Jump to content
HWBOT Community Forums
max1024

CPU-Z Valdation (CPU Frequency benchmark)

Recommended Posts

Once upon a time, when the processors were single-core, the CPU-Z validation process was simple. But when the number of cores began to increase, one nuance arose.

The essence of the nuance is that 1-2 active cores are left in multi-core processors and this means that the processor of this model accelerates to the X frequency. But in fact, not all Cores accelerates. Scrolling through the results of 8-12-16 and more cores processors, I often see that some overclockers disable some of the cores and others do not. But when sorting the result, this aspect is not taken into account. And they get a strange situation for many that the processor overclocked "entirely" to a certain frequency, but in fact 1/6 or 1/10 in general. Often, such results fall into the news feeds of sites and reading users naively believe that all the cores were able to overclock to the stated frequency, but when they click on the result link, they will be disappointed.

I want to suggest changing the rules of this test, it would be more logical to see ALL processor cores at the final result. It would be honest in relation to oveclockers and to the processors as well. What are your thoughts on this situation?

Share this post


Link to post
Share on other sites
18 hours ago, max1024 said:

Once upon a time, when the processors were single-core, the CPU-Z validation process was simple. But when the number of cores began to increase, one nuance arose.

The essence of the nuance is that 1-2 active cores are left in multi-core processors and this means that the processor of this model accelerates to the X frequency. But in fact, not all Cores accelerates. Scrolling through the results of 8-12-16 and more cores processors, I often see that some overclockers disable some of the cores and others do not. But when sorting the result, this aspect is not taken into account. And they get a strange situation for many that the processor overclocked "entirely" to a certain frequency, but in fact 1/6 or 1/10 in general. Often, such results fall into the news feeds of sites and reading users naively believe that all the cores were able to overclock to the stated frequency, but when they click on the result link, they will be disappointed.

I want to suggest changing the rules of this test, it would be more logical to see ALL processor cores at the final result. It would be honest in relation to oveclockers and to the processors as well. What are your thoughts on this situation?

There are many "singlethreaded" benches out there like superpi 32m which people tend to disable cores for. I don't see any reason to do away with that.

 

  • Like 1

Share this post


Link to post
Share on other sites

Yes, there are single-threaded CPU tests, but I also consider CPU-Z as an independent test. The essence is different, overclocking one or two cores does not mean overclocking the processor entirely. If the embedded CPU-Z test  was used, then the first result would be the result with all the cores.

Share this post


Link to post
Share on other sites
28 minutes ago, max1024 said:

Yes, there are single-threaded CPU tests, but I also consider CPU-Z as an independent test. The essence is different, overclocking one or two cores does not mean overclocking the processor entirely. If the embedded CPU-Z test  was used, then the first result would be the result with all the cores.

A cpuz validation doesn't even run any stress, it's literally a test of what clocks cpuz reads out so it is in fact single threaded

  • Like 1

Share this post


Link to post
Share on other sites
21 hours ago, max1024 said:

I want to suggest changing the rules of this test, it would be more logical to see ALL processor cores at the final result. It would be honest in relation to oveclockers and to the processors as well. What are your thoughts on this situation?

Although that would be perfectly valid as a new rule, it's definitely less fun.

Might be even more fun to distribute the global points by cpu architecture.

  • Like 1

Share this post


Link to post
Share on other sites

Raw CPU-Z clock (as a benchmark) is supposed to be all about 1-second stability irrespective of measures necessary. Although I understand your point about disabling cores frequently leading to misleading news articles, I would blame it on the messenger on failing to correctly interpret information, rather than a benchmark flaw or "unethical" behavior by an overclocker. Also, there is no need to create extra complications, such as splitting into multiple categories (1core vs all cores) or rewriting the rule books.

Edited by TaPaKaH
  • Thanks 1

Share this post


Link to post
Share on other sites

A job well done. Someone most likely has invested time into testing all 18 cores individually and finding the best one. This fact is not hidden from you on validation, so why complain?

Share this post


Link to post
Share on other sites
4 hours ago, yosarianilives said:

A cpuz validation doesn't even run any stress, it's literally a test of what clocks cpuz reads out so it is in fact single threaded

This is all clear, you can also set the CPU-Z priority to one core, but this does not negate the fact that ALL of the processor cores have reached the final one — the same frequency.

3 hours ago, _mat_ said:

Although that would be perfectly valid as a new rule

It is rather a suggestion for a future revision of the HWBOT

 

1 hour ago, Splave said:

So how do you feel about enabling all cores and only moving 1

From my point of view this is the wrong result. All cores must be of the same frequency.

TaPaKaH

The software of this test can be oriented to which frequencies using different cooling, the processor is extremely overclocked. It is clear to everyone that there can be no talk of total stability, but since in the database there is now observed, in my opinion, chaos due to the disconnection of the cores, then very often the average frequency of overclocking differs by 20% or more. You have to look, and how many cores have reached the final frequency, you look, but it turns out that only one or two. I think the essence of what you understand my ideas.

Share this post


Link to post
Share on other sites
1 hour ago, yosarianilives said:

You have completely misunderstood the point of this bench. 

Share your concept of essence. I shared my understanding.

Share this post


Link to post
Share on other sites
2 minutes ago, max1024 said:

Share your concept of essence. I shared my understanding.

The point of this bench is to get cpuz to read a number, has nothing to do with comparing to other cpus in any practical sense. If you want to do that look at a bench like x265

Share this post


Link to post
Share on other sites
1 hour ago, yosarianilives said:

he point of this bench is to get cpuz to read a number, has nothing to do with comparing to other cpus in any practical sense.

Thanks :) I fully understand your point, I just wanted to offer for more order and the least deviation of the results to get cpuz to read a number for all cores with the same frequency, and nothing else. I will not write more about this, I just wanted to draw attention to this aspect.

Share this post


Link to post
Share on other sites

The only people that care about CPU valids being on all cores are journalists, Roman and noobs.
which one are you?

  • Like 2
  • Haha 2

Share this post


Link to post
Share on other sites
2 hours ago, zeropluszero said:

The only people that care about CPU valids being on all cores are journalists, Roman and noobs.
which one are you?

You nailed it

  • Thanks 2

Share this post


Link to post
Share on other sites

You do have a valid point max, I too once thought that was the way it was surpassed to be but since then I learned otherwise.

However one big issue would come into play;  Most if not all of those world record 8ghz+ cpuz validations would have to be taken down from the rankings (hell might have to remove ALL cpuz validations) because for good or bad right or wrong validating on less cores has been the norm since it began. That would leave a lot of unhappy campers loosing their minds as to why their hard earned validations were taken down.

Then next a new category for CPUZ validations will have to be created based on and separated by core/thread count. Dont see that happening any time soon.

That idea could however be a suggestion for adding such limitations as part of a future completion.

  • Like 1

Share this post


Link to post
Share on other sites

cpuz core cuts.... Well you always have to validate core 0 only. So if other cores draw too much from the 12v rail, its time to shut them down, increase current and increase overclock. 

Not me. I run naked Ryzen 7 2700x at 4500mhz on 4 cores plenty fine. Not many of my personal needs extend beyond the need for increased core count even gaming. You can see at Warp9-systems.com

I like that hwbot lets us cut cores. I learnt processor does tricks!!

Edited by ShrimpBrime

Share this post


Link to post
Share on other sites

If it validates it is fine, watch out with XOC mode if SPD tab and co is required… Tab might be blank

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.


×
×
  • Create New...