CPU-Z changes benchmark code (Ryzen suffers)

[Cracklingice]

Quote

CPU-Z 1.79 : new benchmark, new scores.

 May 01st, 2017

The latest version of CPU-Z introduced a new version of the benchmark (2017.01), that succeeds to the previous version introduced in 2015. We've received a lot of emails concerning the new scores, and we'll try to answer to the most frequently asked questions :

Why are the scores much lower than the previous version, and can they be compared ?

 

Why do the Ryzen performance decrease in comparison to the Intel processors with the new benchmark ?

What algorithm does the benchmark use, and what instruction set is used ?

When will the benchmark pages be updated with the new benchmark results ?

CPU-Z has changed the code behind it's bench marking and scoring.  The new scores reflect lower for all CPUs.  The scores from version 1.79 cannot be compared to older versions as a result.  The change in the benchmark code also appears the have a larger negative impact on Ryzen because Ryzen was significantly faster at processing a specific segment of the code in the benchmark than any previous architecture.  CPUID says that they could not find any other benchmark or real world performance indicators that reflected this discrepancy exclusive to Ryzen and that their new code more accurately represents real world performance.

 

http://www.cpuid.com/news/51-cpu-z-1-79-new-benchmark-new-scores.html

 

Personally, as long as the new benchmark version stands more in line with other benchmarks, productivity and gaming thread to thread than the previous version, I support the change.  I also think this could make an interesting topic to investigate in a quick video for LTT or possibly a tech quickie.

[Guest Darth Revan]

So basically the old results with Ryzen were false.

[Cracklingice]

2 minutes ago, Darth Revan said:

So basically the old results with Ryzen were false.

Not really - they were accurate for the workload, but the workload may have been flawed and not real world indicative.

[Citadelen]

22 hours ago, Darth Revan said:

So basically the old results with Ryzen were false.

As always you're the first to jump to AMD's throat with a knife...

No, Zen was genuinely faster than Skylake in those scenarios, it has simply been changed to be more inline with with how the architecture performs in more real world situations now.

[raphidy]

Ryzen got nerf because it's too OP.

[porina]

Does anyone actually use the CPU-Z bench? I don't recall them relating it to any particular real world scenario so I've not really paid attention to it.

[mr moose]

It's basically exactly the same as applying an algorithm to temperature data to get an accurate real world average. Or the algorithms used to ensure sensor variance doesn't effect comparisons.   In any case, CPU-z is hardly the be all and end all of benchmarks and you should never rely on only one.

[lawrencep93]

I really want to see some decent premier pro work load orientate bench marks for the CPU's as I think the 1700/1800x is more so going to be used as a multi core workstation alternative to over priced Intel 2011-3 x99 platform

[DocSwag]

15 minutes ago, raphidy said:

Ryzen got nerf because it's too OP.

No, it got nerfed because the part that Ryzen vastly outperformed Intel CPUs in wasn't representative of real world scenarios, and so CPU z changed it to more reflect real world scenarios.

 

I always thought it was weird that Ryzen was so much faster in CPU z than Intel CPUs, this explains it.

[Cracklingice]

Just now, DocSwag said:

No, it got nerfed because the part that Ryzen vastly outperformed Intel CPUs in wasn't representative of real world scenarios, and so CPU z changed it to more reflect real world scenarios.

 

I always thought it was weird that Ryzen was so much faster in CPU z than Intel CPUs, this explains it.

I think it may have been partly due to their 'neural net prediction', but that's entirely a theory of my own and not based on any factual evidence.

[DocSwag]

2 minutes ago, Cracklingice said:

I think it may have been partly due to their 'neural net prediction', but that's entirely a theory of my own and not based on any factual evidence.

If dat was the case tho it would've been significantly better the second run than the first and wouldn't have been winning by much on the first run.

[Guest Darth Revan]

Looks like I pissed off AMD fanboys.

[Cracklingice]

6 minutes ago, yian88 said:

No it means the old results were right and now CPU-Z guys got payed by intel to change the code so that ryzen looks bad.

Get a grip on reality, AMD isnt doing bad just because they arent able to make best cpus, these retarded practices have been going on for ages and intel does enough damage on AMD using them enough to keep market monopoly and sell us more crappy 4 cores with unwanted IGP, 4 cores first released by intel in 2007, 10 years later they still sell us 4 cores. Now look at phones first quad-core released 2011 or so, 6 years later most midrange phones have 8 cores and they are actually being used/usefull(ofcourse very different loads and arhitecture than Desktop but still), and higher end phones have 10-12 cores hybrid cpu combinations, and the integrated IGP in phones is at least as good or much better than crappy intel IGP, yet you pay for i7 with IGP what you pay for a WHOLE phone with 8-12 cores better graphics and a ton of other hardware, quality led touchscreen, quality camera(for a phone)  and a ton more features.

 

With this trend the desktop experience will just become weaker and more expensive over time, if they could actually make a phone that doesnt overheat and has big battery with new technology, desktop PC's will die overnight.

You're joking right?  Have you actually taken the time to compare the scaling between results for Ryzen 4c/8t and Intel 4c/8t in both versions of CPU-Z and other relevant metrics?

[Cracklingice]

1 minute ago, Darth Revan said:

Looks like I pissed off AMD fanboys.

Dunno about that.  I'd still by a R5 1600 over any i5 any day of the week because of the extra cores.  Though those cores are still slower than Intels on a core/thread basis.  Ok - so I guess I'm not really an AMD fanboy - just a hardware enthusiast.

[Guest Darth Revan]

4 minutes ago, Cracklingice said:

Dunno about that.  I'd still by a R5 1600 over any i5 any day of the week because of the extra cores.  Though those cores are still slower than Intels on a core/thread basis.  Ok - so I guess I'm not really an AMD fanboy - just a hardware enthusiast.

The R5 over the i5, yes. With that one I agree.

Intel's current version of i5's compared to what AMD's line-up has to offer the R5, makes no sense.

[Cracklingice]

8 minutes ago, Darth Revan said:

The R5 over the i5, yes. With that one I agree.

Intel's current version of i5's compared to what AMD's line-up has to offer the R5, makes no sense.

Yeah, though the i7 7700k over even the 1600x.  Ryzen just can't keep up single core enough IMO if you can budget i7 (or r7 unless you're productivity > gaming) price on the CPU.

 

[0ld_Chicken]

Thread moved to CPUs, Motherboards, and Memory

 

 

[Cracklingice]

10 minutes ago, 0ld_Chicken said:

Thread moved to CPUs, Motherboards, and Memory

 

 

Sorry, I was unaware that technology news now belongs outside of the news forum.

[0ld_Chicken]

13 minutes ago, Cracklingice said:

Sorry, I was unaware that technology news now belongs outside of the news forum.

Nope, just some basic posting guidelines is all.

Quote

• Your thread must include some original input to tell the reader why it is relevant to them, and what your personal opinion on the topic is.
• Your thread must include a link to at least one reputable source. Most of the time, this should be a respected news site.
• Your thread should also include quotes from the cited source(s). While you shouldn't just copy the entire article, your quote should give the reader a summary of the article in a way that gives the key details, but also leaves room for them to read the full article on the linked website. Please use quote tags (the speech bubble at the top of the editor, under the  :)) to show that you have copied this content from another site.
• The title of your thread must be relevant to the topic and should give a reader a good idea of the contents of the thread. Copying the title of the source is permitted but absolutely not required.
• If your article is about a product or some form of media, images are always appreciated, although they are not required.

 

[done12many2]

1 hour ago, yian88 said:

No it means the old results were right

 

How did you reach this determination? 

 

1 hour ago, yian88 said:

and now CPU-Z guys got payed by intel to change the code so that ryzen looks bad.

Get a grip on reality, AMD isnt doing bad just because they arent able to make best cpus, these retarded practices have been going on for ages and intel does enough damage on AMD using them enough to keep market monopoly and sell us more crappy 4 cores with unwanted IGP, 4 cores first released by intel in 2007, 10 years later they still sell us 4 cores. Now look at phones first quad-core released 2011 or so, 6 years later most midrange phones have 8 cores and they are actually being used/usefull(ofcourse very different loads and arhitecture than Desktop but still), and higher end phones have 10-12 cores hybrid cpu combinations, and the integrated IGP in phones is at least as good or much better than crappy intel IGP, yet you pay for i7 with IGP what you pay for a WHOLE phone with 8-12 cores better graphics and a ton of other hardware, quality led touchscreen, quality camera(for a phone)  and a ton more features.

 

With this trend the desktop experience will just become weaker and more expensive over time, if they could actually make a phone that doesnt overheat and has big battery with new technology, desktop PC's will die overnight.

 

I'm pretty sure you just pulled the Intel pay off theory completely out of your back side.   Additionally, multi core smartphones and desktop parts have nothing in common.  

[Cracklingice]

1 hour ago, 0ld_Chicken said:

Nope, just some basic posting guidelines is all.

 

Didn't include a quote because I wasn't sure what to put that would not negate any reason to click the article.  No idea why it took this long to realize to include the questions but not the answers!  -facepalm-

BTW, I misread earlier and thought it said that tech news belonged in the tech section, not that only tech news should be posted.

[0ld_Chicken]

29 minutes ago, Cracklingice said:

Didn't include a quote because I wasn't sure what to put that would not negate any reason to click the article.  No idea why it took this long to realize to include the questions but not the answers!  -facepalm-

BTW, I misread earlier and thought it said that tech news belonged in the tech section, not that only tech news should be posted.

Got it moved back to TN&R

[Misanthrope]

3 hours ago, Darth Revan said:

So basically the old results with Ryzen were false.

Or Intel and their globalist illuminati vampires successfully got to the CPU-Z people.

 

Joking aside this is kind of why I hate synthetics. 

[79wjd]

2 hours ago, done12many2 said:

I'm pretty sure you just pulled the Intel pay off theory completely out of your back side.

I'm not saying this is the case here, but Intel does have a history of doing that. 

[ApolloFury]

5 hours ago, DocSwag said:

No, it got nerfed because the part that Ryzen vastly outperformed Intel CPUs in wasn't representative of real world scenarios, and so CPU z changed it to more reflect real world scenarios.

 

I always thought it was weird that Ryzen was so much faster in CPU z than Intel CPUs, this explains it.

Joke > your head