CPU-Z v1.74 - MultiThread CPU Benchmark

[Dresdenboy]

3.06 Northwood in an old Dell.

Single 155
Multi 100

Likely HT cache thrashing at its best.

 

[Arachnotronic]

5960x @ 4.5GHz core/4.0GHz cache.

validation link:

http://valid.x86.fr/4ry9w5

 

[SPBHM]

Likely HT cache thrashing at its best.

I think there is a bug with single core CPUs, I got 0 for the MT score, while with "stress" it was showing the same score as I got for ST basically.

 

[biostud]

5820K @ 4.4Ghz

 

[Mercennarius]

Lenovo D20 - X5690(x2) - stock clocks

 

[MongGrel]

 

[Burpo]

not bad 🙂

 

[LTC8K6]

Likely HT cache thrashing at its best.

Yes, the first try at HT on desktop for Intel I believe. 🙂

 

[MrTeal]

http://valid.x86.fr/bx5u7e

 

[crashtech]

 

[Deders]

Fresh install of windows 10 has resolved my issues

 

[Edgemeal]

I forgot to mention, since this bench is still beta I'm going to stop adding scores to the chart at around 50 entries.

Thanks for posting! :thumbsup:

 

[fourdegrees11]

I forgot to mention, since this bench is still beta I'm going to stop adding scores to the chart at around 50 entries.

Thanks for posting! :thumbsup:

Is there something wrong with my entry?

 

[Edgemeal]

Is there something wrong with my entry?

Sorry, missed that one, added.

 

[Burpo]

Since you're only doing 50, there's also a double entry for spinejam you may want to fix..

As always, Thanks Edgemeal for another fun & enlightening thread 🙂

 

[Edgemeal]

Since you're only doing 50, there's also a double entry for spinejam you may want to fix..

As always, Thanks Edgemeal for another fun & enlightening thread 🙂

Your welcome, and thanks for catching the dup, I must be going blind!

 

[techjesse]

Here's Mine 🙂

 

[Edgemeal]

Here's Mine 🙂

Nice one! :thumbsup:

That makes 49 entries, do I hear 50?, going once, going twice,....

 

[warzeta]

Here is mine

 

[Tecnoworld]

I'd like to see the same benchmark implemented also in the android version of cpuz, if possible!

p.s.: my result (5820k@4.4ghz): 12669

 

[Conroe]

i5 6500

Memory timings don't seem to have much effect on this. It's all clock speed.

 

[Dresdenboy]

Memory timings don't seem to have much effect on this. It's all clock speed.

Yes, it seems to be very cache friendly and mainly depends on uarch.

I've seen a nice nearly linear increase in perf/clock over uarchs.

 

[inf64]

I think there is something wrong with this benchmark. Whenever I run it multiple times (more than once), my MT score decrease after each run. ST score remains as it is. After 6th or 7th run my MT score is lower than ST one... I tried this on Win7 and Win10, same thing happens. Also tried it on stock and OCed settings, no difference. Submitted the report to CPUz team, no reply yet.

 

[Dresdenboy]

I think there is something wrong with this benchmark. Whenever I run it multiple times (more than once), my MT score decrease after each run. ST score remains as it is. After 6th or 7th run my MT score is lower than ST one... I tried this on Win7 and Win10, same thing happens. Also tried it on stock and OCed settings, no difference. Submitted the report to CPUz team, no reply yet.

Still seems to be beta. Did you restart CPU-Z each time?

 

[SlowSpyder]

Broadwell i5 5300U (dual core) running @ 2.7GHz during the bench.