• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.
  • The forums have been upgraded with support for dark mode. By default it will follow the setting on your system/browser. You may override it by scrolling to the end of the page and clicking the gears icon.

Is something wrong with my CPU?

  • Thread starter Thread starter Deleted member 212040
  • Start date Start date
There is a bug with CPU-Z on Win 11, it runs single thread benchmark on most efficient core, and not the highest performance core. The only way I found to make it run on correct core for a benchmark, is RDP to your computer from another machine and then run CPU-Z on Win11 machine, and it will run on the highest performance core for single thread portion of the benchmark.
 
What platform? This is an old thread and we were discussing Zen 3 here. Intel may be different, maybe not.

On Zen 2 and Zen 3, CPU-Z has never run its ST test on the best rated CPPC core, or on anything other than Core 0 honestly. An inherent limitation, whether by its own hand or Windows'. It's not exclusive to Windows 11. Setting process affinity or process lasso doesn't always do much either.

If you want to measure ST performance, Cinebench R23 does a better and more consistent job. CPU-Z is way too short and unreliable, varies way too much from run to run even without settings change. Remote Desktop back into your machine all just to run one of the less meaningful benchmarks?
 
Last edited:
I am running 5900x Zen3. On Win 10 I recall it used to run on my highest performance core. My score due to thread change went from 682 to 672. And the only difference I noticed is which thread it runs on.
The RDP hack on win 11 works and forces it to go to highest performance core, for some reason.
 
I am running 5900x Zen3. On Win 10 I recall it used to run on my highest performance core. My score due to thread change went from 682 to 672. And the only difference I noticed is which thread it runs on.
The RDP hack on win 11 works and forces it to go to highest performance core, for some reason.

ST drop from going to 11 is not unheard of on this platform. Same with some other synthetics as well on 11. I dropped from 687 ish to 675, it just kind of hits a wall on 11 even though clocks are the same.

If there is no appreciable difference in CB R23 ST, this isn't worth the time to fuss over. Sounds like performance is still fine.
 
And that is basically my point behind posting it as well, your cpu is performing fine, it is just CPU-Z being weird. So don’t stress.

I would agree with your ST drop point, if I didn’t observe what I am reporting here.
My cpu still scores dead on 682 when it runs on highest performance core.

just an interesting observation. No fuss, just providing input for others who may experience the same. Let’s all relax.
 
I would agree with that, if I didn’t observe what I am reporting here.
My cpu still scores dead on 682 when it runs on highest performance core.

Let’s all relax.

What AGESA? afaik all that behaviour holds true up to 1.2.0.6, not sure about any new beta BIOSes that may have surfaced.

Without any specifics as to the hardware and firmware you use as well, as well as the HWInfo clocks you're seeing, your CPPC core distribution and what exactly the cores are in question, kinda hard to do anything.
 
Correct. These tests were performed on 1.2.0.3 GA, not beta.

As for CPPC and core distribution, that does not change with agesa, that is a hardware spec based on the chiplet bin.

The point is simple, so no need to throw up a word salad to try and sound smart. CPU-Z on win 11 does not run on the fastest thread based on CPPC. Stop trying to make an observation report, which you agree with, into a thing.
 
Correct. These tests were performed on 1.2.0.3 GA, not beta.

As for CPPC and core distribution, that does not change with agesa, that is a hardware spec based on the chiplet bin.

The point is simple, so no need to throw up a word salad to try and sound smart. CPU-Z on win 11 does not run on the fastest thread based on CPPC. Stop trying to make an observation report, which you agree with, into a thing.

Are you for real? Telling others to relax? If you don't include any details aside from "it doesn't run on the fastest core"?

I never said CPU-Z follows core ranking. It never has, that's the point, and what sets it apart from just about every other bench. 1.2.0.3 is a ways back now, pre-11 release, and AGESA can most definitely affect thread scheduling in ST benches.
 
Lol!!!! You are a gem. Then why are you trying to be obtuse? aren’t we in an agreement then?

Additionally, I am now on 1.2.0.6b and it is still the same behaviour.
 
Back
Top