New PassMark PerformanceTest Versions Allow for Expanded Cross-platform CPU Results

[btarunr]

Passmark Software have recently announced the significant expansion of their cross-platform benchmarking with the latest releases of PerformanceTest. For the first time this cornerstone benchmarking software is now available for Windows ARM, included with every PerformanceTest for Windows license. All CPU, 2D, Memory and Disk Tests have been natively compiled to run on Windows ARM, partial 3D support is also available.

PerformanceTest has also been released as free command line versions for Linux x86 64-bit, Linux ARM 32-bit, Linux ARM 64-bit and for MacOS (CPU Test Suite Only). The Android and IOS Mobile versions have also received updates, with all available versions of PerformanceTest now running the same CPU Test Suite found in V10 for Windows. These releases allow for directly comparable cross-platform CPU results, an apparent necessity as the number of ARM based processors such as Apple's M1 and Qualcomm's Snapdragon enter consumer hardware.



The CPU mark used to compare cross-platform results is a composite average of the integer, floating point, prime, sorting, encryption, compression, single threaded, physics and SIMD test scores. Below is a range of PassMark's most benchmarked CPU's, with the inclusion of recent baseline results from Qualcomm and Apple ARM based chips (in orange) for a cross-platform comparison.

Interestingly, initial baseline results for the Mac mini 9.1 show strong single threaded results, placing it in second position under Intel's CORE i9 11900K on the Single Threaded Performance chart.


PerformanceTest Windows ARM is now included with every purchase of PassMark's PerformanceTest for Windows license.

To download the newly released and free command line versions or for further information, visit this page.

View at TechPowerUp Main Site

 

[ViperXTR]

Look at that Apple M1 go >:]

 

[1d10t]

Alternative to highly controversial geekbench?

 

[nguyen]

So a lifestyle company in Cupertino make better CPU than both AMD and Intel

 

[Luminescent]

I never look at benchmarks for buying decisions, always look at specific application i use and how fast it performs.

 

[Searing]

Alternative to highly controversial geekbench?

We can now see how uncontroversial geekbench really is. Pretty much every benchmark agrees with geekbench, across all hardware and all OS. I think geekbench got a bad rap because Antutu is so bad, and people get those confused.

 

[harm9963]

5000 need fast RAM , skew the data with slow ram .

 

[Vya Domus]

Alternative to highly controversial geekbench?

Only time will tell, the issue with Geekbench was that it was blatantly optimized specifically for Apple's chips, you could see that because with every new version their chips always got faster somehow compared to everything else.

In fact I can't really trust any benchmark that isn't open source, no way to be sure of anything if I can't see what's under the hood.

 

[stimpy88]

Interesting choice of CPU's there, especially the single threaded ones... Does Intel pay towards this? /s

 

[birdie]

Alternative to highly controversial geekbench?

Geekbench is alright - it closely matches the SPEC test which is recognized by the industry.

 

[john_]

I wonder.

If I was someone expecting to get a paycheck from AMD and Intel, it would have been logical to assume that Intel's paycheck would be bigger.

But what if I was expecting to get a paycheck from Apple and Intel? Who would give me more?

 

[evernessince]

So a lifestyle company in Cupertino make better CPU than both AMD and Intel

Cross-platform benchmarking really isn't that simple. Comparing two different architectures means having two different code paths that could very well have wide divergence in regards to efficiency.

Mind you, Apple doesn't top the charts in any of these benchmarks either so even if you ignore the prior mentioned note, in fact a lifestyle company is not beating AMD / Intel. It's only a single bench as well so drawing a broad conclusion is foolish at best. Of course, calling Apple a lifestyle company is disingenuous as well. They are worth over a trillion dollars nor is this their first CPU.

 

[TumbleGeorge]

32 bit why? This is so ancient.

 

[Vya Domus]

Mind you, Apple doesn't top the charts in any of these benchmarks either so even if you ignore the prior mentioned note, in fact a lifestyle company is not beating AMD / Intel. It's only a single bench as well so drawing a broad conclusion is foolish at best. Of course, calling Apple a lifestyle company is disingenuous as well. They are worth over a trillion dollars nor is this their first CPU.

And people still have no idea what these companies are about and what is it that they want to achieve with the chips they make. They still have this dumb and simplified view that since they all make CPUs that must mean they wall want to perform the best in X metric and if one of them performs the best in X everyone else is a loser.

 

[nguyen]

Cross-platform benchmarking really isn't that simple. Comparing two different architectures means having two different code paths that could very well have wide divergence in regards to efficiency.

Mind you, Apple doesn't top the charts in any of these benchmarks either so even if you ignore the prior mentioned note, in fact a lifestyle company is not beating AMD / Intel. It's only a single bench as well so drawing a broad conclusion is foolish at best. Of course, calling Apple a lifestyle company is disingenuous as well. They are worth over a trillion dollars nor is this their first CPU.

Mind you I was quoting P.Gelsinger, Intel's new CEO that Apple is a "Lifestyle Company", surely you don't really think you are smarter than Intel new CEO do you
And I was clearly making a light-hearted joke.
Also in the single core benchmark the M1 only lose to unreleased Intel 11th Gen, so clearly there are some truths to my joke

 

[InVasMani]

Doesn't really matter how fast Apple's M1 chip is if it's overtly locked down with a confined usage scope. By contrast Intel/AMD x86 will run circles around it in terms of the swath of programs people can run on said chips. Especially true in the case of programs that are freely available like Fortnite. Speaking of let's compare how Fortnite is on the M1 versus Intel/AMD oh wait.

 

[nguyen]

Doesn't really matter how fast Apple's M1 chip is if it's overtly locked down with a confined usage scope. By contrast Intel/AMD x86 will run circles around it in terms of the swath of programs people can run on said chips. Especially true in the case of programs that are freely available like Fortnite.

I don't use or care about Apple products either, just putting credit where credit is due I guess.

 

[InVasMani]

Just pointing out the difference Intel/AMD are more open and general purpose both in regard to software used with the hardware, but also the hardware platform options while Apple is a state of the art prison complex by comparison.

 

[Nater]

CPU speed clearly scales with market cap.

 

[Makaveli]

5000 need fast RAM , skew the data with slow ram .

Indeed there Zen 3 system is most likely all stock and not tuned.

I'm seeing 3619 on my system for 5800X.

 

[InVasMani]

I wonder how the Xbox X would do at Passmark.

 

[ThrashZone]

Hi,
I find it a little odd people buy this benchmark.

 

[WeeRab]

Interesting that Intel should make their as yet unreleased halo chip available to Passmark and not any other b/marking utility.
Just sayin'.......
As for Apple's M1...Well it's kind of irrelevant. They only have a miniscule share of the PC market.