Closed madoverlord40 closed 1 month ago
At this point i think its a kernel level issue. The kernel is probably not reporting the CPU correctly.
The estimation is computed from a one second interval between two
TSC
collects. CoreFreq relies on kernel just for the delay; theTSC
being collected directly onto CPU register.A kernel issue would be its own clock source but you would have noticed it if one second is not accurate.
Can you however try the following build option and tell if it makes a difference ?
make -j clean make -j DELAY_TSC=0
Also, just in case of a kernel issue, can you post the output of
/proc/cpuinfo
(first CPU will be enough)
Ok so i just rebuilt your software using make -j DELAY_TSC=0 what info are you needing with that change?
doesnt it need to go through the OS to do that?
Nope, all hardware primitives are made by my own.
Ok so i just rebuilt your software using make -j DELAY_TSC=0 what info are you needing with that change?
Just want to know about BCLK. If it's still 70
or what we want 100 MHz
But how can it process so differently from the previous architectures, especially 7950X
BCLK 70 MHz
is my main issue. TSC looks OK and counts invariantly at processor base frequency but my estimation function takes a ratio as an input parameter; ratio is broken.
@madoverlord40 : With the MSR-Tools installed, can you dump that series of P-State registers ?
## MSR_AMD_PSTATE_DEF_BASE
## P0
rdmsr -p 0 0xc0010064
## P1
rdmsr -p 0 0xc0010065
## P2
rdmsr -p 0 0xc0010066
## P3
rdmsr -p 0 0xc0010067
## P4
rdmsr -p 0 0xc0010068
## P5
rdmsr -p 0 0xc0010069
## P6
rdmsr -p 0 0xc001006a
## P7
rdmsr -p 0 0xc001006b
## MSR_AMD_PSTATE_CURRENT_LIMIT
rdmsr -p 0 0xc0010061
## MSR_AMD_PERF_CTL
rdmsr -p 0 0xc0010062
## MSR_AMD_PERF_STATUS
rdmsr -p 0 0xc0010063
## MSR_AMD_F17H_HW_PSTATE_STATUS
rdmsr -p 0 0xc0010293
Ok so using your turbo feature on core 5, ryzen master shows it to be a fast core with a white dot?
Remark: CoreFreq counts CPU number on a zero base whilst RM is one based
Ok so using your turbo feature on core 5, ryzen master shows it to be a fast core with a white dot?
Remark: CoreFreq counts CPU number on a zero base whilst RM is one based
Oh crap lol does that invalidate the PBO max test i did for you?
doesnt it need to go through the OS to do that?
Nope, all hardware primitives are made by my own.
Ok so i just rebuilt your software using make -j DELAY_TSC=0 what info are you needing with that change?
Just want to know about BCLK. If it's still
70
or what we want100 MHz
But how can it process so differently from the previous architectures, especially 7950X
BCLK
70 MHz
is my main issue. TSC looks OK and counts invariantly at processor base frequency but my estimation function takes a ratio as an input parameter; ratio is broken.
ITs still 70 and still showing a boost off the chart past 61. Also, my system still locks up if i leave your software running ilde for more than like 30min.
Ok so using your turbo feature on core 5, ryzen master shows it to be a fast core with a white dot?
Remark: CoreFreq counts CPU number on a zero base whilst RM is one based
Oh crap lol does that invalidate the PBO max test i did for you?
Luckily those adjacent cores are providing the same PBO score
ITs still 70 and still showing a boost off the chart past 61.
Thanks for your answer. I will debug the Coefficient Of Frequency as soon as I received the MSR dump I claimed previously.
Also, my system still locks up if i leave your software running ilde for more than like 30min.
To make sure of Linux distribution driver conflicts and kernel setup, I can suggest you boot my archlinux-corefreq.iso
to download from version 1.98.2
ISO can be flashed on an USB key with a plain dd
command. See instructions in wiki/Live-CD
Boot the key and when facing the startup menu, choose [Bare Metal]
And let it go...
Once CoreFreq is launched, press k
for the kernel window
s
for the Settings window and proceed to the Governor, CPU-Freq and CPU-idle registration
Q: In that configuration please let me know if system is locking up or not ?
Ok so I booted to the thumb drive with arch linux and it installed the corefreq.ko but cpu-freq dirver will not activate in settings windows. It shows 'off' and when I try to turn it on it says error file exists. It did not lock up however I wonder if not having freq driver on is causing that?
when I try to turn it on it says error file exists
This means a Linux default driver is already acting as CPU-Freq.
That's why there is that [Bare Metal]
entry you have to select immediately when thumb drive boots. You have a few seconds to make the selection.
Fyi, that boot menu entry attempts to blacklist all default CPU-Freq and CPU-Idle kernel modules to leave the seat to CoreFreq
~~Next, press k
for Kernel window: you should read Missing
for the name of drivers.
If true then you are allowed to register from the Settings window.~~
To verify things are OK, reopen the Kernel window: drivers are now assigned to corefreq
In short, select [Bare Metal]
during the thumb drive startup.
It did not lock up however I wonder if not having freq driver on is causing that?
That's the point of the above procedure: to find which Linux kernel module(s) CoreFreq conflicts with.
Most of the time, it has been the k10temp
module.
Q: while running the ISO, had you read a
EDIT: BCLK
of 100
or 70 MHz
?70
in your screenshot.
As soon as you can please provide me the MSR dump as it should give me some clues to debug the frequency ratios.
@madoverlord40 : With the MSR-Tools installed, can you dump that series of P-State registers ?
## MSR_AMD_PSTATE_DEF_BASE ## P0 rdmsr -p 0 0xc0010064 ## P1 rdmsr -p 0 0xc0010065 ## P2 rdmsr -p 0 0xc0010066 ## P3 rdmsr -p 0 0xc0010067 ## P4 rdmsr -p 0 0xc0010068 ## P5 rdmsr -p 0 0xc0010069 ## P6 rdmsr -p 0 0xc001006a ## P7 rdmsr -p 0 0xc001006b ## MSR_AMD_PSTATE_CURRENT_LIMIT rdmsr -p 0 0xc0010061 ## MSR_AMD_PERF_CTL rdmsr -p 0 0xc0010062 ## MSR_AMD_PERF_STATUS rdmsr -p 0 0xc0010063 ## MSR_AMD_F17H_HW_PSTATE_STATUS rdmsr -p 0 0xc0010293
800000004ab2435c 80000000479e4258 0 0 0 0 0 0 10 0 0 4000320479
Something seems off here, do you need my CPU to be more than idle? Like do you want it at full load or is what i gave you ok?
@madoverlord40 Hello
Manually I'm computing the same as CoreFreq
Let's take the main P0 P-state register dump with value 800000004ab2435c
FID = 0x5c = 92
DID = 0x3 = 3
COF = ( FID << 1 ) / DID
COF = ( 92 * 2 ) / 3 = 61
BCLK = 4300 / 61 = 70 MHz
This never happened with previous Zen and EPYC generations!
I dont think thats correct though, i am not saying your math is wrong, just that 61. Both RM and bios show BLCK at 100. The highest multiplier should be 57.5 and BLCK of 100 for 5750mhz. Im also quite sure that a BLCK of 4300 would be a multiplier of 43 x 100, i dont think 61 for a multiplier is correct at all. So what i wonder is, if by chance, talking directly to the cpu when its idle, is getting you those values, maybe during idle(lower power state) it drops the blck to 70, but i think the biggest issue is a bios problem. Ive been reading online that 9000s were rushed by AMD new microcode patches are on the way to adjust performance. Could it be that bios is configuring the CPU at boot with the values you are pulling, and the problem is incorrect bios config. With that thought, ryzen master might be assuming a blck of 100 when it performs its calculations to show me the clocks of all cores when in fact its really doing what your math says. Which would explain why the 9000s seems to not quite perform as AMD said it would. So answer me this, what do you think of bios microcode being an issue with a rushed CPU release? If so that would make your calculations accurate, and bios updates might fix it in the near future.
So answer me this, what do you think of bios microcode being an issue with a rushed CPU release?
Honestly, you're right and I've been thinking about it ever since math blew up into my eyes. But RM, CPU-Z and HWINFO agree on the same technical answer so I can only question my calculations.
It would be interesting to try other Zen5 and other architectures like Strix Point and the next EPYC; not sure if its CPUID is confirmed and can be programmed into CoreFreq https://x.com/9550pro/status/1825870149498597760
Yeah i wouldn't beat yourself up over it, ill watch for bios updates, and on the next bios update ill post some new screenshots of your software's behavior. In the mean time, your software basically just does not support 9000s until newer microcode comes out and maybe makes your life easier lol.
Yeah i wouldn't beat yourself up over it, ill watch for bios updates, and on the next bios update ill post some new screenshots of your software's behavior. In the mean time, your software basically just does not support 9000s until newer microcode comes out and maybe makes your life easier lol.
Thank you for your help.
Meanwhile another Ryzen 7 9700X got a max frequency of 6347 MHz
leaked from Geekbench whereas its max Boost Clock is advertised to 5.5 GHz
Only a change in the AMD formula could explain it. But to confirm it or not, the Zen5 PPR datasheet is urgently expected. Especially if no firmware fix happens in the future.
@madoverlord40 Hello,
If I extend the FID
to the full first 14 bits (from 0 to 13 position included)
800000004ab2435c P0 = 0x35c P0 = 860
80000000479e4258 P1 = 0x258 P1 = 600
Then I divide value by 20
, I'm getting the following COF
COF(P0) = 860/20 = 43
COF(P1) = 600/20 = 30
Thus multiply by a factory BCLK
of 100 MHz
, frequency is as below
Freq@P0 = 43 * 100 = 4300 MHz
Freq@P1 = 30 * 100 = 3000 MHz
:-)
COF = (FID >> 1) / 10
where: FID
is PStateDef[13:0]
or PStateDef[10:0]
@madoverlord40 Can you pull and try latest commit of branch develop
?
I will need this output: corefreq-cli -s -n -z -n -M
and two screenshots of the UI in Absolute frequency
(idle and full stressed cases):
corefreq-cli -Oa -t frequency
Thank you
@madoverlord40 Can you pull and try latest commit of branch
develop
?I will need this output:
corefreq-cli -s -n -z -n -M
and two screenshots of the UI inAbsolute frequency
(idle and full stressed cases):corefreq-cli -Oa -t frequency
Thank you
Sure thing. As a software developer myself(game developer/UE5/C++) i enjoy working with other developers on projects like this. Ill go pull latest and compile, will let you know.
Processor [AMD Ryzen 9 9950X 16-Core Processor]
|- Architecture [Zen5/Granite Ridge]
|- Vendor ID [AuthenticAMD]
|- Microcode [0x0b40401a]
|- Signature [ BF_44]
|- Stepping [ 0]
|- Online CPU [ 32/ 32]
|- Base Clock [ 99.983]
|- Frequency (MHz) Ratio
Min 2999.50 < 30 >
Max 4299.28 < 43 >
|- Factory [100.000]
4300 [ 43 ]
|- Performance
TGT 4299.28 < 43 >
|- CPPC
Min 3699.38 < 37 >
Max 399.93 < 4 >
TGT AUTO < 0 >
|- Boost [ UNLOCK]
XFR 5699.05 [ 57 ]
CPB 5699.05 [ 57 ]
|- P-State
P1 2999.50 < 30 >
|- Uncore [ LOCK]
CLK 1499.75 [ 15 ]
MEM 2999.50 [ 30 ]
Instruction Set Extensions
|- 3DNow!/Ext [N/N] ADX [Y] AES [Y] AVX/AVX2 [Y/Y]
|- AVX512-F [Y] AVX512-DQ [Y] AVX512-IFMA [Y] AVX512-PF [N]
|- AVX512-ER [N] AVX512-CD [Y] AVX512-BW [Y] AVX512-VL [Y]
|- AVX512-VBMI [Y] AVX512-VBMI2 [Y] AVX512-VNNI [Y] AVX512-ALG [Y]
|- AVX512-VPOP [Y] AVX512-VNNIW [N] AVX512-FMAPS [N] AVX512-VP2I [Y]
|- AVX512-BF16 [Y] AVX-VNNI-VEX [Y] AVX-FP128 [N] AVX-FP256 [N]
|- BMI1/BMI2 [Y/Y] CLWB [Y] CLFLUSH [Y] CLFLUSH-OPT [Y]
|- CLAC-STAC [Y] CMOV [Y] CMPXCHG8B [Y] CMPXCHG16B [Y]
|- F16C [Y] FPU [Y] FXSR [Y] LAHF-SAHF [Y]
|- MMX/Ext [Y/Y] MON/MWAITX [Y/Y] MOVBE [Y] PCLMULQDQ [Y]
|- POPCNT [Y] RDRAND [Y] RDSEED [Y] RDTSCP [Y]
|- SEP [Y] SHA [Y] SSE [Y] SSE2 [Y]
|- SSE3 [Y] SSSE3 [Y] SSE4.1/4A [Y/Y] SSE4.2 [Y]
|- SERIALIZE [N] SYSCALL [Y] RDPID [Y] UMIP [Y]
|- VAES [Y] VPCLMULQDQ [Y] PREFETCH/W [Y] LZCNT [Y]
Features
|- 1 GB Pages Support 1GB-PAGES [Capable]
|- 100 MHz multiplier Control 100MHzSteps [Missing]
|- Advanced Configuration & Power Interface ACPI [Capable]
|- Advanced Programmable Interrupt Controller APIC [Capable]
|- Advanced Virtual Interrupt Controller AVIC [Missing]
|- APIC Timer Invariance ARAT [Capable]
|- LOCK prefix to read CR8 AltMov [Capable]
|- Clear Zero Instruction CLZERO [Capable]
|- Core Multi-Processing CMP Legacy [Capable]
|- L1 Data Cache Context ID CNXT-ID [Missing]
|- Collaborative Processor Performance Control CPPC [Capable]
|- Direct Cache Access DCA [Missing]
|- Debugging Extension DE [Capable]
|- Debug Store & Precise Event Based Sampling DS, PEBS [Missing]
|- CPL Qualified Debug Store DS-CPL [Missing]
|- 64-Bit Debug Store DTES64 [Missing]
|- Fast Short REP MOVSB FSRM [Capable]
|- Fast-String Operation ERMS [Capable]
|- Fused Multiply Add FMA4 [Missing]
|- Fused Multiply Add FMA [Capable]
|- Hardware Lock Elision HLE [Missing]
|- Hyper-Threading Technology HTT [Capable]
|- Hardware P-state control HwP [Capable]
|- Instruction Based Sampling IBS [Capable]
|- Instruction INVLPGB INVLPGB [Missing]
|- Instruction INVPCID INVPCID [Capable]
|- Long Mode 64 bits IA64 | LM [Capable]
|- LightWeight Profiling LWP [Missing]
|- Memory Bandwidth Enforcement MBE [Capable]
|- Machine-Check Architecture MCA [Capable]
|- Instruction MCOMMIT MCOMMIT [Missing]
|- Model Specific Registers MSR [Capable]
|- Memory Type Range Registers MTRR [Capable]
|- No-Execute Page Protection NX [Capable]
|- OS-Enabled Ext. State Management OSXSAVE [Capable]
|- OS Visible Work-around OSVW [Capable]
|- Physical Address Extension PAE [Capable]
|- Page Attribute Table PAT [Capable]
|- Pending Break Enable PBE [Missing]
|- Process Context Identifiers PCID [Missing]
|- Perfmon and Debug Capability PDCM [Missing]
|- Page Global Enable PGE [Capable]
|- Page Size Extension PSE [Capable]
|- 36-bit Page Size Extension PSE36 [Capable]
|- Processor Serial Number PSN [Missing]
|- Resource Director Technology/PQE RDT-A [Capable]
|- Resource Director Technology/PQM RDT-M [Capable]
|- Read Processor Register at User level RDPRU [Capable]
|- Restricted Transactional Memory RTM [Missing]
|- Safer Mode Extensions SMX [Missing]
|- Self-Snoop SS [Missing]
|- Supervisor-Mode Access Prevention SMAP [Capable]
|- Supervisor-Mode Execution Prevention SMEP [Capable]
|- Trailing Bit Manipulation TBM [Missing]
|- Translation Cache Extension TCE [Capable]
|- Time Stamp Counter TSC [Invariant]
|- Time Stamp Counter Deadline TSC-DEADLINE [Missing]
|- TSX Force Abort MSR Register TSX-ABORT [Missing]
|- TSX Suspend Load Address Tracking TSX-LDTRK [Missing]
|- User-Mode Instruction Prevention UMIP [Capable]
|- Virtual Mode Extension VME [Capable]
|- Virtual Machine Extensions VMX [Missing]
|- Write Back & Do Not Invalidate Cache WBNOINVD [Capable]
|- Extended xAPIC Support x2APIC [ xAPIC]
|- AVIC controller for x2APIC x2AVIC [Missing]
|- XSAVE/XSTOR States XSAVE [Capable]
|- xTPR Update Control xTPR [Missing]
|- Extended Operation Support XOP [Missing]
Mitigation mechanisms
|- Indirect Branch Restricted Speculation IBRS [Capable]
|- IBRS Always-On preferred by processor [ Unable]
|- IBRS preferred over software solution [Capable]
|- IBRS provides same speculation limits [Capable]
|- Indirect Branch Prediction Barrier IBPB [Capable]
|- Single Thread Indirect Branch Predictor STIBP [ Enable]
|- Speculative Store Bypass Disable SSBD [Capable]
|- SSBD use VIRT_SPEC_CTRL register [ Unable]
|- SSBD not needed on this processor [ Unable]
|- No Branch Type Confusion BTC_NO [Capable]
|- BTC on Non-Branch instruction BTC-NOBR [Capable]
|- Limited Early Redirect Window AGENPICK [ Unable]
|- Arch - No Fast Predictive Store Forwarding PSFD [Capable]
|- Arch - Enhanced Predictive Store Forwarding EPSF [Capable]
|- Arch - Cross Processor Information Leak XPROC_LEAK [ Unable]
Security Features
|- CET Shadow Stack features CET-SS [Capable]
|- Secure Init and Jump with Attestation SKINIT [Capable]
|- Secure Encrypted Virtualization SEV [Missing]
|- SEV - Encrypted State SEV-ES [Missing]
|- SEV - Secure Nested Paging SEV-SNP [Missing]
|- Guest Mode Execute Trap GMET [Capable]
|- Supervisor Shadow Stack SSS [Capable]
|- VM Permission Levels VMPL [Missing]
|- VMPL Supervisor Shadow Stack VMPL-SSS [Missing]
|- Secure Memory Encryption SME [Capable]
|- Transparent SME TSME [ Enable]
|- Secure Multi-Key Memory Encryption SME-MK [Missing]
|- DRAM Data Scrambling Scrambler [ Enable]
Technologies
|- Instruction Cache Unit
|- L1 IP Prefetcher L1 HW IP < ON>
|- Data Cache Unit
|- L1 Prefetcher L1 HW < ON>
|- Cache Prefetchers
|- L2 Prefetcher L2 HW < ON>
|- L1 Stride Prefetcher L1 Stride < ON>
|- L1 Region Prefetcher L1 Region < ON>
|- L1 Burst Prefetch Mode L1 Burst < ON>
|- L2 Stream HW Prefetcher L2 Stream < ON>
|- L2 Up/Down Prefetcher L2 Up/Down < ON>
|- System Management Mode SMM-Lock [ ON]
|- Simultaneous Multithreading SMT [ ON]
|- PowerNow! CnQ [ ON]
|- Core C-States CCx [ ON]
|- Core Performance Boost CPB < ON>
|- Watchdog Timer WDT < ON>
|- Virtualization SVM [ ON]
|- I/O MMU AMD-V [ ON]
|- Version [ 0.1]
|- Hypervisor [OFF]
|- Vendor ID [ N/A]
Performance Monitoring
|- Version PM [ 2]
|- Counters: General Fixed
| { 6, 4, 16 } x 48 bits 3 x 64 bits
|- Enhanced Halt State C1E <OFF>
|- C2 UnDemotion C2U < ON>
|- C3 UnDemotion C3U < ON>
|- Core C6 State CC6 < ON>
|- Package C6 State PC6 < ON>
|- Legacy Frequency ID control FID [OFF]
|- Legacy Voltage ID control VID [OFF]
|- P-State Hardware Coordination Feedback MPERF/APERF [ ON]
|- Core C-States
|- C-States Base Address BAR [ 0x413 ]
|- ACPI Processor C-States _CST [ 3]
|- MONITOR/MWAIT
|- State index: #0 #1 #2 #3 #4 #5 #6 #7
|- Sub C-State: 1 2 0 0 0 0 0 0
|- Monitor-Mwait Extensions EMX [Capable]
|- Interrupt Break-Event IBE [Capable]
|- Core Cycles [Capable]
|- Instructions Retired [Capable]
|- Reference Cycles [Capable]
|- Last Level Cache References [Capable]
|- Global Time Stamp Counter [Missing]
|- Data Fabric Performance Counter [Capable]
|- Core Performance Counter [Capable]
|- Processor Performance Control _PCT [ Enable]
|- Performance Supported States _PSS [ 2]
|- Performance Present Capabilities _PPC [ 0]
|- Continuous Performance Control _CPC [Missing]
Power, Current & Thermal
|- Temperature Offset:Junction TjMax [ 49: 95 C]
|- CPPC Energy Preference EPP < 0>
|- Digital Thermal Sensor DTS [Capable]
|- Power Limit Notification PLN [Missing]
|- Package Thermal Management PTM [Missing]
|- Thermal Monitor 1 TTP [ Enable]
|- Thermal Monitor 2 HTC [ Enable]
|- Thermal Design Power TDP [Missing]
|- Minimum Power Min [Missing]
|- Maximum Power Max [Missing]
|- Thermal Design Power Package [Disable]
|- Power Limit PL1 [ 0 W]
|- Time Window TW1 [ 0 ns]
|- Power Limit PL2 [ 0 W]
|- Time Window TW2 [ 0 ns]
|- Thermal Design Power Core [Disable]
|- Power Limit PL1 [ 0 W]
|- Time Window TW1 [ 0 ns]
|- Thermal Design Power Uncore [Disable]
|- Power Limit PL1 [ 0 W]
|- Time Window TW1 [ 0 ns]
|- Thermal Design Power DRAM [Disable]
|- Power Limit PL1 [ 0 W]
|- Time Window TW1 [ 0 ns]
|- Thermal Design Power Platform [Disable]
|- Power Limit PL1 [ 0 W]
|- Time Window TW1 [ 0 ns]
|- Power Limit PL2 [ 0 W]
|- Time Window TW2 [ 0 ns]
|- Package Power Tracking PPT [Missing]
|- Electrical Design Current EDC [Missing]
|- Thermal Design Current TDC [Missing]
|- Core Thermal Point
|- Package Thermal Point
|- Thermal Monitor Trip Limit [ 115 C]
|- HTC Temperature Limit Limit [ 127 C]
|- HTC Temperature Hysteresis Threshold [ 2 C]
|- Units
|- Power watt [ Missing]
|- Energy joule [ 0.000015259]
|- Window second [ 0.000976562]
|- Collaborative Processor Performance Control CPPC < ON>
|- Capabilities Lowest Efficient Guaranteed Highest
|- CPU #0 399.93 ( 4) 1899.68 ( 19) 2799.53 ( 28) 5199.13 ( 52)
|- CPU #1 399.94 ( 4) 1899.70 ( 19) 2799.56 ( 28) 5299.17 ( 53)
|- CPU #2 399.94 ( 4) 1899.70 ( 19) 2799.56 ( 28) 4899.24 ( 49)
|- CPU #3 399.93 ( 4) 1899.68 ( 19) 2799.54 ( 28) 4799.20 ( 48)
|- CPU #4 399.93 ( 4) 1899.69 ( 19) 2799.54 ( 28) 5299.12 ( 53)
|- CPU #5 399.94 ( 4) 1899.70 ( 19) 2799.56 ( 28) 5099.20 ( 51)
|- CPU #6 399.93 ( 4) 1899.69 ( 19) 2799.54 ( 28) 4599.25 ( 46)
|- CPU #7 399.93 ( 4) 1899.69 ( 19) 2799.54 ( 28) 4699.23 ( 47)
|- CPU #8 399.93 ( 4) 1899.69 ( 19) 2799.54 ( 28) 3999.35 ( 40)
|- CPU #9 399.93 ( 4) 1899.69 ( 19) 2799.54 ( 28) 4499.26 ( 45)
|- CPU #10 399.93 ( 4) 1899.69 ( 19) 2799.54 ( 28) 4199.32 ( 42)
|- CPU #11 399.93 ( 4) 1899.69 ( 19) 2799.54 ( 28) 4399.28 ( 44)
|- CPU #12 399.93 ( 4) 1899.69 ( 19) 2799.54 ( 28) 3899.36 ( 39)
|- CPU #13 399.93 ( 4) 1899.69 ( 19) 2799.54 ( 28) 4299.30 ( 43)
|- CPU #14 399.93 ( 4) 1899.69 ( 19) 2799.54 ( 28) 3799.38 ( 38)
|- CPU #15 399.93 ( 4) 1899.69 ( 19) 2799.54 ( 28) 3699.40 ( 37)
|- CPU #16 399.93 ( 4) 1899.69 ( 19) 2799.54 ( 28) 5199.14 ( 52)
|- CPU #17 399.94 ( 4) 1899.69 ( 19) 2799.55 ( 28) 5299.14 ( 53)
|- CPU #18 399.93 ( 4) 1899.69 ( 19) 2799.54 ( 28) 4899.20 ( 49)
|- CPU #19 399.93 ( 4) 1899.69 ( 19) 2799.54 ( 28) 4799.22 ( 48)
|- CPU #20 399.93 ( 4) 1899.69 ( 19) 2799.54 ( 28) 5299.13 ( 53)
|- CPU #21 399.93 ( 4) 1899.69 ( 19) 2799.54 ( 28) 5099.17 ( 51)
|- CPU #22 399.93 ( 4) 1899.69 ( 19) 2799.54 ( 28) 4599.25 ( 46)
|- CPU #23 399.93 ( 4) 1899.69 ( 19) 2799.54 ( 28) 4699.23 ( 47)
|- CPU #24 399.93 ( 4) 1899.69 ( 19) 2799.54 ( 28) 3999.35 ( 40)
|- CPU #25 399.93 ( 4) 1899.69 ( 19) 2799.54 ( 28) 4499.27 ( 45)
|- CPU #26 399.93 ( 4) 1899.69 ( 19) 2799.54 ( 28) 4199.31 ( 42)
|- CPU #27 399.93 ( 4) 1899.69 ( 19) 2799.54 ( 28) 4399.28 ( 44)
|- CPU #28 399.93 ( 4) 1899.69 ( 19) 2799.54 ( 28) 3899.36 ( 39)
|- CPU #29 399.93 ( 4) 1899.69 ( 19) 2799.54 ( 28) 4299.30 ( 43)
|- CPU #30 399.93 ( 4) 1899.69 ( 19) 2799.54 ( 28) 3799.38 ( 38)
|- CPU #31 399.93 ( 4) 1899.69 ( 19) 2799.54 ( 28) 3699.40 ( 37)
Zen UMC [14E0]
Controller #0 Dual Channel
Bus Rate 3000 MHz Bus Speed 3000 MHz DDR5 Speed 6000 MT/s
Cha CL RCDr RCDw RP RAS RC RRDs RRDl FAW WTRs WTRl WR clRR clWW
#0 32 39 39 39 102 146 8 15 32 8 30 90 8 23
#1 32 39 39 39 102 146 8 15 32 8 30 90 8 23
CWL RTP RdWr WrRd scWW sdWW ddWW scRR sdRR ddRR drRR drWW drWR drRRD
#0 30 23 21 8 1 10 10 1 11 11 0 0 0 0
#1 30 23 22 8 1 10 10 1 11 11 0 0 0 0
REFI RFC1 RFC2 RFCsb RCPB RPPB BGS:Alt Ban Page CKE CMD GDM ECC
#0 11677 312 192 390 0 0 ON OFF R0W0 0 0 1T ON 0
#1 11677 312 192 390 0 0 ON OFF R0W0 0 0 1T ON 0
MRD:PDA MOD:PDA WRMPR STAG PDM RDDATA WRD WRL RDL XS XP CPDED
#0 42 32 42 32 24 7 0:F:1 20 6 18 36 914 23 15
#1 42 32 42 32 24 7 0:F:1 20 6 18 36 914 23 15
DIMM Geometry for channel #0
Slot Bank Rank Rows Columns Memory Size (MB)
#0
#1 32 1 65536 1024 16384 F5-6400J3239G32G
DIMM Geometry for channel #1
Slot Bank Rank Rows Columns Memory Size (MB)
#0
#1 32 1 65536 1024 16384 F5-6400J3239G32G
corefreq-cli -Oa -t frequency
Its usually 3600 all the way down idle, when it hit the screenshot button one core boosted lol
FYI i think its show 5825 because i have a +100 boost clock in bios, but it seems to want to not give the entire 100 over.
FYI i think its show 5825 because i have a +100 boost clock in bios, but it seems to want to not give the entire 100 over.
Thanks for your returns
It works much better. CPU and Memory frequencies are getting accurate 😊
On screenshots I'm noticing that Vcore is going in the opposite direction of the system load.
That's something I had fixed for 7950X in issue #496 without knowing the root cause.
But previously on 9950X we used to have a coherent voltage Core.
So what did happen in your 9950X configuration? BIOS? Microcode? Hardware change?
the only thing i have done is lowered my curve optimizer value down for stability. This may have changed the voltages curves for you.
the only thing i have done is lowered my curve optimizer value down for stability. This may have changed the voltages curves for you.
If you cancel this change, are you reading back the good Vcore?
Hmm i did notice that one of my curve optimizer settings recently had my system hard locked, and i had to clear cmos to get it to post again. At the time i was just trying to find the limits it would boot at. After cmos clear i put PBO back on and curve optimizer -20. Maybe the recent clear cmos changed something? Should i just load setup defaults and get you new screen shots with no bios settings changed?
FYI i think its show 5825 because i have a +100 boost clock in bios, but it seems to want to not give the entire 100 over.
You mean you have set a frequency ratio of 58
in BIOS but that value is not shown in the corefreq-cli -s
output and not in the UI [Processor]
window ?
Observe that you can increase the UI ruler with a build parameter.
For example to display up to a maximum of 60
make -j UI_RULER_MAXIMUM=60
You mean you have set a frequency ratio of 58 in BIOS but that value is not shown in the corefreq-cli -s output and not in the UI [Processor] window ?
No i mean i added 100 to boost clock override in PBO, which means its max boost clock gets bumped up by 100 when PBO fires. I didnt change any multipliers. Would it be helpful if in my bios, i reset everything to setup defaults incase my bios custom settings are throwing off your calculations?
Here are some custom bios settings i have for PBO if it helps you:
Here i turned up the LLC to try and balance high curve optimizer:
Maybe these settings are making it harder for you to get VCore values you expect?
Yeah i see what you mean now, this next one is your software with no parameters, showing relative, check out the V at the bottom:
Maybe these settings are making it harder for you to get VCore values you expect?
Yes I believe they change the relation between the VID
and the computed voltage.
AMD has specified a voltage formula for Zen and that's the only one I know. This formula does not take into account any optimizer.
Please let me know if setting back the BIOS to the voltage stock options is showing the right Vcore in CoreFreq ?
Maybe these settings are making it harder for you to get VCore values you expect?
Yes I believe they change the relation between the
VID
and the computed voltage.AMD has specified a voltage formula for Zen and that's the only one I know. This formula does not take into account any optimizer.
Please let me know if setting back the BIOS to the voltage stock options is showing the right Vcore in CoreFreq ?
Ok i will go set LLC to auto, and set PBO to just enabled, Ill also remove the boost clock +100 and set the curve optimizer to off and i will get you some new screenshots.
Yeah i see what you mean now, this next one is your software with no parameters, showing relative, check out the V at the bottom:
Indeed, as you can see a Vcore of 1.39
is not feasible when Processor is idle.
Thus my request to test with stock options.
Yeah i see what you mean now, this next one is your software with no parameters, showing relative, check out the V at the bottom:
Indeed, as you can see a Vcore of
1.39
is not feasible when Processor is idle. Thus my request to test with stock options.
Ok all i see is that the VCore is higher now, here is LLC auto, curve optimizer off, PBO enabled not max: While screen capturing i saw VCore spike to 1.45 without frequency load to match. Also it seems to bounce back and fourth from 0.88 to 1.43ish without frequency load to match.
This is using your software to turn on turbo for all cores, nearly normal bios, no options passed to your software:
When i stop it, the V goes down to 0.88 for a short time the spikes back up to 1.43ish and stays there with no load.
@madoverlord40
0.88
when idle, despite LLC, confirms a good reading. Thank you for this test.
Unfortunately I'm not aware of Registers to compensate voltage when BIOS tricks are enabled.
@madoverlord40 Voltage subject transfered to issue #506 as Optimizer concerns more architectures
@madoverlord40 Voltage subject transfered to issue #506 as Optimizer concerns more architectures
Ok cool.
"Why can't I compile version 1.98.3 correctly on the rocky9.2 system with a 9950X CPU?"
"Should I switch to a lower version?"
1.96.5version is ok but no temperature
"Why can't I compile version 1.98.3 correctly on the rocky9.2 system with a 9950X CPU?"
Edit file at these lines:
Replace the build condition with:
#if LINUX_VERSION_CODE >= KERNEL_VERSION(6, 4, 0) \
|| (defined(RHEL_MAJOR) && (RHEL_MAJOR >= 9) \
&& (RHEL_MINOR > 2) && (RHEL_MINOR < 99))
Save file and rebuild all
make -j clean
make -j
i try
When single thread boosting occurs on 9950x, it goes all the way to 61 and seems to be off the chart while reporting 5676 in the upper left corner. I assume you have not yet added support for this cpu as it just released. I will keep using CoreFreq anyway hoping you will release an update soon to correct this.
Thanks.