Question about Passmark PerformanceTest v9.0 B1030

[aldiney@yahoo.com]

have a question about CPU Benchmark scores.
First question I have is about score variation. What is the percentage variation that could the considered acceptable ? Can I consider 3% ?

Considering I am running Passmark v9.0 B1030 using the same CPU- for example, i3-8100, using different motherboards, memory size, HDD or SSD, the CPU Benchmark score should be the same, considering the variation of the test ?

I understand that if PerformanceTest is evaluating CPU only, it shouldn’t be affected by different memory size and/or if the Windows is installed in SSD or HDD.

[David (PassMark)]

PerformanceTest is a system benchmark. But the CPUMark portion does (mostly) only test the CPU.

But the CPU can't function without using some RAM, so the RAM speed and setup (e.g. dual channel) can impact the results. The RAM speed & configuration is in turn influenced (slightly) by the motherboard and the BIOS. But the hard drive should have no impact (but see comments below).

Variations from one run to the next depend on what background tasks are running (often randomly) on the machine. e.g. having a windows update or Anti-virus update occur mid test might result in a large variation. Thermal throttling can also impact the results.

[aldiney@yahoo.com]

Dear David
Thanks for you answer.
I did some tests and considering the same configuration, using a clean install of Windows 10 without any new update and the same drivers. What can be the percentage variation in values ?
I have 3 computers with the same configuration, bios version, Windows 10 RS6 clean installation with the same drivers version and no Windows update, to be sure that all the 3 samples are the same. The target score in CPU Mark is 9500, running only this test. I got score values from 9340 to 9620 points, running more than 500 tests in these 3 samples. These values are considered correct ?
If I run the benchmark repeatedly on a consistently performing system in a well controlled environment will produce scores that fall within a 3% range ? If not, what value can be considered ?

[David (PassMark)]

On a modern Windows machine there is always a massive amount of background activity taking place. Most of is random when it occurs. This impacts the consistency of benchmarking.

If you want to take a look download Microsoft's Process Monitor tool
https://docs.microsoft.com/en-us/sys...nloads/procmon

And also have a look at the background Network activity with Wireshark.
https://www.wireshark.org/

You'll be amazed what is going on without your knowledge.