Passmark suddenly adding AVX512, intel ahead once again

[P4-630]

https://twitter.com/x/status/1238257119557832706

Intel opeens weer bovenaan PassMark Performance Test na wijzigingen - update

Passmark heeft voor het weekend diverse onderdelen van zijn PerformanceTest-benchmark aangepast, waardoor processors van Intel opeens weer bovenaan de g...

nl.hardware.info

PassMark verändert CPU-Benchmark erheblich zu Gunsten von Intel [Update] - Planet 3DNow!

Auf Planet 3DNow! gibt es alle wichtigen Informationen fr AMD-User: News, Downloads, Support, Tests

www.planet3dnow.de

 

[Solaris17]

And what are your thoughts on the matter?

 

[P4-630]

And what are your thoughts on the matter?

I was reading this on hardwareinfo and thought I'd share this.
Delete the post if you wish.

 

[TheoneandonlyMrK]

I was reading this on hardwareinfo and thought I'd share this.
Delete the post if you wish.

It's valid news elsewhere, I can't see them holding onto it long, Ryzens getting discounted, I see new chips all round on the way so it might be new Intel but maybe not.

 

[Vya Domus]

Coffee Lake doesn't even have AVX 512, so these Intel CPUs got faster by some other mysterious means.

Anyway I find it at the very least interesting how every time one of these benchmarks get's updated , i.e cpu-z, geekbench, etc it's always AMD that loses some of the scores and almost never Intel. It's just, you know, kind of odd.

 

[Assimilator]

News is supposed to be impartial and unbiased. An article linking to a tweet claiming "bribe money" is not news, it's scurrilous rumour-mill bulls**t. Sadly this is what is considered "journalism" in the tech space nowadays.

 

[Deleted member 50521]

AVX512 work wonders in non GPU related works when implemented for our bioinformatics work

 

[Vya Domus]

CPU benchmarks huge changes? - PassMark Support Forums

Hi, Today I wanted to check some CPUs benchmarks and was really surprised to see that the charts have changed significantly. I'm using Passmark benchmarks for many years and have never seen such changes. Here are some examples: Passmark few days ago Passmark today Change in % i7-2600 8181...

www.passmark.com

But for PerformanceTest V10 we did really major changes to the CPU test algorithms. These changes included
- Using new CPU instructions (e.g. AVX512) only available in modern CPUs.
- Use a more up to date compiler (Visual Studio 2019 instead of 2013) which also brings some code optimization.
- Have better support for out of order execution, which is a feature of newer CPUs.
- Updated the 3rd libraries we use for some of the tests (including more modern versions of GZip, Crypto++ and Bullet Physics.
- Fixed up a bunch of bugs that hurt performance (like some variable alignment issues and compiler optimization flags).
- Completely rewrote some of the tests. e.g. removed TwoFish encryption and replaced it with the more common Elliptic curve encryption.
- Improving the algorithms to push more data through the CPU also results in more load on the cache and memory subsystem. So older CPUs, those with inadequate cache or memory bandwidth are expected not to perform so well with PT10.

Bribe money or not I can't take seriously anyone that lists "better support for out of order execution" when that's something purely hardware related and has been around for about 20 years and not just in newer CPUs. It's like this has been written by some one pretending to know about these things, had I not known better it would have fooled me.

Also, MSVC for a CPU performance benchmark ? Big yikes.

 

[biffzinker]

Also, VSC for a performance benchmark ? Big yikes.

Would you rather have them run source code through ICC instead?

I'm curious about the: "- Have better support for out of order execution, which is a feature of newer CPUs."

 

[TheoneandonlyMrK]

News is supposed to be impartial and unbiased. An article linking to a tweet claiming "bribe money" is not news, it's scurrilous rumour-mill bulls**t. Sadly this is what is considered "journalism" in the tech space nowadays.

True, worthy of forum chat though I think.

 

[Vya Domus]

Would you rather have them run source code through ICC instead?

It wouldn't have been surprising to find out that ICC was the compiler all along.

On a more serious note, MSVC is notorious for generating sub-optimal code compared to other compilers like GCC and Clang. I do use MSVC , it has gotten much better over the years but it's still behind, for a CPU benchmark it would literally be my last choice.