after first concluded with success. Put default bios on it and focus on ram first. Try 1t GDM off, 3800/1900 mclk==fclk. You need to test ram a lot. And after that come to this thread to improve your cpu, but next time discuss on Aida thread.

First advice, focus only one thing and the other after first concluded with success. Put default bios on it and focus on ram first. Try 1t GDM off, 3800/1900 mclk==fclk. You need to test ram a lot. And after that come to this thread to improve your cpu, but next time discuss on Aida thread.

I've used default XMP 3000 settings, bear in mind I'm running at 3800 now with the same basic settings, have tweaked trfc and tfaw used curve optimiser to boost to 4450mhz with 3800mhz RAM for 1:1

Ibizadr

I've used default XMP 3000 settings, bear in mind I'm running at 3800 now with the same basic settings, have tweaked trfc and tfaw used curve optimiser to boost to 4450mhz with 3800mhz RAM for 1:1

Your timings needs improvements take a look in Aida topic

Octopuss

Is it possible to be able to use higher negative offsets a year after thoroughly testing a CPU? I don't think BIOS updates make much of a difference.
On the other hand, I was running Prime95 manually with the smallest or small FFTs back then. Also ran two threads of it on the core I waz testing, and I'm not sure what the CoreCycler script actually does, because I don't quite understand that from the config file. Logic says you should run two threads with SMT/HT on, but...

AVATARAT

Is it possible to be able to use higher negative offsets a year after thoroughly testing a CPU? I don't think BIOS updates make much of a difference.
On the other hand, I was running Prime95 manually with the smallest or small FFTs back then. Also ran two threads of it on the core I waz testing, and I'm not sure what the CoreCycler script actually does, because I don't quite understand that from the config file. Logic says you should run two threads with SMT/HT on, but...

I wrote here how I test it.

Octopuss

Also wtf, I increased the negative offset on two cores and now my all core (I think, it's what Afterburner shows) boosts are lower? That makes no sense. Previously I usually saw 4800MHz and now it fluctuates between 4650 and 4725.

I wrote here how I test it.

How would the script ovevolt cores? That's nonsense.

tabascosauz

I wrote here how I test it.

Overvolting cores? PBO On causing Vcore to spike over 1.5V is common but CO causing it is a new idea.

Also wtf, I increased the negative offset on two cores and now my all core (I think, it's what Afterburner shows) boosts are lower? That makes no sense. Previously I usually saw 4800MHz and now it fluctuates between 4650 and 4725.

How would the script ovevolt cores? That's nonsense.

If you want to properly gauge your clocks in any sort of sustained load, use HWInfo. Either look only at the Effective Clock metric, or enable Snapshot Polling in settings (right click tray icon) and look at Core Clock/Effective Clock. Nothing else (except maybe Ryzen Master) properly shows a realistic clock. Hitting 5150/5250MHz "core clock" (just an example) without having similar Effective numbers to back it up means it will never sustain long enough to actually make a difference in benchmarks or effective clock.

Real 4.8GHz all-core is very, very hot on Ryzen. Afterburner OSD is a classic culprit of bamboozling you into thinking the CPU is working a lot harder than it actually is.

AVATARAT

Also wtf, I increased the negative offset on two cores and now my all core (I think, it's what Afterburner shows) boosts are lower? That makes no sense. Previously I usually saw 4800MHz and now it fluctuates between 4650 and 4725.

How would the script ovevolt cores? That's nonsense.

The problem is not in the script, but if you use it, it will throw errors until you put enough or more voltage in the Curve on the core that fails. So you just need to test it.
With SSE the cores work on a higher frequency and need not so high voltage to not fail, and that is what I use with OCCT.

Octopuss

I have just realized I never touched the PBO limit. Does increasing it by the usual +200MHz have any negative implications? What exactly does it do anyway?

And what is the default frequency cap of 5800X?

oobymach

I have just realized I never touched the PBO limit. Does increasing it by the usual +200MHz have any negative implications? What exactly does it do anyway?

And what is the default frequency cap of 5800X?

Stability in CoreCycler is not guaranteed with +200 but regular p95 will be stable. It pushes every core to perform an additional 200mhz faster so if you were getting 4.65ghz you'll get 4.85ghz. Also you may have different results in curve with the extra speed you may need more voltage.

tabascosauz

I have just realized I never touched the PBO limit. Does increasing it by the usual +200MHz have any negative implications? What exactly does it do anyway?

And what is the default frequency cap of 5800X?

+200 raises the global frequency limit. Stock for 5800X is 4850MHz, so it goes up to 5050. Global limit is only a theoretical ceiling. The problem is it's really silicon dependent. You will notice in HWinfo the global limit goes up to 5050MHz, but only works if the cores usually picked for ST tasks (usually top 2 ranked cores) are actually capable of sustaining more than 4850MHz effective. eg. on a mediocre CPU you can change Boost Override all you want, but it won't make a difference.

Also, stable undervolt at 4850MHz may no longer be stable with higher boost override. You'll just have to test and see. If you are already maxing out at 4850MHz effective, good chance you can go further.

AVATARAT

I have just realized I never touched the PBO limit. Does increasing it by the usual +200MHz have any negative implications? What exactly does it do anyway?

And what is the default frequency cap of 5800X?

You will increase performance with this 200MHz but you need to play again with CO to stabilize it (add more voltage to be stable).


Page 17

Octopuss

But if raising the limit by 200MHz requires lower CO offets, doesn't that mean the cores would only hit lower frequencies than before?

Octopuss

It's just confusing from defaults point of view. If lowering the core voltages by negative offsets lets the cores clock higher, increasing the voltages doesn't seem logical to be beneficial for frequencies even if the "cap" is removed. I just can't wrap my head around this. Btw I couldn't google the max boost frequency anywhere. AMD's website only says 4,7GHz without any details. So I ran two rounds of iterations of 6m Huge FFT size and these are the maximum frequencies:

1651681547210.png

Do you think I should increase the limits or the CPU is not good enough?

The current offsets are 30, 15, 20, 15, 30, 25, 15, 25.

AVATARAT

PBO, +200MHz, you probably need to lower current offsets. Or something like this, because your mobo is like main (with bios 1602):

1651684131669.png

1651684239976.png

Det0x

Maybe this is the wrong thread for this, but could not fine one dedicated to the 5800x3d (?) Let me know and i will delete the post if there is a problem Have finally completed my first game comparison between my maxed 5950x, my 5800x3d and golden samples Alderlake cpus @ ~5.6 7200MT/s vs
Shadow of the Tomb Raider: 1080p lowest:

  • 5950x @ ~5100/5000mhz = 353fps average cpu game
  • 12900k @ 5750mhz 4300MT/s CL14 = 373fps average cpu game
  • 5800x3d @ 4450mhz = 402fps average cpu game

Horizon Zero Dawn: 1080p performance preset, lowest res scale:

  • 5950x @ ~5100mhz = 301fps average cpu game
  • 12900k @ 5580mhz 7160MT/s CL30 = 321fps average cpu game
  • 5800x3d @ 4450mhz = 313fps average cpu game

F1 2020 1080p low dx11: Australia benchmark location and dry weather:

  • 5950x @ ~5100mhz = 490 average fps
  • 5800x3d @ 4450mhz = 555 average fps

Farcry6 1080p ultra, HD-texture enabled, FSR QTY:

  • 5950x @ ~5100mhz = 163 average fps
  • 12900k @ 5580mhz 7160MT/s CL30 = 203 average fps
  • 5800x3d @ 4450mhz = 176 average fps

Cyberpunk 2077: 1080p low

  • 5950x @ ~5100mhz = 252 average fps
  • 12900k @ 5580mhz 7160MT/s CL30 = 304 average fps
  • 5800x3d @ 4450mhz = 268 average fps

Final Fantasy XV 1080p low: (game engine limited, results with grain of salt)

  • 5950x @ ~5100mhz = 23426 score
  • 12900k @ ~5700mhz(?) = 23585 score
  • 5800x3d @ 4450mhz = 23489 score

Final Fantasy XIV: Endwalker 1440p maximum: (game engine limited, results with grain of salt)

  • 5950x @ ~5100mhz = 30553 score
  • 12900k @ ~5700mhz(?) = 33891 score
  • 5800x3d @ 4450mhz = 33764 score
  • 12900k @ 5500mhz 4133MT/s CL16 = 954 fps
  • 5800x3d @ 4450mhz = 716 fps

Max tuned Alderlake beats out Zen3 in games, but i have to say i'm pretty impressed by the5800x3 :) ...Considering zero binning/golden samples are required for 5800x3d and they dont scale with memory, so you can use a cheapo x470 together with 3200/3600MT/s memory for nice "low-end gaming machine".

edit

Added some of the Alder lake screens:

1651736354010.png1651736373212.png1651736386493.png1651736402355.png

*edit2* My old Spectre install was bloated at this point, so i made a new win10 install for the 5800x3d runs.. (earlier when i did the 5950x runs the spectre install were pretty slim)

And i have no control over the Alder Lake windows installs, but those runs are cherry picked super golden samples with the highest scores i could find.. (binned cores+binned IMC+super cooling+best of the best memory) Anyone feel free to improve the AL numbers if you can :)

For this to be a even more fair comparison of "maxed out gaming systems" i would need a new motherboard with an external clockgen so i could run the 5800x3d at ~4900mhz instead of 4450mhz, but i dont think that would change the outcome much in the end either.. Alder Lake should still come out on top.

Mussels

I have just realized I never touched the PBO limit. Does increasing it by the usual +200MHz have any negative implications? What exactly does it do anyway?

And what is the default frequency cap of 5800X?

Adds 200Mhz to the low core count boost, from 4850Mhz to 5050Mhz on the 5800x - If there is thermal and power room to spare

If you're undervolting with CO, you may need to reduce it for stability

Ibizadr

Adds 200Mhz to the low core count boost, from 4850Mhz to 5050Mhz on the 5800x - If there is thermal and power room to spare

If you're undervolting with CO, you may need to reduce it for stability

I'm at +200 with CO. And 4 of my 5800x cores goes to -30

tabascosauz

But if raising the limit by 200MHz requires lower CO offets, doesn't that mean the cores would only hit lower frequencies than before?

It's just confusing from defaults point of view. If lowering the core voltages by negative offsets lets the cores clock higher, increasing the voltages doesn't seem logical to be beneficial for frequencies even if the "cap" is removed. I just can't wrap my head around this. Btw I couldn't google the max boost frequency anywhere. AMD's website only says 4,7GHz without any details. So I ran two rounds of iterations of 6m Huge FFT size and these are the maximum frequencies:

View attachment 246070

Do you think I should increase the limits or the CPU is not good enough?

The current offsets are 30, 15, 20, 15, 30, 25, 15, 25.

Like I said, you'll just keep fruitlessly going in circles unless you either enable Snapshot Polling or look at Effective metrics. These default "Core Clock" numbers with 25MHz granularity are pointless data. I can see 5050 on "core clock" when I know my 5900X barely manages 4930MHz effective on a good day. Once you do that, test with +0 and see after corecycler whether you are capping out at exactly 4850MHz. If so, then you have additional headroom on those cores by expanding boost override. Stock 5800X boost ceiling is 4850, 5900X is 4950 and 5950X is 5050. It's not advertised explicitly, but all Vermeer and Cezanne CPUs have their Global limit set 50-150MHz above their official rated boost. Setting +200 doesn't hurt anything. If your cores are 4800MHz duds, then nothing happens. If your cores are excellent, then worst case you may need to dial back your undervolt a little bit, but you won't know until you test

Maybe this is the wrong thread for this, but could not fine one dedicated to the 5800x3d (?) Let me know and i will delete the post if there is a problem

Interesting results particularly for 12900K. Might gain more traction to start a new thread for this in General Hardware - I'm sure many people are very interested in seeing more of these comparisons between these 3, it's a hot topic and certainly deserves to have its own thread

Makaveli

I have a golden sample 5800X and using 200+ with these CO settings

1651935958058.png

1651935992945.png

Det0x

Interesting results particularly for 12900K. Might gain more traction to start a new thread for this in General Hardware - I'm sure many people are very interested in seeing more of these comparisons between these 3, it's a hot topic and certainly deserves to have its own thread

Some more numbers, i found the framtimes very interesting.. :) Metro Exodus Benchmark: 1080p low:

  • 5950x @ ~5100mhz = 322 average fps
  • 12900k @ 5580mhz 7160MT/s CL30 = 408 average fps
  • 5800x3d @ 4450mhz = 338 average fps
  • 5800x3d @ 4450mhz, dx11 = 524 average fps

1651948651911.png1651948664716.png1651948681169.png1651948693446.png Metro Exodus Enhanced Benchmark: 1080p low:

  • 12900k @ 5580mhz 7160MT/s CL30 = 250 average fps
  • 5800x3d @ 4450mhz = 237 average fps

1651948751736.png1651948762055.png

**The frametimes for the 5800x3d look much better, no ?
**

1651948798807.png

Where do I find that?
I use HWiNFO for information.

You can use boosttester to see max clocks. It's low load array shuffle. I wouldn't tweak CO with high offset like +200. Not worth much since it only hits 1 or 2 cores at most and will get very unstable. On my 5900x, +150 will get 5.1 boost in games on two best cores and it's not stable past CO-6. With +100 offset, I get 5050 for best two cores and can go CO -19 After the best core, the others will always boost -25mhz,-50mhz,-75mhz etc even with +200 offset; so there's nothing there to gain on dual ccds.

Not sure about 5800x but boosttester will show you the max possible freq for all cores.

all core Y-cruncher (options 7) seem to catch errors very fast vs Prime95 or corecycler that can pass for days with unstable CO.

1652144140095.png

I'm currently at -19 -19 -19 -6 -19 -18 -14 -14 -15 [] []-19 for y-cruncher. Two more cores to test & I can run Prime95 blend & corecycler. 210w and +100 offset I think it's mostly a waste of time on my 5900x b2. It already scores high with low temps. Here's 4 cores -6 & 8 cores -12.

1652143462853.png

Last edited: May 10, 2022

HD64G

I used the Ryzen Master that tested my new 5600 and concluded that the optimal values are -27-28 for all cores apart from the 3rd one that managed -22. I put manually -25 for the strong ones and -20 for the weak one and added +150MHz for the PBO increased limit at the default 76W value and all good for now.

Mussels

I used the Ryzen Master that tested my new 5600 and concluded that the optimal values are -27-28 for all cores apart from the 3rd one that managed -22. I put manually -25 for the strong ones and -20 for the weak one and added +150MHz for the PBO increased limit at the default 76W value and all good for now.

Oh yeah i forgot they added that, now i'm on the new AGESA i'll let it test that overnight

Octopuss

So what am I doing wrong if no cores are boosting over 4850MHz despite adding +200 limit? edit: WTF, I have just gone in the BIOS and found the power limits changed to manual and the frequency override disabled. I didn't change anything.

Looks like a week of testing wasted, because the thing decided to change itself.

Last edited: May 12, 2022

Mussels

So what am I doing wrong if no cores are boosting over 4850MHz despite adding +200 limit? edit: WTF, I have just gone in the BIOS and found the power limits changed to manual and the frequency override disabled. I didn't change anything.

Looks like a week of testing wasted, because the thing decided to change itself.

ryzen master can change those settings if you used its test settings

RM has mine set to -30 but the voltages didnt actually change, CO is broken on my current beta BIOS

AVATARAT

So what am I doing wrong if no cores are boosting over 4850MHz despite adding +200 limit? edit: WTF, I have just gone in the BIOS and found the power limits changed to manual and the frequency override disabled. I didn't change anything.

Looks like a week of testing wasted, because the thing decided to change itself.

Adrenaline video driver can change it. Be careful, it's known problem.

1652347693374.png

Octopuss

I don't use Ryzen Master (tried, looked like bloatware), and I don't have that option in Adrenaline. No idea what happened there. And the stupid ass CPU just doesn't boost past 4800MHz anymore no matter what I do. Doesn't matter if I add 200MHz to the limit or not. I even have slightly lower (actually, higher since negative) offsets on a few cores, and it's actually worse.

I updated the BIOS to the latest version before I started messing with all this again too, hoping I'd get better overclock, but no.

HD64G

ryzen master can change those settings if you used its test settings

RM has mine set to -30 but the voltages didnt actually change, CO is broken on my current beta BIOS

I think if one want to test with Ryzen Master the same settings altered there need to be at auto in UEFI. My PC stopped booting because I already had applied the CO in Ryzen Master and thought that without deactivation there I could do the same in UEFI. It seemed that the settings added to the previous ones and V was so low for the CPU that didn't booted at all.

Mussels

I don't use Ryzen Master (tried, looked like bloatware), and I don't have that option in Adrenaline. No idea what happened there. And the stupid ass CPU just doesn't boost past 4800MHz anymore no matter what I do. Doesn't matter if I add 200MHz to the limit or not. I even have slightly lower (actually, higher since negative) offsets on a few cores, and it's actually worse.

I updated the BIOS to the latest version before I started messing with all this again too, hoping I'd get better overclock, but no.

Theres only been two AGESA's since the 5800x3d came out, and it seems like the regular X gets treated like the 3D with some limits

That said... i like the lower temps, and i see no performance loss.

Octopuss

I think I was on a BIOS from late 2021.

oobymach

I don't use Ryzen Master (tried, looked like bloatware), and I don't have that option in Adrenaline. No idea what happened there. And the stupid ass CPU just doesn't boost past 4800MHz anymore no matter what I do. Doesn't matter if I add 200MHz to the limit or not. I even have slightly lower (actually, higher since negative) offsets on a few cores, and it's actually worse.

I updated the BIOS to the latest version before I started messing with all this again too, hoping I'd get better overclock, but no.

Some bios versions are better than others, see if you can find 1.2.0.3c for your board, that's the best one for 5000 series (except the 5800X3D)

Also the boost is affected by PPT, TDC, and EDC so you may need to adjust PPT, TDC, and EDC to achieve higher clocks, and don't bee too aggressive with values, higher isn't always better especially TDC and EDC.

Octopuss

I don't want to mess with that, I just set limits to motherboard and that's it.
Also not interested in using old BIOSes with bugs. I'm just curious why is significantly newer BIOS with numerous improvements over the course of months affecting performance in negative way.

Tóm tắt
The article discusses the author's latest stability testing results for their AMD 5800X CPU on a B550-F motherboard with BIOS version 1801. The author shares specific core curve offsets and performance benchmarks, noting adjustments made to voltage settings for certain cores after testing with Prime95 and Testmem5. They observed that Testmem5 allowed higher clock speeds due to lower power consumption, leading to a more stable configuration. The author concluded that AMD's default settings are already optimized, and while curve offsets can enhance multi-core performance, they complicate stability testing. The author also mentions plans to update to a newer BIOS version that supports Nvidia's Smart Access Memory. Performance metrics from various benchmarks, including Cinebench and Geekbench, are provided, showcasing the system's capabilities. The author emphasizes caution when adjusting curve offsets and PBO settings, as individual core stability testing is challenging. Overall, the findings suggest that while some performance gains are possible, the risks and complexities involved in overclocking may not yield significant benefits.