AMD accuses BAPCo and Intel of cheating with Sysmark benchmarks

[CAT-THE-FIFTH]

Here it is :

http://www.xtremesystems.org/forums...y-in-Tablets&p=5150663&viewfull=1#post5150663

It was in relationship to the XPRT series of benchmarks used for mobile benchmarks(interestingly some of the links are now defunct now..),like comparing Atom towards competitors.

But later on a few websites said the XPRT series of benchmarks had Intel involvement.

However,if you search the name of the guy mentioned you will he was/is a longterm Intel employee and founding member of BAPCo.

The new Intel multimedia benchmark used by AnandTech: TouchXPRT 2013 is another brainchild
of dr Who's boss Shervin Kheradpir, General Manager of Intel's Performance Benchmarking and Analysis
Group and founding President of Bapco (via HDXPRT/Principled Technologies)

http://www.hdxprt.com/blog/2012/10/2...the-fast-lane/ AnandTech was the first to use the test
http://intel-mydreampc-1829796403.us...Whitepaper.pdf

If you google him you find he is a longterm Intel person and founding member of BAPCo.

It does not indicate he or Intel has done anything dodgy,but still it does put a cloud on whether the benchmark is viable,especially with 3/4 of the chip makers(AMD,Nvidia and VIA) leaving the said benchmark years ago.

Second Edit!!

Some stuff about him:

https://www.crunchbase.com/person/shervin-kheradpir#/entity
http://newsroom.intel.com/community/intel_newsroom/bios?n=Shervin Kheradpir&f=searchAll

Shervin KheradpirVice President, Sales and Marketing Group
General Manager, Platform Evaluation and Competitive Assessment

Kheradpir joined Intel in 1989 as a technical marketing engineer shortly after completing his education.

http://www.businesswire.com/news/home/20060925006184/en/BAPCo-Announces-EECoMark-TM----Joint-Venture

Shervin Kheradpir, President of BAPCo

Third Edit!!

Did some more digging:

https://semiaccurate.com/forums/showpost.php?p=119950&postcount=50

In that Futuremark release in 1999 it has:

John Peterson BAPCo 2200 Mission College Blvd., SC12-608, Santa Clara, CA 95052 408-988-7654 http://www.bapco.com

Another from The Register who commented on BAPCo and Intel being at the same address in 2000:

http://www.theregister.co.uk/2000/07/24/intel_and_bapco_just_good/

Not sure if they have moved from there now but Intel still is at the same address:

https://www.sec.gov/comments/4-600/4600-36.pdf

 

[Nasher]

Everyone knows Intel BS about the performance. In the real world, Skylake has no noticeable performance difference vs 5 year old i5/i7s in most applications.

 

[mmj_uk]

Everyone knows Intel BS about the performance. In the real world, Skylake has no noticeable performance difference vs 5 year old i5/i7s in most applications.

They're all a lot better than AMD in the majority of applications though, which is why AMD complaining about this benchmark is little more than mud slinging. If they have any evidence of foul play and that it impacted their sales I'm sure they'd have filed a law suit by now, rather than running to the media crying as usual.

 

[faceman123]

AMD still used Intel to show of their FuryX

 

[David Bisset]

A couple of posts in here effectively saying "Intel CPUs are faster so it's fine for them to lie about how fast they are" ... that's a pretty unhealthy attitude!

 

[mmj_uk]

A couple of posts in here effectively saying "Intel CPUs are faster so it's fine for them to lie about how fast they are" ... that's a pretty unhealthy attitude!

Innocent until proven guilty.

If Intel have indeed cheated then I'm sure AMD will be suing them for damages, as opposed to crying about it in amateurish Youtube videos that serve mainly to rile up the few customers they have left against their far more successful competitors.

 

[Martini1991]

AMD still used Intel to show of their FuryX

Because their own CPU's aren't a viable option for those GPU's, because AMD know that they will bottleneck them (Doesn't stop AMD users arguing blind in the face that their CPU isn't a bottleneck)

 

[CAT-THE-FIFTH]

Because their own CPU's aren't a viable option for those GPU's, because AMD know that they will bottleneck them (Doesn't stop AMD users arguing blind in the face that their CPU isn't a bottleneck)

Ironic how their CPUs generally were better in MT benchmarks,but their own drivers tended to be more ST heavy than Nvidia,who tended to work better on AMD in comparison.

 

[David Bisset]

Innocent until proven guilty.

If Intel have indeed cheated then I'm sure AMD will be suing them for damages, as opposed to crying about it in amateurish Youtube videos that serve mainly to rile up the few customers they have left against their far more successful competitors.

Agreed, not saying we should lynch Intel because of an unsubstantiated claim. However, two users were suggesting even if it was proved to be true it's fine as Intel are faster, which makes no sense at all.

 

[CAT-THE-FIFTH]

Agreed, not saying we should lynch Intel because of an unsubstantiated claim. However, two users were suggesting even if it was proved to be true it's fine as Intel are faster, which makes no sense at all.

The problem is at least a couple of years ago,the General Manager, Platform Evaluation and Competitive Assessment at Intel was founding president of BAPCo,another Intel guy used to be their main contact and at least at the beginning they were located in the main Intel location in the US AFAIK.

It does not mean Intel cheated or changed anything,but it does cloud it as a benchmark.

It would be like the president of Futuremark was an AMD VP and their offices were located at AMD HQ - I don't think many Nvidia users like myself would trust them,even though you could argue like Sysmark,3DMark is not really a good indication of real world performance.

 

[Jono8]

They're all a lot better than AMD in the majority of applications though, which is why AMD complaining about this benchmark is little more than mud slinging. If they have any evidence of foul play and that it impacted their sales I'm sure they'd have filed a law suit by now, rather than running to the media crying as usual.

This

AMD need to simply concentrate on releasing something competitive instead of wasting time slinging around fairly weak accusations.

 

[CAT-THE-FIFTH]

This

AMD need to simply concentrate on releasing something competitive instead of wasting time slinging around fairly weak accusations.

Yet it does not change the fact that an Intel VP and General Manager, Platform Evaluation and Competitive Assessment at Intel was president of BAPCo,its main contact a while back worked at Intel and they were located at Intel HQ in Santa Clara. That has been known for a very long time it appears - The Register reported on it in 2000.

I am suprised at how much of that info was around after finding that first post - it seems to be something levelled at BAPCo over a decade ago and it might be why for nearly a decade AMD,Nvidia and VIA have left the consortium.

Plus,you don't seem to understand the Sysmark is still used for UK and US government contracts,so even if they have something competitive(Zen) then what happens then...so I can see why they are trying to do this. Its purely a business move if anything.

Nvidia,VIA and AMD have left - so all the Sysmark/BAPCo defenders are contradicting themselves. Thats 3/4 of the companies which supplies parts for X86 PCs,and luckily for VIA and Nvidia their exposure is not as big. None of you are questioning why Nvidia or VIA left too.

It makes me wonder what would happen if Futuremark decided to suddently relocate to the AMD HQ and had Lisa Su as president.

Then Fury X starts edging out a GTX980TI in all the 3DMark benchmarks run by a few percent..

I expect everybody would say,sure blood,thats proper OK,innit??

I could imagine the ****storm if that happened.

This is why proper reviews,not silly benchmark suites, are important as is the user testing done on forums which are far more useful for people.

Its just far too easy to run data on benchmarks suites that could sway the results for a company. This is why even for independent reviews of CPUs and GPUs,there is variability in relative results.

But,then I expect you all know that anyway!

 

[kirts]

So correct me if I am wrong but what they say in the video is:
One benchmark uses all PC components and the other benchmark which they accuse uses only the CPU.

Then they compare the results of the two benchmarks and they find the performance is very different.

Well...of course it is! By using all components it is expected to have completely different results!