Perpetual benchmark thread... - Page 23

James Heinrich · 2011-08-05, 00:05

Quote:

Originally Posted by davieddy

Nicht Wahr?

Despite an appropriate-sounding family name, what I wrote is about as much as I actually understand the language

Even with some assistance from Google, I'm not quite sure I understand what you're saying -- that FFTs have sizes, but those sizes are not measured in bytes, so I should express it as 3M, but not 3MB? If that's all, why didn't you say so?

davieddy · 2011-08-05, 00:51

Quote:

Originally Posted by James Heinrich

Despite an appropriate-sounding family name, what I wrote is about as much as I actually understand the language

Even with some assistance from Google, I'm not quite sure I understand what you're saying -- that FFTs have sizes, but those sizes are not measured in bytes, so I should express it as 3M, but not 3MB? If that's all, why didn't you say so?

That was precisely what I was saying, and hence a big deal it
of out to make I wasn't.

I passed German "O level" in 1965, and am grateful for it, although
predictably the several agreeable German volk I have met speaky Englisch
besser than I Deutch do.
Soon after we started to learn the "word order" (age~12) my best friend
and I would in Englisch that order always speak.

David

cheesehead · 2011-08-10, 12:22

Quote:

Originally Posted by davieddy

What means the "B" in "FFT size"?

David

Though James is usually spot-on, I'm afraid that this time you've spotted that he's inserted a "B" where there should not be one.

Quote:

Originally Posted by James Heinrich

Here is the actual comparison between the two CPUs. At the same clock speed (2160MHz), other than a small difference between 1MB and 3MB FFT sizes, performance is nearly identical.

Quote:

Originally Posted by James Heinrich

The only place I see "B" is in my above quote about 1MB-3MB, which would be 1024K-3072K FFT sizes (or 1048576-3145728 if you prefer), as illustrated by the attached graph.

James,

Perhaps you've momentarily forgotten that FFT sizes are customarily measured by the number of floating-point "digits" rather than bytes. (That's why the mersenne-aries site lists "768K", not "768KB", for instance, as an FFT size.)

- - -

In Prime95, the FFT "digits" are each a double-precision FP number, spanning 8 bytes.

So, in GIMPS, an FFT of size "1024K" or "1M" occupies 8MB, not 1 MB, and a "3M" FFT occupies 24MB.

If some other FFT software were using single-precision FP for "digits", there a "1M" FFT would occupy 4MB and a "3M" FFT would span 12 MB. That shows why FFT sizes are not listed in MB or KB: because then one would have to also specify the size of each FFT "digit" in order to get a proper FFT picture.

4MB would contain 1M digits of single-precision FP, but only 512K digits of double-precision FP, or only 256K digits of quad-precision FP.

James Heinrich · 2011-08-10, 13:17

Quote:

Originally Posted by cheesehead

Perhaps you've momentarily forgotten...

A nice, diplomatic way of saying I don't know of what I speak. Which in this case is entirely true.

Thanks to your cogent explanation, I now understand why I was wrong.

davieddy · 2011-08-10, 19:28

Quote:

Originally Posted by James Heinrich

A nice, diplomatic way of saying I don't know of what I speak. Which in this case is entirely true.

Thanks to your cogent explanation, I now understand why I was wrong.

These sayings into German for me please translate:

"Of which we cannot speak, we cannot know" (Wittgenstein)
"We can know. We must know" (Hilbert)
"Manuel, this wasn't a proposition from Wittgenstein: I said
GET THOSE DEAD PIGEONS OUT OF THE WATER
TANK BEFORE THE HOTEL INSPECTORS ARRIVE" (Basil Fawlty)
"That's my sunlounger. I arose earlier than you and reserved
it by placing my towel thereon". (Tourist)

(Nuff quotes for now - Ed.)

David

PS Don't think I mentioned the War or Geoff Hurst did I?
Well I may have done once, but I think I got away with it.

cheesehead · 2011-08-10, 23:03

Quote:

Originally Posted by James Heinrich

A nice, diplomatic way of saying I don't know of what I speak. Which in this case is entirely true.

Thanks to your cogent explanation, I now understand why I was wrong.

You're welcome. :-)

ciic · 2011-08-28, 14:51

Hi !
-------------------------------------------------------------------------------------------------------------------
Here's just some more perf data on the new Sandy Bridge Pentium:

Intel(R) Pentium(R) CPU G620T @ 2.20GHz 35W TDP (System ~45-55W)

[Thu Aug 4 04:25:30 2011]
Compare your results to other computers at http://www.mersenne.org/report_benchmarks
Intel(R) Pentium(R) CPU G620T @ 2.20GHz
CPU speed: 2160.98 MHz, 2 cores
CPU features: Prefetch, MMX, SSE, SSE2, SSE4
L1 cache size: 32 KB
L2 cache size: 256 KB, L3 cache size: 3 MB
L1 cache line size: 64 bytes
L2 cache line size: 64 bytes
TLBS: 64
Prime95 64-bit version 26.6, RdtscTiming=1
Best time for 768K FFT length: 11.801 ms., avg: 11.817 ms.
Best time for 896K FFT length: 14.451 ms., avg: 14.473 ms.
Best time for 1024K FFT length: 15.775 ms., avg: 15.802 ms.
Best time for 1280K FFT length: 20.394 ms., avg: 20.422 ms.
Best time for 1536K FFT length: 24.889 ms., avg: 24.912 ms.
Best time for 1792K FFT length: 30.053 ms., avg: 30.067 ms.
Best time for 2048K FFT length: 33.272 ms., avg: 33.296 ms.
Best time for 2560K FFT length: 42.139 ms., avg: 42.156 ms.
Best time for 3072K FFT length: 51.821 ms., avg: 51.844 ms.
Best time for 3584K FFT length: 62.475 ms., avg: 62.494 ms.
Best time for 4096K FFT length: 70.048 ms., avg: 70.056 ms.
Best time for 5120K FFT length: 90.898 ms., avg: 90.910 ms.
Best time for 6144K FFT length: 112.647 ms., avg: 112.688 ms.
Best time for 7168K FFT length: 137.351 ms., avg: 137.421 ms.
Best time for 8192K FFT length: 148.731 ms., avg: 148.790 ms.
Timing FFTs using 2 threads.
Best time for 768K FFT length: 6.337 ms., avg: 6.366 ms.
Best time for 896K FFT length: 7.789 ms., avg: 7.807 ms.
Best time for 1024K FFT length: 8.484 ms., avg: 8.528 ms.
Best time for 1280K FFT length: 10.847 ms., avg: 10.868 ms.
Best time for 1536K FFT length: 13.067 ms., avg: 13.080 ms.
Best time for 1792K FFT length: 16.338 ms., avg: 16.372 ms.
Best time for 2048K FFT length: 17.974 ms., avg: 18.129 ms.
Best time for 2560K FFT length: 22.592 ms., avg: 22.624 ms.
Best time for 3072K FFT length: 27.876 ms., avg: 27.908 ms.
Best time for 3584K FFT length: 33.386 ms., avg: 33.512 ms.
Best time for 4096K FFT length: 36.597 ms., avg: 36.630 ms.
Best time for 5120K FFT length: 47.321 ms., avg: 47.347 ms.
Best time for 6144K FFT length: 59.132 ms., avg: 59.219 ms.
Best time for 7168K FFT length: 77.364 ms., avg: 77.638 ms.
Best time for 8192K FFT length: 78.744 ms., avg: 79.095 ms.
Best time for 61 bit trial factors: 3.599 ms.
Best time for 62 bit trial factors: 3.640 ms.
Best time for 63 bit trial factors: 4.084 ms.
Best time for 64 bit trial factors: 4.256 ms.
Best time for 65 bit trial factors: 4.954 ms.
Best time for 66 bit trial factors: 5.820 ms.
Best time for 67 bit trial factors: 5.773 ms.
Best time for 75 bit trial factors: 5.624 ms.
Best time for 76 bit trial factors: 5.610 ms.
Best time for 77 bit trial factors: 5.607 ms.
[Thu Aug 4 04:45:53 2011]
Compare your results to other computers at http://www.mersenne.org/report_benchmarks
Intel(R) Pentium(R) CPU G620T @ 2.20GHz
CPU speed: 2159.71 MHz, 2 cores
CPU features: Prefetch, MMX, SSE, SSE2, SSE4
L1 cache size: 32 KB
L2 cache size: 256 KB, L3 cache size: 3 MB
L1 cache line size: 64 bytes
L2 cache line size: 64 bytes
TLBS: 64
Prime95 64-bit version 26.6, RdtscTiming=1
Best time for 768K FFT length: 11.781 ms., avg: 11.802 ms.
Best time for 896K FFT length: 14.514 ms., avg: 14.542 ms.
Best time for 1024K FFT length: 15.906 ms., avg: 15.929 ms.
Best time for 1280K FFT length: 20.630 ms., avg: 20.656 ms.
Best time for 1536K FFT length: 25.333 ms., avg: 25.360 ms.
Best time for 1792K FFT length: 30.390 ms., avg: 30.434 ms.
Best time for 2048K FFT length: 33.557 ms., avg: 33.578 ms.
Best time for 2560K FFT length: 42.977 ms., avg: 43.029 ms.
Best time for 3072K FFT length: 52.514 ms., avg: 52.534 ms.
Best time for 3584K FFT length: 63.344 ms., avg: 63.374 ms.
Best time for 4096K FFT length: 70.448 ms., avg: 70.475 ms.
Best time for 5120K FFT length: 91.047 ms., avg: 91.087 ms.
Best time for 6144K FFT length: 112.786 ms., avg: 112.865 ms.
Best time for 7168K FFT length: 136.480 ms., avg: 136.524 ms.
Best time for 8192K FFT length: 148.853 ms., avg: 148.892 ms.
Timing FFTs using 2 threads.
Best time for 768K FFT length: 6.298 ms., avg: 6.368 ms.
Best time for 896K FFT length: 7.776 ms., avg: 7.803 ms.
Best time for 1024K FFT length: 8.452 ms., avg: 8.480 ms.
Best time for 1280K FFT length: 10.881 ms., avg: 10.893 ms.
Best time for 1536K FFT length: 13.060 ms., avg: 13.086 ms.
Best time for 1792K FFT length: 16.126 ms., avg: 16.152 ms.
Best time for 2048K FFT length: 18.057 ms., avg: 18.085 ms.
Best time for 2560K FFT length: 22.595 ms., avg: 22.654 ms.
Best time for 3072K FFT length: 27.772 ms., avg: 27.859 ms.
Best time for 3584K FFT length: 33.517 ms., avg: 33.549 ms.
Best time for 4096K FFT length: 37.300 ms., avg: 37.351 ms.
Best time for 5120K FFT length: 48.189 ms., avg: 48.293 ms.
Best time for 6144K FFT length: 58.756 ms., avg: 58.806 ms.
Best time for 7168K FFT length: 80.770 ms., avg: 80.839 ms.
Best time for 8192K FFT length: 80.001 ms., avg: 80.046 ms.
Best time for 61 bit trial factors: 3.597 ms.
Best time for 62 bit trial factors: 3.638 ms.
Best time for 63 bit trial factors: 4.083 ms.
Best time for 64 bit trial factors: 4.256 ms.
Best time for 65 bit trial factors: 4.955 ms.
Best time for 66 bit trial factors: 5.817 ms.
Best time for 67 bit trial factors: 5.772 ms.
Best time for 75 bit trial factors: 5.622 ms.
Best time for 76 bit trial factors: 5.606 ms.
Best time for 77 bit trial factors: 5.616 ms.
[Thu Aug 4 04:54:47 2011]
Compare your results to other computers at http://www.mersenne.org/report_benchmarks
Intel(R) Pentium(R) CPU G620T @ 2.20GHz
CPU speed: 2160.56 MHz, 2 cores
CPU features: Prefetch, MMX, SSE, SSE2, SSE4
L1 cache size: 32 KB
L2 cache size: 256 KB, L3 cache size: 3 MB
L1 cache line size: 64 bytes
L2 cache line size: 64 bytes
TLBS: 64
Prime95 64-bit version 26.6, RdtscTiming=1
Best time for 768K FFT length: 11.771 ms., avg: 11.796 ms.
Best time for 896K FFT length: 14.500 ms., avg: 14.523 ms.
Best time for 1024K FFT length: 15.912 ms., avg: 15.938 ms.
Best time for 1280K FFT length: 20.677 ms., avg: 20.692 ms.
Best time for 1536K FFT length: 25.299 ms., avg: 25.323 ms.
Best time for 1792K FFT length: 30.449 ms., avg: 30.471 ms.
Best time for 2048K FFT length: 33.537 ms., avg: 33.556 ms.
Best time for 2560K FFT length: 43.151 ms., avg: 43.161 ms.
Best time for 3072K FFT length: 52.367 ms., avg: 52.384 ms.
Best time for 3584K FFT length: 63.243 ms., avg: 63.290 ms.
Best time for 4096K FFT length: 70.376 ms., avg: 70.398 ms.
Best time for 5120K FFT length: 91.385 ms., avg: 91.410 ms.
Best time for 6144K FFT length: 111.428 ms., avg: 111.459 ms.
Best time for 7168K FFT length: 136.892 ms., avg: 136.968 ms.
Best time for 8192K FFT length: 148.951 ms., avg: 149.047 ms.
Timing FFTs using 2 threads.
Best time for 768K FFT length: 6.336 ms., avg: 6.359 ms.
Best time for 896K FFT length: 7.778 ms., avg: 7.827 ms.
Best time for 1024K FFT length: 8.479 ms., avg: 8.515 ms.
Best time for 1280K FFT length: 10.831 ms., avg: 10.948 ms.
Best time for 1536K FFT length: 13.049 ms., avg: 13.069 ms.
Best time for 1792K FFT length: 16.066 ms., avg: 16.089 ms.
Best time for 2048K FFT length: 17.947 ms., avg: 17.961 ms.
Best time for 2560K FFT length: 22.665 ms., avg: 22.689 ms.
Best time for 3072K FFT length: 27.663 ms., avg: 27.783 ms.
Best time for 3584K FFT length: 33.524 ms., avg: 33.572 ms.
Best time for 4096K FFT length: 36.449 ms., avg: 36.526 ms.
Best time for 5120K FFT length: 47.151 ms., avg: 47.301 ms.
Best time for 6144K FFT length: 59.534 ms., avg: 59.658 ms.
Best time for 7168K FFT length: 77.943 ms., avg: 77.981 ms.
Best time for 8192K FFT length: 79.611 ms., avg: 79.806 ms.
Best time for 61 bit trial factors: 3.599 ms.
Best time for 62 bit trial factors: 3.637 ms.
Best time for 63 bit trial factors: 4.083 ms.
Best time for 64 bit trial factors: 4.256 ms.
Best time for 65 bit trial factors: 4.956 ms.
Best time for 66 bit trial factors: 5.815 ms.
Best time for 67 bit trial factors: 5.773 ms.
Best time for 75 bit trial factors: 5.621 ms.
Best time for 76 bit trial factors: 5.610 ms.
Best time for 77 bit trial factors: 5.608 ms.

Think my benchmarks seem a bit faster than that of the mersenne-aries.sili.net/throughput.php webpage??

Cheers.. ciic

ciic · 2011-08-28, 15:02

Hi!
---------------------------------------------------
Found some new Wikipedia-Edits:

* http://en.wikipedia.org/wiki/Prime95

and

* http://de.wikipedia.org/wiki/Prime95

Would like some of you to comment - constructively - on them (accuracy, collum/row-titles, physical quatities & units of measurements) or even would like to appeal you for contributional edits - please.

Cheers, ciic

James Heinrich · 2011-12-16, 16:45

Some Sandy-E benchmarks:

v26.3 64-bit:

Code:

Compare your results to other computers at http://www.mersenne.org/report_benchmarks
Intel(R) Core(TM) i7-3930K CPU @ 3.20GHz
CPU speed: 4902.50 MHz, 6 hyperthreaded cores
CPU features: Prefetch, MMX, SSE, SSE2, SSE4, AVX
L1 cache size: 32 KB
L2 cache size: 256 KB, L3 cache size: 12 MB
L1 cache line size: 64 bytes
L2 cache line size: 64 bytes
TLBS: 64
Prime95 64-bit version 26.6, RdtscTiming=1
Best time for 768K FFT length: 5.249 ms., avg: 5.372 ms.
Best time for 896K FFT length: 6.434 ms., avg: 6.797 ms.
Best time for 1024K FFT length: 7.101 ms., avg: 7.272 ms.
Best time for 1280K FFT length: 9.211 ms., avg: 9.615 ms.
Best time for 1536K FFT length: 11.204 ms., avg: 11.408 ms.
Best time for 1792K FFT length: 13.669 ms., avg: 13.980 ms.
Best time for 2048K FFT length: 15.108 ms., avg: 15.709 ms.
Best time for 2560K FFT length: 19.325 ms., avg: 19.803 ms.
Best time for 3072K FFT length: 23.788 ms., avg: 24.124 ms.
Best time for 3584K FFT length: 28.732 ms., avg: 29.131 ms.
Best time for 4096K FFT length: 31.759 ms., avg: 32.335 ms.
Best time for 5120K FFT length: 41.248 ms., avg: 41.896 ms.
Best time for 6144K FFT length: 51.552 ms., avg: 52.863 ms.
Best time for 7168K FFT length: 63.371 ms., avg: 64.978 ms.
Best time for 8192K FFT length: 69.446 ms., avg: 70.034 ms.
Timing FFTs using 2 threads on 1 physical CPUs.
Best time for 768K FFT length: 5.150 ms., avg: 5.271 ms.
Best time for 896K FFT length: 6.206 ms., avg: 6.267 ms.
Best time for 1024K FFT length: 6.975 ms., avg: 7.008 ms.
Best time for 1280K FFT length: 8.971 ms., avg: 9.087 ms.
Best time for 1536K FFT length: 11.051 ms., avg: 11.096 ms.
Best time for 1792K FFT length: 13.366 ms., avg: 13.421 ms.
Best time for 2048K FFT length: 15.002 ms., avg: 15.033 ms.
Best time for 2560K FFT length: 18.984 ms., avg: 19.038 ms.
Best time for 3072K FFT length: 23.587 ms., avg: 23.628 ms.
Best time for 3584K FFT length: 28.255 ms., avg: 28.329 ms.
Best time for 4096K FFT length: 31.588 ms., avg: 31.640 ms.
Best time for 5120K FFT length: 40.923 ms., avg: 40.967 ms.
Best time for 6144K FFT length: 50.793 ms., avg: 51.226 ms.
Best time for 7168K FFT length: 59.152 ms., avg: 59.181 ms.
Best time for 8192K FFT length: 66.920 ms., avg: 67.097 ms.
Timing FFTs using 4 threads on 2 physical CPUs.
Best time for 768K FFT length: 2.663 ms., avg: 2.741 ms.
Best time for 896K FFT length: 3.235 ms., avg: 3.252 ms.
Best time for 1024K FFT length: 3.630 ms., avg: 3.985 ms.
Best time for 1280K FFT length: 4.670 ms., avg: 4.787 ms.
Best time for 1536K FFT length: 5.688 ms., avg: 5.759 ms.
Best time for 1792K FFT length: 6.940 ms., avg: 6.981 ms.
Best time for 2048K FFT length: 7.805 ms., avg: 7.824 ms.
Best time for 2560K FFT length: 9.868 ms., avg: 9.896 ms.
Best time for 3072K FFT length: 12.214 ms., avg: 12.257 ms.
Best time for 3584K FFT length: 14.579 ms., avg: 14.602 ms.
Best time for 4096K FFT length: 16.381 ms., avg: 16.442 ms.
Best time for 5120K FFT length: 21.108 ms., avg: 21.162 ms.
Best time for 6144K FFT length: 26.151 ms., avg: 26.263 ms.
Best time for 7168K FFT length: 30.675 ms., avg: 30.770 ms.
Best time for 8192K FFT length: 34.609 ms., avg: 35.164 ms.
Timing FFTs using 6 threads on 3 physical CPUs.
Best time for 768K FFT length: 2.004 ms., avg: 2.254 ms.
Best time for 896K FFT length: 2.229 ms., avg: 2.240 ms.
Best time for 1024K FFT length: 2.771 ms., avg: 3.001 ms.
Best time for 1280K FFT length: 3.298 ms., avg: 3.424 ms.
Best time for 1536K FFT length: 3.990 ms., avg: 4.050 ms.
Best time for 1792K FFT length: 4.802 ms., avg: 5.015 ms.
Best time for 2048K FFT length: 5.409 ms., avg: 5.432 ms.
Best time for 2560K FFT length: 6.825 ms., avg: 6.937 ms.
Best time for 3072K FFT length: 8.378 ms., avg: 8.816 ms.
Best time for 3584K FFT length: 9.979 ms., avg: 10.013 ms.
Best time for 4096K FFT length: 11.182 ms., avg: 11.218 ms.
Best time for 5120K FFT length: 14.448 ms., avg: 14.647 ms.
Best time for 6144K FFT length: 17.762 ms., avg: 18.054 ms.
Best time for 7168K FFT length: 21.398 ms., avg: 21.788 ms.
Best time for 8192K FFT length: 23.608 ms., avg: 24.434 ms.
Timing FFTs using 8 threads on 4 physical CPUs.
Best time for 768K FFT length: 1.775 ms., avg: 1.934 ms.
Best time for 896K FFT length: 1.744 ms., avg: 1.782 ms.
Best time for 1024K FFT length: 2.445 ms., avg: 2.479 ms.
Best time for 1280K FFT length: 2.880 ms., avg: 3.749 ms.
Best time for 1536K FFT length: 3.403 ms., avg: 3.427 ms.
Best time for 1792K FFT length: 3.717 ms., avg: 3.894 ms.
Best time for 2048K FFT length: 4.282 ms., avg: 4.655 ms.
Best time for 2560K FFT length: 5.603 ms., avg: 5.638 ms.
Best time for 3072K FFT length: 6.641 ms., avg: 6.824 ms.
Best time for 3584K FFT length: 7.756 ms., avg: 7.806 ms.
Best time for 4096K FFT length: 8.872 ms., avg: 9.270 ms.
Best time for 5120K FFT length: 11.347 ms., avg: 12.227 ms.
Best time for 6144K FFT length: 14.006 ms., avg: 14.029 ms.
Best time for 7168K FFT length: 17.100 ms., avg: 17.281 ms.
Best time for 8192K FFT length: 18.540 ms., avg: 18.695 ms.
Timing FFTs using 10 threads on 5 physical CPUs.
Best time for 768K FFT length: 1.665 ms., avg: 1.882 ms.
Best time for 896K FFT length: 1.613 ms., avg: 1.627 ms.
Best time for 1024K FFT length: 2.284 ms., avg: 2.684 ms.
Best time for 1280K FFT length: 2.687 ms., avg: 2.723 ms.
Best time for 1536K FFT length: 3.174 ms., avg: 3.898 ms.
Best time for 1792K FFT length: 3.428 ms., avg: 3.860 ms.
Best time for 2048K FFT length: 4.033 ms., avg: 4.343 ms.
Best time for 2560K FFT length: 5.202 ms., avg: 6.459 ms.
Best time for 3072K FFT length: 6.246 ms., avg: 7.263 ms.
Best time for 3584K FFT length: 7.092 ms., avg: 7.355 ms.
Best time for 4096K FFT length: 8.392 ms., avg: 8.579 ms.
Best time for 5120K FFT length: 10.516 ms., avg: 12.441 ms.
Best time for 6144K FFT length: 12.917 ms., avg: 15.059 ms.
Best time for 7168K FFT length: 15.478 ms., avg: 15.564 ms.
Best time for 8192K FFT length: 17.182 ms., avg: 17.832 ms.
Timing FFTs using 12 threads on 6 physical CPUs.
Best time for 768K FFT length: 1.572 ms., avg: 1.849 ms.
Best time for 896K FFT length: 1.536 ms., avg: 1.605 ms.
Best time for 1024K FFT length: 2.178 ms., avg: 2.451 ms.
Best time for 1280K FFT length: 2.575 ms., avg: 2.907 ms.
Best time for 1536K FFT length: 3.001 ms., avg: 3.316 ms.
Best time for 1792K FFT length: 3.294 ms., avg: 3.962 ms.
Best time for 2048K FFT length: 3.877 ms., avg: 3.973 ms.
Best time for 2560K FFT length: 4.992 ms., avg: 5.436 ms.
Best time for 3072K FFT length: 5.927 ms., avg: 6.781 ms.
Best time for 3584K FFT length: 6.951 ms., avg: 7.774 ms.
Best time for 4096K FFT length: 7.927 ms., avg: 10.767 ms.
Best time for 5120K FFT length: 10.002 ms., avg: 10.090 ms.
Best time for 6144K FFT length: 12.233 ms., avg: 12.538 ms.
Best time for 7168K FFT length: 17.659 ms., avg: 18.278 ms.
Best time for 8192K FFT length: 16.496 ms., avg: 20.024 ms.
Best time for 61 bit trial factors: 1.574 ms.
Best time for 62 bit trial factors: 1.587 ms.
Best time for 63 bit trial factors: 1.794 ms.
Best time for 64 bit trial factors: 1.862 ms.
Best time for 65 bit trial factors: 2.180 ms.
Best time for 66 bit trial factors: 2.576 ms.
Best time for 67 bit trial factors: 2.547 ms.
Best time for 75 bit trial factors: 2.481 ms.
Best time for 76 bit trial factors: 2.479 ms.
Best time for 77 bit trial factors: 2.476 ms.

James Heinrich · 2011-12-16, 16:47

v27.1 32-bit:

Code:

[Fri Dec 16 10:36:06 2011]
Compare your results to other computers at http://www.mersenne.org/report_benchmarks
Intel(R) Core(TM) i7-3930K CPU @ 3.20GHz
CPU speed: 4902.50 MHz, 6 hyperthreaded cores
CPU features: Prefetch, MMX, SSE, SSE2, SSE4, AVX
L1 cache size: 32 KB
L2 cache size: 256 KB, L3 cache size: 12 MB
L1 cache line size: 64 bytes
L2 cache line size: 64 bytes
TLBS: 64
Prime95 32-bit version 27.1, RdtscTiming=1
Best time for 768K FFT length: 3.677 ms., avg: 3.760 ms.
Best time for 1024K FFT length: 5.186 ms., avg: 5.267 ms.
Best time for 1280K FFT length: 6.910 ms., avg: 7.204 ms.
Best time for 1536K FFT length: 8.306 ms., avg: 8.447 ms.
Best time for 2048K FFT length: 11.293 ms., avg: 11.723 ms.
Best time for 2560K FFT length: 14.685 ms., avg: 14.878 ms.
Best time for 3072K FFT length: 17.964 ms., avg: 18.225 ms.
Best time for 4096K FFT length: 26.006 ms., avg: 26.514 ms.
Best time for 5120K FFT length: 32.573 ms., avg: 33.259 ms.
Best time for 6144K FFT length: 38.950 ms., avg: 39.090 ms.
Best time for 8192K FFT length: 54.119 ms., avg: 55.810 ms.
Timing FFTs using 2 threads on 1 physical CPUs.
Best time for 768K FFT length: 4.128 ms., avg: 4.139 ms.
Best time for 1024K FFT length: 5.534 ms., avg: 5.575 ms.
Best time for 1280K FFT length: 7.372 ms., avg: 7.525 ms.
Best time for 1536K FFT length: 8.963 ms., avg: 9.026 ms.
Best time for 2048K FFT length: 12.040 ms., avg: 12.236 ms.
Best time for 2560K FFT length: 15.943 ms., avg: 16.192 ms.
Best time for 3072K FFT length: 19.521 ms., avg: 19.635 ms.
Best time for 4096K FFT length: 28.020 ms., avg: 28.191 ms.
Best time for 5120K FFT length: 34.828 ms., avg: 35.223 ms.
Best time for 6144K FFT length: 42.825 ms., avg: 43.272 ms.
Best time for 8192K FFT length: 57.334 ms., avg: 57.615 ms.
Timing FFTs using 4 threads on 2 physical CPUs.
Best time for 768K FFT length: 3.069 ms., avg: 3.460 ms.
Best time for 1024K FFT length: 4.057 ms., avg: 4.234 ms.
Best time for 1280K FFT length: 5.331 ms., avg: 5.898 ms.
Best time for 1536K FFT length: 5.438 ms., avg: 5.505 ms.
Best time for 2048K FFT length: 6.870 ms., avg: 7.100 ms.
Best time for 2560K FFT length: 8.982 ms., avg: 9.064 ms.
Best time for 3072K FFT length: 10.922 ms., avg: 11.323 ms.
Best time for 4096K FFT length: 15.655 ms., avg: 15.923 ms.
Best time for 5120K FFT length: 19.384 ms., avg: 19.493 ms.
Best time for 6144K FFT length: 23.860 ms., avg: 24.282 ms.
Best time for 8192K FFT length: 32.104 ms., avg: 32.649 ms.
Timing FFTs using 6 threads on 3 physical CPUs.
Best time for 768K FFT length: 2.762 ms., avg: 2.879 ms.
Best time for 1024K FFT length: 3.664 ms., avg: 4.008 ms.
Best time for 1280K FFT length: 4.748 ms., avg: 5.452 ms.
Best time for 1536K FFT length: 4.778 ms., avg: 5.322 ms.
Best time for 2048K FFT length: 5.758 ms., avg: 5.887 ms.
Best time for 2560K FFT length: 7.476 ms., avg: 7.697 ms.
Best time for 3072K FFT length: 8.956 ms., avg: 10.168 ms.
Best time for 4096K FFT length: 12.729 ms., avg: 14.813 ms.
Best time for 5120K FFT length: 15.632 ms., avg: 17.910 ms.
Best time for 6144K FFT length: 19.205 ms., avg: 20.806 ms.
Best time for 8192K FFT length: 25.762 ms., avg: 28.414 ms.
Timing FFTs using 8 threads on 4 physical CPUs.
Best time for 768K FFT length: 2.577 ms., avg: 3.193 ms.
Best time for 1024K FFT length: 3.483 ms., avg: 4.013 ms.
Best time for 1280K FFT length: 4.511 ms., avg: 5.158 ms.
Best time for 1536K FFT length: 4.490 ms., avg: 4.824 ms.
Best time for 2048K FFT length: 5.361 ms., avg: 6.380 ms.
Best time for 2560K FFT length: 6.916 ms., avg: 7.287 ms.
Best time for 3072K FFT length: 8.313 ms., avg: 9.672 ms.
Best time for 4096K FFT length: 11.683 ms., avg: 13.380 ms.
Best time for 5120K FFT length: 14.410 ms., avg: 15.375 ms.
Best time for 6144K FFT length: 17.554 ms., avg: 18.603 ms.
Best time for 8192K FFT length: 23.688 ms., avg: 27.285 ms.
Timing FFTs using 10 threads on 5 physical CPUs.
Best time for 768K FFT length: 2.478 ms., avg: 2.701 ms.
Best time for 1024K FFT length: 3.347 ms., avg: 3.571 ms.
Best time for 1280K FFT length: 4.310 ms., avg: 4.446 ms.
Best time for 1536K FFT length: 4.316 ms., avg: 5.180 ms.
Best time for 2048K FFT length: 5.132 ms., avg: 5.437 ms.
Best time for 2560K FFT length: 6.579 ms., avg: 7.673 ms.
Best time for 3072K FFT length: 7.893 ms., avg: 8.491 ms.
Best time for 4096K FFT length: 11.130 ms., avg: 13.262 ms.
Best time for 5120K FFT length: 13.623 ms., avg: 16.028 ms.
Best time for 6144K FFT length: 16.623 ms., avg: 16.670 ms.
Best time for 8192K FFT length: 22.197 ms., avg: 23.030 ms.
Timing FFTs using 12 threads on 6 physical CPUs.
Best time for 768K FFT length: 2.428 ms., avg: 2.933 ms.
Best time for 1024K FFT length: 3.282 ms., avg: 3.673 ms.
Best time for 1280K FFT length: 4.194 ms., avg: 4.492 ms.
Best time for 1536K FFT length: 4.208 ms., avg: 4.904 ms.
Best time for 2048K FFT length: 4.985 ms., avg: 5.005 ms.
Best time for 2560K FFT length: 6.359 ms., avg: 6.488 ms.
Best time for 3072K FFT length: 7.679 ms., avg: 8.297 ms.
Best time for 4096K FFT length: 10.776 ms., avg: 11.733 ms.
Best time for 5120K FFT length: 13.139 ms., avg: 15.356 ms.
Best time for 6144K FFT length: 16.029 ms., avg: 16.111 ms.
Best time for 8192K FFT length: 21.620 ms., avg: 22.925 ms.
Best time for 61 bit trial factors: 1.990 ms.
Best time for 62 bit trial factors: 1.991 ms.
Best time for 63 bit trial factors: 3.223 ms.
Best time for 64 bit trial factors: 3.235 ms.
Best time for 65 bit trial factors: 3.271 ms.
Best time for 66 bit trial factors: 3.251 ms.
Best time for 67 bit trial factors: 3.252 ms.
Best time for 75 bit trial factors: 3.413 ms.
Best time for 76 bit trial factors: 3.412 ms.
Best time for 77 bit trial factors: 3.407 ms.

Note, for both the detected CPU speed was off a little, but monitored CPU speed varied between 4900 and 4910 during both tests, and 90% of the time at 4902.5

Dubslow · 2011-12-16, 22:37

It still seems that small FFT's do not parallelize well in AVX.

(Someone else posted the same in the v27 thread.)

Paulie · 2012-01-10, 06:01

[Main thread Jan 9 23:50] Mersenne number primality test program version 26.6
[Main thread Jan 9 23:50] Optimizing for CPU architecture: Core i3/i5/i7, L2 cache size: 256 KB, L3 cache size: 8 MB

Code:

[Jan 9 23:52] Your timings will be written to the results.txt file.
[Jan 9 23:52] Compare your results to other computers at http://www.mersenne.org/report_benchmarks
[Jan 9 23:52] Timing 58 iterations of 768K FFT length.  Best time: 7.172 ms., avg time: 7.306 ms.
[Jan 9 23:52] Timing 50 iterations of 896K FFT length.  Best time: 8.694 ms., avg time: 9.086 ms.
[Jan 9 23:52] Timing 43 iterations of 1024K FFT length.  Best time: 9.624 ms., avg time: 10.360 ms.
[Jan 9 23:52] Timing 35 iterations of 1280K FFT length.  Best time: 12.776 ms., avg time: 14.163 ms.
[Jan 9 23:52] Timing 29 iterations of 1536K FFT length.  Best time: 15.193 ms., avg time: 15.392 ms.
[Jan 9 23:52] Timing 25 iterations of 1792K FFT length.  Best time: 18.431 ms., avg time: 18.734 ms.
[Jan 9 23:52] Timing 21 iterations of 2048K FFT length.  Best time: 20.503 ms., avg time: 21.560 ms.
[Jan 9 23:52] Timing 17 iterations of 2560K FFT length.  Best time: 25.894 ms., avg time: 27.357 ms.
[Jan 9 23:52] Timing 14 iterations of 3072K FFT length.  Best time: 32.464 ms., avg time: 36.067 ms.
[Jan 9 23:52] Timing 12 iterations of 3584K FFT length.  Best time: 38.640 ms., avg time: 38.970 ms.
[Jan 9 23:52] Timing 10 iterations of 4096K FFT length.  Best time: 46.608 ms., avg time: 49.861 ms.
[Jan 9 23:52] Timing 10 iterations of 5120K FFT length.  Best time: 55.650 ms., avg time: 60.454 ms.
[Jan 9 23:52] Timing 10 iterations of 6144K FFT length.  Best time: 69.431 ms., avg time: 76.871 ms.
[Jan 9 23:52] Timing 10 iterations of 7168K FFT length.  Best time: 95.081 ms., avg time: 97.739 ms.
[Jan 9 23:52] Timing 10 iterations of 8192K FFT length.  Best time: 102.999 ms., avg time: 105.876 ms.
[Jan 9 23:52] Timing FFTs using 2 threads on 1 physical CPUs.
[Jan 9 23:52] Timing 58 iterations of 768K FFT length.  Best time: 5.305 ms., avg time: 5.679 ms.
[Jan 9 23:52] Timing 50 iterations of 896K FFT length.  Best time: 6.326 ms., avg time: 6.828 ms.
[Jan 9 23:52] Timing 43 iterations of 1024K FFT length.  Best time: 7.162 ms., avg time: 7.425 ms.
[Jan 9 23:52] Timing 35 iterations of 1280K FFT length.  Best time: 9.100 ms., avg time: 9.950 ms.
[Jan 9 23:52] Timing 29 iterations of 1536K FFT length.  Best time: 11.225 ms., avg time: 11.517 ms.
[Jan 9 23:52] Timing 25 iterations of 1792K FFT length.  Best time: 13.221 ms., avg time: 13.755 ms.
[Jan 9 23:52] Timing 21 iterations of 2048K FFT length.  Best time: 11.341 ms., avg time: 15.243 ms.
[Jan 9 23:53] Timing 17 iterations of 2560K FFT length.  Best time: 18.807 ms., avg time: 20.219 ms.
[Jan 9 23:53] Timing 14 iterations of 3072K FFT length.  Best time: 23.177 ms., avg time: 24.234 ms.
[Jan 9 23:53] Timing 12 iterations of 3584K FFT length.  Best time: 27.807 ms., avg time: 29.115 ms.
[Jan 9 23:53] Timing 10 iterations of 4096K FFT length.  Best time: 26.987 ms., avg time: 31.704 ms.
[Jan 9 23:53] Timing 10 iterations of 5120K FFT length.  Best time: 40.554 ms., avg time: 42.378 ms.
[Jan 9 23:53] Timing 10 iterations of 6144K FFT length.  Best time: 48.297 ms., avg time: 51.667 ms.
[Jan 9 23:53] Timing 10 iterations of 7168K FFT length.  Best time: 60.508 ms., avg time: 63.019 ms.
[Jan 9 23:53] Timing 10 iterations of 8192K FFT length.  Best time: 67.488 ms., avg time: 69.501 ms.
[Jan 9 23:53] Timing FFTs using 4 threads on 2 physical CPUs.
[Jan 9 23:53] Timing 58 iterations of 768K FFT length.  Best time: 3.939 ms., avg time: 4.157 ms.
[Jan 9 23:53] Timing 50 iterations of 896K FFT length.  Best time: 3.299 ms., avg time: 4.208 ms.
[Jan 9 23:53] Timing 43 iterations of 1024K FFT length.  Best time: 4.080 ms., avg time: 5.663 ms.
[Jan 9 23:53] Timing 35 iterations of 1280K FFT length.  Best time: 5.996 ms., avg time: 6.636 ms.
[Jan 9 23:53] Timing 29 iterations of 1536K FFT length.  Best time: 7.098 ms., avg time: 7.413 ms.
[Jan 9 23:53] Timing 25 iterations of 1792K FFT length.  Best time: 8.624 ms., avg time: 9.016 ms.
[Jan 9 23:53] Timing 21 iterations of 2048K FFT length.  Best time: 7.957 ms., avg time: 9.731 ms.
[Jan 9 23:53] Timing 17 iterations of 2560K FFT length.  Best time: 11.776 ms., avg time: 12.339 ms.
[Jan 9 23:53] Timing 14 iterations of 3072K FFT length.  Best time: 13.920 ms., avg time: 15.042 ms.
[Jan 9 23:53] Timing 12 iterations of 3584K FFT length.  Best time: 14.955 ms., avg time: 17.700 ms.
[Jan 9 23:53] Timing 10 iterations of 4096K FFT length.  Best time: 19.732 ms., avg time: 20.394 ms.
[Jan 9 23:53] Timing 10 iterations of 5120K FFT length.  Best time: 23.945 ms., avg time: 25.534 ms.
[Jan 9 23:53] Timing 10 iterations of 6144K FFT length.  Best time: 30.103 ms., avg time: 32.071 ms.
[Jan 9 23:53] Timing 10 iterations of 7168K FFT length.  Best time: 34.029 ms., avg time: 38.118 ms.
[Jan 9 23:53] Timing 10 iterations of 8192K FFT length.  Best time: 40.634 ms., avg time: 42.347 ms.
[Jan 9 23:53] Timing FFTs using 6 threads on 3 physical CPUs.
[Jan 9 23:53] Timing 58 iterations of 768K FFT length.  Best time: 3.357 ms., avg time: 3.616 ms.
[Jan 9 23:53] Timing 50 iterations of 896K FFT length.  Best time: 3.013 ms., avg time: 3.272 ms.
[Jan 9 23:53] Timing 43 iterations of 1024K FFT length.  Best time: 4.531 ms., avg time: 4.980 ms.
[Jan 9 23:53] Timing 35 iterations of 1280K FFT length.  Best time: 5.465 ms., avg time: 5.774 ms.
[Jan 9 23:53] Timing 29 iterations of 1536K FFT length.  Best time: 6.024 ms., avg time: 6.280 ms.
[Jan 9 23:54] Timing 25 iterations of 1792K FFT length.  Best time: 6.344 ms., avg time: 6.730 ms.
[Jan 9 23:54] Timing 21 iterations of 2048K FFT length.  Best time: 7.114 ms., avg time: 7.568 ms.
[Jan 9 23:54] Timing 17 iterations of 2560K FFT length.  Best time: 9.498 ms., avg time: 9.915 ms.
[Jan 9 23:54] Timing 14 iterations of 3072K FFT length.  Best time: 10.816 ms., avg time: 11.386 ms.
[Jan 9 23:54] Timing 12 iterations of 3584K FFT length.  Best time: 12.979 ms., avg time: 13.569 ms.
[Jan 9 23:54] Timing 10 iterations of 4096K FFT length.  Best time: 14.447 ms., avg time: 14.944 ms.
[Jan 9 23:54] Timing 10 iterations of 5120K FFT length.  Best time: 18.407 ms., avg time: 18.997 ms.
[Jan 9 23:54] Timing 10 iterations of 6144K FFT length.  Best time: 22.638 ms., avg time: 23.380 ms.
[Jan 9 23:54] Timing 10 iterations of 7168K FFT length.  Best time: 30.555 ms., avg time: 31.647 ms.
[Jan 9 23:54] Timing 10 iterations of 8192K FFT length.  Best time: 30.210 ms., avg time: 31.288 ms.
[Jan 9 23:54] Timing FFTs using 8 threads on 4 physical CPUs.
[Jan 9 23:54] Timing 58 iterations of 768K FFT length.  Best time: 2.599 ms., avg time: 2.671 ms.
[Jan 9 23:54] Timing 50 iterations of 896K FFT length.  Best time: 2.869 ms., avg time: 3.117 ms.
[Jan 9 23:54] Timing 43 iterations of 1024K FFT length.  Best time: 3.513 ms., avg time: 3.610 ms.
[Jan 9 23:54] Timing 35 iterations of 1280K FFT length.  Best time: 4.120 ms., avg time: 4.391 ms.
[Jan 9 23:54] Timing 29 iterations of 1536K FFT length.  Best time: 5.371 ms., avg time: 5.456 ms.
[Jan 9 23:54] Timing 25 iterations of 1792K FFT length.  Best time: 5.923 ms., avg time: 6.113 ms.
[Jan 9 23:54] Timing 21 iterations of 2048K FFT length.  Best time: 6.604 ms., avg time: 6.843 ms.
[Jan 9 23:54] Timing 17 iterations of 2560K FFT length.  Best time: 8.557 ms., avg time: 8.677 ms.
[Jan 9 23:54] Timing 14 iterations of 3072K FFT length.  Best time: 10.306 ms., avg time: 10.469 ms.
[Jan 9 23:54] Timing 12 iterations of 3584K FFT length.  Best time: 12.233 ms., avg time: 12.539 ms.
[Jan 9 23:54] Timing 10 iterations of 4096K FFT length.  Best time: 13.585 ms., avg time: 13.860 ms.
[Jan 9 23:54] Timing 10 iterations of 5120K FFT length.  Best time: 17.318 ms., avg time: 17.604 ms.
[Jan 9 23:54] Timing 10 iterations of 6144K FFT length.  Best time: 21.106 ms., avg time: 21.584 ms.
[Jan 9 23:54] Timing 10 iterations of 7168K FFT length.  Best time: 34.146 ms., avg time: 34.428 ms.
[Jan 9 23:54] Timing 10 iterations of 8192K FFT length.  Best time: 28.237 ms., avg time: 28.692 ms.
[Jan 9 23:54] Timing trial factoring of M35000011 with 61 bit length factors.  Best time: 2.257 ms.
[Jan 9 23:54] Timing trial factoring of M35000011 with 62 bit length factors.  Best time: 2.246 ms.
[Jan 9 23:54] Timing trial factoring of M35000011 with 63 bit length factors.  Best time: 2.510 ms.
[Jan 9 23:54] Timing trial factoring of M35000011 with 64 bit length factors.  Best time: 2.672 ms.
[Jan 9 23:54] Timing trial factoring of M35000011 with 65 bit length factors.  Best time: 3.046 ms.
[Jan 9 23:54] Timing trial factoring of M35000011 with 66 bit length factors.  Best time: 3.652 ms.
[Jan 9 23:54] Timing trial factoring of M35000011 with 67 bit length factors.  Best time: 3.569 ms.
[Jan 9 23:54] Timing trial factoring of M35000011 with 75 bit length factors.  Best time: 3.525 ms.
[Jan 9 23:54] Timing trial factoring of M35000011 with 76 bit length factors.  Best time: 3.481 ms.
[Jan 9 23:54] Timing trial factoring of M35000011 with 77 bit length factors.  Best time: 3.451 ms.
[Jan 9 23:54] Benchmark complete.

petrw1 · 2012-06-12, 14:57

This may not exactly fit in this thread but I didn't want to start yet another thread for a minor topic.

My i5-750 is OC'd at 3.2 Ghz.
It has been doing DC for all of 2012 so far.

The per iteration timings in the 1344K FFT range have been a very consistent 20Ms on all 4 cores.

Just recently it started a 1440K FFT on 1 core and 1344K FFT on the other 3 cores.

The 1440K FFT timing went to 20.25Ms; about a 1.25% increase as expected.
The timings on the 3 1344K FFT cores DROPPED to 19.8; 19.55 and 19.45; decreases of about 1%, 2.25% and 2.75%.

Could I be correct in extrapolating from this that LL/DC performance is slightly better if you mix FFT sizes across the cores.

kracker · 2012-06-15, 13:41

Benchmark for my i5-3450M (mobile)

Code:

Intel(R) Core(TM) i5-2450M CPU @ 2.50GHz
CPU speed: 2874.46 MHz, 2 hyperthreaded cores
CPU features: Prefetch, SSE, SSE2, SSE4, AVX
L1 cache size: 32 KB
L2 cache size: 256 KB, L3 cache size: 3 MB
L1 cache line size: 64 bytes
L2 cache line size: 64 bytes
TLBS: 64
Prime95 64-bit version 27.7, RdtscTiming=1
Best time for 768K FFT length: 5.692 ms., avg: 5.765 ms.
Best time for 896K FFT length: 6.833 ms., avg: 7.028 ms.
Best time for 1024K FFT length: 7.709 ms., avg: 7.966 ms.
Best time for 1280K FFT length: 9.766 ms., avg: 9.941 ms.
Best time for 1536K FFT length: 11.941 ms., avg: 12.060 ms.
Best time for 1792K FFT length: 14.206 ms., avg: 14.720 ms.
Best time for 2048K FFT length: 15.908 ms., avg: 16.055 ms.
Best time for 2560K FFT length: 20.135 ms., avg: 20.244 ms.
Best time for 3072K FFT length: 37.234 ms., avg: 42.681 ms.
Best time for 3584K FFT length: 45.562 ms., avg: 50.985 ms.
Best time for 4096K FFT length: 33.819 ms., avg: 36.377 ms.
Best time for 5120K FFT length: 43.823 ms., avg: 44.184 ms.
Best time for 6144K FFT length: 52.934 ms., avg: 54.029 ms.
Best time for 7168K FFT length: 63.753 ms., avg: 64.011 ms.
Best time for 8192K FFT length: 73.712 ms., avg: 74.059 ms.
Timing FFTs using 2 threads on 1 physical CPUs.
Best time for 768K FFT length: 5.861 ms., avg: 6.003 ms.
Best time for 896K FFT length: 6.902 ms., avg: 7.095 ms.
Best time for 1024K FFT length: 7.889 ms., avg: 11.216 ms.
Best time for 1280K FFT length: 10.082 ms., avg: 10.336 ms.
Best time for 1536K FFT length: 12.520 ms., avg: 16.615 ms.
Best time for 1792K FFT length: 14.564 ms., avg: 14.800 ms.
Best time for 2048K FFT length: 16.654 ms., avg: 16.965 ms.
Best time for 2560K FFT length: 21.274 ms., avg: 21.486 ms.
Best time for 3072K FFT length: 26.087 ms., avg: 26.536 ms.
Best time for 3584K FFT length: 31.041 ms., avg: 31.446 ms.
Best time for 4096K FFT length: 35.178 ms., avg: 35.576 ms.
Best time for 5120K FFT length: 46.825 ms., avg: 47.295 ms.
Best time for 6144K FFT length: 58.684 ms., avg: 59.951 ms.
Best time for 7168K FFT length: 70.104 ms., avg: 71.844 ms.
Best time for 8192K FFT length: 80.488 ms., avg: 82.946 ms.
Timing FFTs using 4 threads on 2 physical CPUs.
Best time for 768K FFT length: 3.345 ms., avg: 3.440 ms.
Best time for 896K FFT length: 3.955 ms., avg: 4.072 ms.
Best time for 1024K FFT length: 4.536 ms., avg: 4.734 ms.
Best time for 1280K FFT length: 5.643 ms., avg: 5.817 ms.
Best time for 1536K FFT length: 7.056 ms., avg: 7.344 ms.
Best time for 1792K FFT length: 8.214 ms., avg: 8.452 ms.
Best time for 2048K FFT length: 9.299 ms., avg: 9.526 ms.
Best time for 2560K FFT length: 11.952 ms., avg: 12.175 ms.
Best time for 3072K FFT length: 14.299 ms., avg: 14.791 ms.
Best time for 3584K FFT length: 17.204 ms., avg: 17.430 ms.
Best time for 4096K FFT length: 19.448 ms., avg: 19.813 ms.
Best time for 5120K FFT length: 25.892 ms., avg: 26.440 ms.
Best time for 6144K FFT length: 31.723 ms., avg: 32.643 ms.
Best time for 7168K FFT length: 37.808 ms., avg: 39.033 ms.
Best time for 8192K FFT length: 44.753 ms., avg: 45.206 ms.
Best time for 61 bit trial factors: 2.672 ms.
Best time for 62 bit trial factors: 2.691 ms.
Best time for 63 bit trial factors: 3.049 ms.
Best time for 64 bit trial factors: 3.160 ms.
Best time for 65 bit trial factors: 3.699 ms.
Best time for 66 bit trial factors: 4.378 ms.
Best time for 67 bit trial factors: 4.322 ms.
Best time for 75 bit trial factors: 4.207 ms.
Best time for 76 bit trial factors: 4.203 ms.
Best time for 77 bit trial factors: 4.203 ms.

kracker · 2012-07-01, 16:46

A8-3850 benchmark. Slow, I know.

Code:

Compare your results to other computers at http://www.mersenne.org/report_benchmarks
AMD A8-3850 APU with Radeon(tm) HD Graphics
CPU speed: 2894.58 MHz, 4 cores
CPU features: 3DNow!, SSE, SSE2
L1 cache size: 64 KB
L2 cache size: 1 MB
L1 cache line size: 64 bytes
L2 cache line size: 64 bytes
L1 TLBS: 48
L2 TLBS: 1024
Prime95 64-bit version 27.7, RdtscTiming=1
Best time for 768K FFT length: 12.236 ms., avg: 13.571 ms.
Best time for 896K FFT length: 14.965 ms., avg: 16.765 ms.
Best time for 1024K FFT length: 16.129 ms., avg: 17.553 ms.
Best time for 1280K FFT length: 20.826 ms., avg: 22.737 ms.
Best time for 1536K FFT length: 25.569 ms., avg: 26.787 ms.
Best time for 1792K FFT length: 31.100 ms., avg: 32.634 ms.
Best time for 2048K FFT length: 33.968 ms., avg: 35.616 ms.
Best time for 2560K FFT length: 43.661 ms., avg: 43.751 ms.
Best time for 3072K FFT length: 53.810 ms., avg: 53.917 ms.
Best time for 3584K FFT length: 64.816 ms., avg: 64.934 ms.
Best time for 4096K FFT length: 71.589 ms., avg: 71.860 ms.
Best time for 5120K FFT length: 91.189 ms., avg: 91.404 ms.
Best time for 6144K FFT length: 113.710 ms., avg: 116.038 ms.
Best time for 7168K FFT length: 137.195 ms., avg: 137.645 ms.
Best time for 8192K FFT length: 154.629 ms., avg: 155.025 ms.
Timing FFTs using 2 threads.
Best time for 768K FFT length: 6.779 ms., avg: 7.150 ms.
Best time for 896K FFT length: 8.183 ms., avg: 8.415 ms.
Best time for 1024K FFT length: 9.012 ms., avg: 9.338 ms.
Best time for 1280K FFT length: 11.416 ms., avg: 11.976 ms.
Best time for 1536K FFT length: 14.034 ms., avg: 14.361 ms.
Best time for 1792K FFT length: 17.012 ms., avg: 17.173 ms.
Best time for 2048K FFT length: 18.532 ms., avg: 19.010 ms.
Best time for 2560K FFT length: 23.874 ms., avg: 26.422 ms.
Best time for 3072K FFT length: 29.229 ms., avg: 30.394 ms.
Best time for 3584K FFT length: 35.178 ms., avg: 36.362 ms.
Best time for 4096K FFT length: 38.816 ms., avg: 41.027 ms.
Best time for 5120K FFT length: 50.805 ms., avg: 56.381 ms.
Best time for 6144K FFT length: 62.464 ms., avg: 63.677 ms.
Best time for 7168K FFT length: 74.902 ms., avg: 77.450 ms.
Best time for 8192K FFT length: 85.885 ms., avg: 91.092 ms.
Timing FFTs using 3 threads.
Best time for 768K FFT length: 5.107 ms., avg: 5.796 ms.
Best time for 896K FFT length: 6.096 ms., avg: 6.735 ms.
Best time for 1024K FFT length: 7.089 ms., avg: 8.646 ms.
Best time for 1280K FFT length: 8.761 ms., avg: 10.176 ms.
Best time for 1536K FFT length: 10.726 ms., avg: 11.558 ms.
Best time for 1792K FFT length: 12.647 ms., avg: 13.416 ms.
Best time for 2048K FFT length: 14.195 ms., avg: 15.441 ms.
Best time for 2560K FFT length: 17.852 ms., avg: 19.652 ms.
Best time for 3072K FFT length: 21.842 ms., avg: 23.374 ms.
Best time for 3584K FFT length: 26.609 ms., avg: 30.221 ms.
Best time for 4096K FFT length: 31.230 ms., avg: 34.899 ms.
Best time for 5120K FFT length: 37.097 ms., avg: 43.213 ms.
Best time for 6144K FFT length: 50.667 ms., avg: 77.316 ms.
Best time for 7168K FFT length: 57.295 ms., avg: 60.101 ms.
Best time for 8192K FFT length: 60.651 ms., avg: 64.744 ms.
Timing FFTs using 4 threads.
Best time for 768K FFT length: 4.472 ms., avg: 4.755 ms.
Best time for 896K FFT length: 5.323 ms., avg: 5.766 ms.
Best time for 1024K FFT length: 6.600 ms., avg: 7.144 ms.
Best time for 1280K FFT length: 8.165 ms., avg: 8.848 ms.
Best time for 1536K FFT length: 9.416 ms., avg: 10.528 ms.
Best time for 1792K FFT length: 11.052 ms., avg: 11.883 ms.
Best time for 2048K FFT length: 12.631 ms., avg: 14.123 ms.
Best time for 2560K FFT length: 16.836 ms., avg: 20.168 ms.
Best time for 3072K FFT length: 20.746 ms., avg: 24.202 ms.
Best time for 3584K FFT length: 23.041 ms., avg: 25.604 ms.
[Sun Jul 01 09:43:55 2012]
Best time for 4096K FFT length: 26.068 ms., avg: 29.135 ms.
Best time for 5120K FFT length: 33.852 ms., avg: 39.417 ms.
Best time for 6144K FFT length: 42.212 ms., avg: 46.086 ms.
Best time for 7168K FFT length: 50.169 ms., avg: 58.227 ms.
Best time for 8192K FFT length: 55.377 ms., avg: 60.012 ms.
Best time for 61 bit trial factors: 2.667 ms.
Best time for 62 bit trial factors: 2.689 ms.
Best time for 63 bit trial factors: 3.090 ms.
Best time for 64 bit trial factors: 3.590 ms.
Best time for 65 bit trial factors: 4.435 ms.
Best time for 66 bit trial factors: 5.239 ms.
Best time for 67 bit trial factors: 5.199 ms.
Best time for 75 bit trial factors: 5.088 ms.
Best time for 76 bit trial factors: 5.064 ms.
Best time for 77 bit trial factors: 5.084 ms.

Sprocket · 2012-09-30, 04:15

Severe lack of Ivy Bridge + Linux in this thread. Its not a big improvement per-clock over Sandy Bridge, but you don't need exotic cooling either since the TDP is 20-30W less than equivalent Sandy Bridge!

Intel Core i7-3770 @ 3.4GHz, running on Ubuntu Linux 12.04 64-bit. Total system memory is 16GB.

Code:

Intel(R) Core(TM) i7-3770 CPU @ 3.40GHz
CPU speed: 3400.05 MHz, 4 hyperthreaded cores
CPU features: Prefetch, SSE, SSE2, SSE4, AVX
L1 cache size: 32 KB
L2 cache size: 256 KB, L3 cache size: 8 MB
L1 cache line size: 64 bytes
L2 cache line size: 64 bytes
TLBS: 64
Prime95 64-bit version 27.7, RdtscTiming=1
Time FFTlen=768K, Type=3, Arch=3, Pass1=512, Pass2=1536, clm=2: 4.261 ms., 4.271 ms.
Time FFTlen=896K, Type=3, Arch=3, Pass1=448, Pass2=2048, clm=2: 5.136 ms., 5.150 ms.
Time FFTlen=1024K, Type=3, Arch=3, Pass1=512, Pass2=2048, clm=2: 5.784 ms., 5.832 ms.
Time FFTlen=1280K, Type=3, Arch=3, Pass1=512, Pass2=2560, clm=2: 7.372 ms., 7.389 ms.
Time FFTlen=1536K, Type=3, Arch=3, Pass1=512, Pass2=3072, clm=2: 9.084 ms., 9.104 ms.
Time FFTlen=1792K, Type=3, Arch=3, Pass1=448, Pass2=4096, clm=2: 11.008 ms., 11.035 ms.
Time FFTlen=2048K, Type=3, Arch=3, Pass1=512, Pass2=4096, clm=2: 12.277 ms., 12.290 ms.
Time FFTlen=2560K, Type=3, Arch=3, Pass1=512, Pass2=5120, clm=2: 15.511 ms., 15.718 ms.
Time FFTlen=3072K, Type=3, Arch=3, Pass1=512, Pass2=6144, clm=2: 19.389 ms., 19.405 ms.
Time FFTlen=3584K, Type=3, Arch=3, Pass1=448, Pass2=8192, clm=2: 23.521 ms., 23.534 ms.
Time FFTlen=4096K, Type=3, Arch=3, Pass1=512, Pass2=8192, clm=2: 26.150 ms., 26.198 ms.
Time FFTlen=5120K, Type=3, Arch=3, Pass1=640, Pass2=8192, clm=2: 34.144 ms., 34.164 ms.
Time FFTlen=6144K, Type=3, Arch=3, Pass1=768, Pass2=8192, clm=2: 41.202 ms., 41.225 ms.
Time FFTlen=7168K, Type=3, Arch=3, Pass1=896, Pass2=8192, clm=2: 50.692 ms., 50.747 ms.
Time FFTlen=8192K, Type=3, Arch=3, Pass1=512, Pass2=16384, clm=2: 58.568 ms., 59.484 ms.
Timing FFTs using 2 threads on 1 physical CPUs.
Time FFTlen=768K, Type=3, Arch=3, Pass1=512, Pass2=1536, clm=2: 4.355 ms., 4.397 ms.
Time FFTlen=896K, Type=3, Arch=3, Pass1=448, Pass2=2048, clm=2: 5.076 ms., 5.104 ms.
Time FFTlen=1024K, Type=3, Arch=3, Pass1=512, Pass2=2048, clm=2: 5.882 ms., 6.117 ms.
Time FFTlen=1280K, Type=3, Arch=3, Pass1=512, Pass2=2560, clm=2: 7.588 ms., 7.817 ms.
Time FFTlen=1536K, Type=3, Arch=3, Pass1=512, Pass2=3072, clm=2: 9.357 ms., 9.534 ms.
Time FFTlen=1792K, Type=3, Arch=3, Pass1=448, Pass2=4096, clm=2: 11.069 ms., 11.154 ms.
Time FFTlen=2048K, Type=3, Arch=3, Pass1=512, Pass2=4096, clm=2: 12.660 ms., 12.778 ms.
Time FFTlen=2560K, Type=3, Arch=3, Pass1=512, Pass2=5120, clm=2: 16.057 ms., 16.209 ms.
Time FFTlen=3072K, Type=3, Arch=3, Pass1=512, Pass2=6144, clm=2: 19.825 ms., 19.966 ms.
Time FFTlen=3584K, Type=3, Arch=3, Pass1=448, Pass2=8192, clm=2: 23.547 ms., 23.677 ms.
Time FFTlen=4096K, Type=3, Arch=3, Pass1=512, Pass2=8192, clm=2: 26.762 ms., 26.993 ms.
Time FFTlen=5120K, Type=3, Arch=3, Pass1=640, Pass2=8192, clm=2: 36.077 ms., 38.463 ms.
Time FFTlen=6144K, Type=3, Arch=3, Pass1=768, Pass2=8192, clm=2: 44.905 ms., 45.841 ms.
Time FFTlen=7168K, Type=3, Arch=3, Pass1=896, Pass2=8192, clm=2: 53.921 ms., 55.234 ms.
Time FFTlen=8192K, Type=3, Arch=3, Pass1=512, Pass2=16384, clm=2: 63.335 ms., 63.984 ms.
Timing FFTs using 4 threads on 2 physical CPUs.
Time FFTlen=768K, Type=3, Arch=3, Pass1=512, Pass2=1536, clm=2: 2.225 ms., 2.241 ms.
Time FFTlen=896K, Type=3, Arch=3, Pass1=448, Pass2=2048, clm=2: 2.592 ms., 2.630 ms.
Time FFTlen=1024K, Type=3, Arch=3, Pass1=512, Pass2=2048, clm=2: 3.020 ms., 3.062 ms.
Time FFTlen=1280K, Type=3, Arch=3, Pass1=512, Pass2=2560, clm=2: 3.859 ms., 4.339 ms.
Time FFTlen=1536K, Type=3, Arch=3, Pass1=512, Pass2=3072, clm=2: 4.916 ms., 5.016 ms.
Time FFTlen=1792K, Type=3, Arch=3, Pass1=448, Pass2=4096, clm=2: 5.710 ms., 5.774 ms.
Time FFTlen=2048K, Type=3, Arch=3, Pass1=512, Pass2=4096, clm=2: 6.569 ms., 6.728 ms.
Time FFTlen=2560K, Type=3, Arch=3, Pass1=512, Pass2=5120, clm=2: 8.332 ms., 8.412 ms.
Time FFTlen=3072K, Type=3, Arch=3, Pass1=512, Pass2=6144, clm=2: 10.232 ms., 10.386 ms.
Time FFTlen=3584K, Type=3, Arch=3, Pass1=448, Pass2=8192, clm=2: 12.188 ms., 12.223 ms.
Time FFTlen=4096K, Type=3, Arch=3, Pass1=512, Pass2=8192, clm=2: 13.812 ms., 14.908 ms.
Time FFTlen=5120K, Type=3, Arch=3, Pass1=640, Pass2=8192, clm=2: 18.397 ms., 19.777 ms.
Time FFTlen=6144K, Type=3, Arch=3, Pass1=768, Pass2=8192, clm=2: 23.013 ms., 23.546 ms.
Time FFTlen=7168K, Type=3, Arch=3, Pass1=896, Pass2=8192, clm=2: 28.781 ms., 29.174 ms.
Time FFTlen=8192K, Type=3, Arch=3, Pass1=512, Pass2=16384, clm=2: 31.712 ms., 31.900 ms.
Timing FFTs using 6 threads on 3 physical CPUs.
Time FFTlen=768K, Type=3, Arch=3, Pass1=512, Pass2=1536, clm=2: 1.529 ms., 1.552 ms.
Time FFTlen=896K, Type=3, Arch=3, Pass1=448, Pass2=2048, clm=2: 1.780 ms., 1.807 ms.
Time FFTlen=1024K, Type=3, Arch=3, Pass1=512, Pass2=2048, clm=2: 2.076 ms., 2.109 ms.
Time FFTlen=1280K, Type=3, Arch=3, Pass1=512, Pass2=2560, clm=2: 2.702 ms., 2.755 ms.
Time FFTlen=1536K, Type=3, Arch=3, Pass1=512, Pass2=3072, clm=2: 3.460 ms., 3.977 ms.
Time FFTlen=1792K, Type=3, Arch=3, Pass1=448, Pass2=4096, clm=2: 4.121 ms., 4.167 ms.
Time FFTlen=2048K, Type=3, Arch=3, Pass1=512, Pass2=4096, clm=2: 4.780 ms., 4.845 ms.
Time FFTlen=2560K, Type=3, Arch=3, Pass1=512, Pass2=5120, clm=2: 6.135 ms., 6.184 ms.
Time FFTlen=3072K, Type=3, Arch=3, Pass1=512, Pass2=6144, clm=2: 7.510 ms., 7.606 ms.
Time FFTlen=3584K, Type=3, Arch=3, Pass1=448, Pass2=8192, clm=2: 8.984 ms., 9.050 ms.
Time FFTlen=4096K, Type=3, Arch=3, Pass1=512, Pass2=8192, clm=2: 10.247 ms., 10.355 ms.
Time FFTlen=5120K, Type=3, Arch=3, Pass1=640, Pass2=8192, clm=2: 13.393 ms., 13.727 ms.
Time FFTlen=6144K, Type=3, Arch=3, Pass1=768, Pass2=8192, clm=2: 16.719 ms., 17.999 ms.
Time FFTlen=7168K, Type=3, Arch=3, Pass1=896, Pass2=8192, clm=2: 21.018 ms., 21.525 ms.
Time FFTlen=8192K, Type=3, Arch=3, Pass1=512, Pass2=16384, clm=2: 24.616 ms., 24.695 ms.
Timing FFTs using 8 threads on 4 physical CPUs.
Time FFTlen=768K, Type=3, Arch=3, Pass1=512, Pass2=1536, clm=2: 1.182 ms., 1.453 ms.
Time FFTlen=896K, Type=3, Arch=3, Pass1=448, Pass2=2048, clm=2: 1.411 ms., 1.758 ms.
Time FFTlen=1024K, Type=3, Arch=3, Pass1=512, Pass2=2048, clm=2: 1.673 ms., 1.717 ms.
Time FFTlen=1280K, Type=3, Arch=3, Pass1=512, Pass2=2560, clm=2: 2.296 ms., 2.337 ms.
Time FFTlen=1536K, Type=3, Arch=3, Pass1=512, Pass2=3072, clm=2: 3.020 ms., 3.057 ms.
Time FFTlen=1792K, Type=3, Arch=3, Pass1=448, Pass2=4096, clm=2: 3.639 ms., 4.194 ms.
Time FFTlen=2048K, Type=3, Arch=3, Pass1=512, Pass2=4096, clm=2: 4.380 ms., 4.440 ms.
Time FFTlen=2560K, Type=3, Arch=3, Pass1=512, Pass2=5120, clm=2: 5.648 ms., 5.686 ms.
Time FFTlen=3072K, Type=3, Arch=3, Pass1=512, Pass2=6144, clm=2: 6.975 ms., 7.030 ms.
Time FFTlen=3584K, Type=3, Arch=3, Pass1=448, Pass2=8192, clm=2: 8.263 ms., 8.313 ms.
Time FFTlen=4096K, Type=3, Arch=3, Pass1=512, Pass2=8192, clm=2: 9.541 ms., 9.572 ms.
Time FFTlen=5120K, Type=3, Arch=3, Pass1=640, Pass2=8192, clm=2: 12.129 ms., 12.190 ms.
Time FFTlen=6144K, Type=3, Arch=3, Pass1=768, Pass2=8192, clm=2: 14.721 ms., 14.935 ms.
Time FFTlen=7168K, Type=3, Arch=3, Pass1=896, Pass2=8192, clm=2: 18.592 ms., 18.732 ms.
Time FFTlen=8192K, Type=3, Arch=3, Pass1=512, Pass2=16384, clm=2: 22.041 ms., 22.218 ms.
Best time for 61 bit trial factors: 1.895 ms.
Best time for 62 bit trial factors: 1.964 ms.
Best time for 63 bit trial factors: 2.215 ms.
Best time for 64 bit trial factors: 2.244 ms.
Best time for 65 bit trial factors: 2.659 ms.
Best time for 66 bit trial factors: 3.081 ms.
Best time for 67 bit trial factors: 3.055 ms.
Best time for 75 bit trial factors: 2.975 ms.
Best time for 76 bit trial factors: 2.965 ms.
Best time for 77 bit trial factors: 2.967 ms.

henryzz · 2012-10-06, 12:39

It looks to me like you might get better results if you turn off hyperthreading (at least in prime95. Everything else will be more responsive to normal usage if it is on while prime95 is running).
2 threads on one physical core is slower than 1 thread in your benchmark.

dbaugh · 2012-10-28, 05:45

[Sat Oct 27 22:11:02 2012]
Compare your results to other computers at http://www.mersenne.org/report_benchmarks
Intel(R) Core(TM) i7-3960X CPU @ 3.30GHz
CPU speed: 3990.97 MHz, 6 hyperthreaded cores
CPU features: Prefetch, SSE, SSE2, SSE4, AVX
L1 cache size: 32 KB
L2 cache size: 256 KB, L3 cache size: 15 MB
L1 cache line size: 64 bytes
L2 cache line size: 64 bytes
TLBS: 64
Prime95 64-bit version 27.7, RdtscTiming=1
Best time for 768K FFT length: 3.930 ms., avg: 3.949 ms.
Best time for 896K FFT length: 4.658 ms., avg: 4.667 ms.
Best time for 1024K FFT length: 5.276 ms., avg: 5.287 ms.
Best time for 1280K FFT length: 6.744 ms., avg: 6.760 ms.
Best time for 1536K FFT length: 8.267 ms., avg: 8.290 ms.
Best time for 1792K FFT length: 9.898 ms., avg: 9.937 ms.
Best time for 2048K FFT length: 11.163 ms., avg: 11.196 ms.
Best time for 2560K FFT length: 14.344 ms., avg: 14.386 ms.
Best time for 3072K FFT length: 17.935 ms., avg: 17.960 ms.
Best time for 3584K FFT length: 21.781 ms., avg: 21.813 ms.
Best time for 4096K FFT length: 24.356 ms., avg: 24.418 ms.
Best time for 5120K FFT length: 31.909 ms., avg: 31.934 ms.
Best time for 6144K FFT length: 38.349 ms., avg: 38.406 ms.
Best time for 7168K FFT length: 46.684 ms., avg: 46.772 ms.
Best time for 8192K FFT length: 54.303 ms., avg: 54.359 ms.
Timing FFTs using 2 threads on 1 physical CPUs.
Best time for 768K FFT length: 4.081 ms., avg: 4.144 ms.
Best time for 896K FFT length: 4.745 ms., avg: 4.771 ms.
Best time for 1024K FFT length: 5.486 ms., avg: 5.552 ms.
Best time for 1280K FFT length: 7.089 ms., avg: 7.202 ms.
Best time for 1536K FFT length: 8.838 ms., avg: 8.958 ms.
Best time for 1792K FFT length: 10.206 ms., avg: 10.357 ms.
Best time for 2048K FFT length: 11.821 ms., avg: 12.070 ms.
Best time for 2560K FFT length: 15.299 ms., avg: 15.692 ms.
Best time for 3072K FFT length: 18.920 ms., avg: 19.217 ms.
Best time for 3584K FFT length: 22.490 ms., avg: 22.588 ms.
Best time for 4096K FFT length: 25.571 ms., avg: 25.949 ms.
Best time for 5120K FFT length: 34.642 ms., avg: 35.489 ms.
Best time for 6144K FFT length: 43.871 ms., avg: 44.261 ms.
Best time for 7168K FFT length: 52.427 ms., avg: 53.362 ms.
Best time for 8192K FFT length: 61.596 ms., avg: 61.833 ms.
Timing FFTs using 4 threads on 2 physical CPUs.
Best time for 768K FFT length: 2.089 ms., avg: 2.115 ms.
Best time for 896K FFT length: 2.414 ms., avg: 2.428 ms.
Best time for 1024K FFT length: 2.794 ms., avg: 2.841 ms.
Best time for 1280K FFT length: 3.563 ms., avg: 3.602 ms.
Best time for 1536K FFT length: 4.505 ms., avg: 4.600 ms.
Best time for 1792K FFT length: 5.202 ms., avg: 5.283 ms.
Best time for 2048K FFT length: 6.041 ms., avg: 6.165 ms.
Best time for 2560K FFT length: 7.750 ms., avg: 7.857 ms.
Best time for 3072K FFT length: 9.603 ms., avg: 9.719 ms.
Best time for 3584K FFT length: 11.424 ms., avg: 11.644 ms.
Best time for 4096K FFT length: 13.042 ms., avg: 13.181 ms.
Best time for 5120K FFT length: 17.650 ms., avg: 18.130 ms.
Best time for 6144K FFT length: 22.769 ms., avg: 23.073 ms.
Best time for 7168K FFT length: 27.653 ms., avg: 27.968 ms.
Best time for 8192K FFT length: 30.269 ms., avg: 30.349 ms.
Timing FFTs using 6 threads on 3 physical CPUs.
Best time for 768K FFT length: 1.428 ms., avg: 1.451 ms.
Best time for 896K FFT length: 1.630 ms., avg: 1.714 ms.
Best time for 1024K FFT length: 1.893 ms., avg: 1.926 ms.
Best time for 1280K FFT length: 2.397 ms., avg: 2.419 ms.
Best time for 1536K FFT length: 3.042 ms., avg: 3.069 ms.
Best time for 1792K FFT length: 3.539 ms., avg: 3.570 ms.
Best time for 2048K FFT length: 4.084 ms., avg: 4.147 ms.
Best time for 2560K FFT length: 5.260 ms., avg: 5.319 ms.
Best time for 3072K FFT length: 6.517 ms., avg: 6.589 ms.
Best time for 3584K FFT length: 7.791 ms., avg: 7.846 ms.
Best time for 4096K FFT length: 8.824 ms., avg: 8.919 ms.
Best time for 5120K FFT length: 12.011 ms., avg: 12.308 ms.
Best time for 6144K FFT length: 15.231 ms., avg: 15.625 ms.
Best time for 7168K FFT length: 18.785 ms., avg: 19.083 ms.
Best time for 8192K FFT length: 21.429 ms., avg: 21.449 ms.
Timing FFTs using 8 threads on 4 physical CPUs.
Best time for 768K FFT length: 1.088 ms., avg: 1.109 ms.
Best time for 896K FFT length: 1.237 ms., avg: 1.247 ms.
Best time for 1024K FFT length: 1.432 ms., avg: 1.457 ms.
Best time for 1280K FFT length: 1.809 ms., avg: 1.834 ms.
Best time for 1536K FFT length: 2.282 ms., avg: 2.301 ms.
Best time for 1792K FFT length: 2.657 ms., avg: 2.688 ms.
Best time for 2048K FFT length: 3.082 ms., avg: 3.119 ms.
Best time for 2560K FFT length: 3.989 ms., avg: 4.028 ms.
Best time for 3072K FFT length: 4.997 ms., avg: 5.049 ms.
Best time for 3584K FFT length: 5.960 ms., avg: 5.998 ms.
Best time for 4096K FFT length: 6.777 ms., avg: 6.853 ms.
Best time for 5120K FFT length: 9.183 ms., avg: 9.339 ms.
Best time for 6144K FFT length: 11.694 ms., avg: 11.898 ms.
Best time for 7168K FFT length: 14.237 ms., avg: 14.608 ms.
Best time for 8192K FFT length: 15.414 ms., avg: 15.438 ms.
Timing FFTs using 10 threads on 5 physical CPUs.
Best time for 768K FFT length: 0.889 ms., avg: 0.900 ms.
Best time for 896K FFT length: 1.008 ms., avg: 1.018 ms.
Best time for 1024K FFT length: 1.169 ms., avg: 1.184 ms.
Best time for 1280K FFT length: 1.480 ms., avg: 1.505 ms.
Best time for 1536K FFT length: 1.845 ms., avg: 1.865 ms.
Best time for 1792K FFT length: 2.192 ms., avg: 2.210 ms.
Best time for 2048K FFT length: 2.529 ms., avg: 2.558 ms.
Best time for 2560K FFT length: 3.273 ms., avg: 3.319 ms.
Best time for 3072K FFT length: 4.100 ms., avg: 4.149 ms.
Best time for 3584K FFT length: 4.896 ms., avg: 4.966 ms.
Best time for 4096K FFT length: 5.600 ms., avg: 5.668 ms.
Best time for 5120K FFT length: 7.549 ms., avg: 7.661 ms.
Best time for 6144K FFT length: 9.534 ms., avg: 9.687 ms.
Best time for 7168K FFT length: 11.499 ms., avg: 11.742 ms.
Best time for 8192K FFT length: 13.041 ms., avg: 13.064 ms.
Timing FFTs using 12 threads on 6 physical CPUs.
Best time for 768K FFT length: 0.756 ms., avg: 0.767 ms.
Best time for 896K FFT length: 0.858 ms., avg: 0.870 ms.
Best time for 1024K FFT length: 0.992 ms., avg: 1.006 ms.
Best time for 1280K FFT length: 1.264 ms., avg: 1.285 ms.
Best time for 1536K FFT length: 1.549 ms., avg: 1.576 ms.
Best time for 1792K FFT length: 1.836 ms., avg: 1.860 ms.
Best time for 2048K FFT length: 2.142 ms., avg: 2.171 ms.
Best time for 2560K FFT length: 2.825 ms., avg: 2.865 ms.
Best time for 3072K FFT length: 3.589 ms., avg: 3.624 ms.
Best time for 3584K FFT length: 4.316 ms., avg: 4.348 ms.
Best time for 4096K FFT length: 4.937 ms., avg: 4.977 ms.
Best time for 5120K FFT length: 6.463 ms., avg: 6.667 ms.
Best time for 6144K FFT length: 8.084 ms., avg: 8.229 ms.
Best time for 7168K FFT length: 9.734 ms., avg: 9.919 ms.
Best time for 8192K FFT length: 11.081 ms., avg: 11.130 ms.
Best time for 61 bit trial factors: 1.920 ms.
Best time for 62 bit trial factors: 1.934 ms.
Best time for 63 bit trial factors: 2.190 ms.
Best time for 64 bit trial factors: 2.269 ms.
Best time for 65 bit trial factors: 2.656 ms.
Best time for 66 bit trial factors: 3.137 ms.
Best time for 67 bit trial factors: 3.108 ms.
Best time for 75 bit trial factors: 3.025 ms.
Best time for 76 bit trial factors: 3.019 ms.
Best time for 77 bit trial factors: 3.021 ms.

henryzz · 2012-10-31, 13:32

Looks to me like 6 threads on 6 physical cpus might be better. Basically this means turning hyperthreading off.

James Heinrich · 2012-10-31, 14:22

Quote:

Originally Posted by henryzz

Basically this means turning hyperthreading off.

Or just running 6 threads and leaving HyperThreading enabled, assuming other work will be done on the machine and it's not a dedicated Prime95 box. At least that's what I do.

kracker · 2012-11-08, 01:45

AMD E-300 APU

Code:

Compare your results to other computers at http://www.mersenne.org/report_benchmarks
AMD E-300 APU with Radeon(tm) HD Graphics
CPU speed: 1297.38 MHz, 2 cores
CPU features: 3DNow! Prefetch, SSE, SSE2
L1 cache size: 32 KB
L2 cache size: 512 KB
L1 cache line size: 64 bytes
L2 cache line size: 64 bytes
L1 TLBS: 40
L2 TLBS: 512
Prime95 64-bit version 27.7, RdtscTiming=1
Best time for 768K FFT length: 87.391 ms., avg: 141.794 ms.
Best time for 896K FFT length: 104.511 ms., avg: 108.164 ms.
Best time for 1024K FFT length: 115.036 ms., avg: 118.470 ms.
Best time for 1280K FFT length: 149.465 ms., avg: 153.412 ms.
Best time for 1536K FFT length: 190.291 ms., avg: 195.114 ms.
Best time for 1792K FFT length: 223.167 ms., avg: 229.565 ms.
Best time for 2048K FFT length: 250.200 ms., avg: 255.183 ms.
Best time for 2560K FFT length: 309.440 ms., avg: 317.220 ms.
Best time for 3072K FFT length: 397.815 ms., avg: 412.169 ms.
Best time for 3584K FFT length: 467.972 ms., avg: 484.672 ms.
Best time for 4096K FFT length: 521.833 ms., avg: 532.843 ms.
Best time for 5120K FFT length: 699.955 ms., avg: 717.164 ms.
Best time for 6144K FFT length: 875.186 ms., avg: 895.617 ms.
Best time for 7168K FFT length: 989.090 ms., avg: 995.739 ms.
Best time for 8192K FFT length: 1113.674 ms., avg: 1146.645 ms.
Timing FFTs using 2 threads.
Best time for 768K FFT length: 45.104 ms., avg: 47.919 ms.
Best time for 896K FFT length: 52.750 ms., avg: 54.765 ms.
Best time for 1024K FFT length: 58.753 ms., avg: 68.861 ms.
Best time for 1280K FFT length: 77.048 ms., avg: 86.149 ms.
Best time for 1536K FFT length: 99.087 ms., avg: 106.154 ms.
Best time for 1792K FFT length: 116.079 ms., avg: 121.478 ms.
Best time for 2048K FFT length: 130.418 ms., avg: 140.236 ms.
Best time for 2560K FFT length: 174.972 ms., avg: 184.304 ms.
Best time for 3072K FFT length: 420.825 ms., avg: 456.376 ms.
Best time for 3584K FFT length: 492.879 ms., avg: 537.556 ms.
Best time for 4096K FFT length: 538.925 ms., avg: 579.044 ms.
Best time for 5120K FFT length: 727.217 ms., avg: 776.924 ms.
Best time for 6144K FFT length: 477.650 ms., avg: 622.764 ms.
Best time for 7168K FFT length: 523.489 ms., avg: 525.292 ms.
Best time for 8192K FFT length: 570.619 ms., avg: 584.927 ms.
Best time for 61 bit trial factors: 9.923 ms.
Best time for 62 bit trial factors: 10.192 ms.
Best time for 63 bit trial factors: 11.900 ms.
Best time for 64 bit trial factors: 15.045 ms.
Best time for 65 bit trial factors: 16.751 ms.
Best time for 66 bit trial factors: 19.999 ms.
Best time for 67 bit trial factors: 20.022 ms.
Best time for 75 bit trial factors: 20.588 ms.
Best time for 76 bit trial factors: 20.526 ms.
Best time for 77 bit trial factors: 20.531 ms.

Intel Core2 Duo E6550

Code:

Compare your results to other computers at http://www.mersenne.org/report_benchmarks
Intel(R) Core(TM)2 Duo CPU     E6550  @ 2.33GHz
CPU speed: 2327.49 MHz, 2 cores
CPU features: Prefetch, SSE, SSE2
L1 cache size: 32 KB
L2 cache size: 4 MB
L1 cache line size: 64 bytes
L2 cache line size: 64 bytes
TLBS: 256
Prime95 32-bit version 27.7, RdtscTiming=1
Best time for 768K FFT length: 14.868 ms., avg: 17.768 ms.
Best time for 896K FFT length: 17.810 ms., avg: 19.278 ms.
Best time for 1024K FFT length: 19.631 ms., avg: 20.166 ms.
Best time for 1280K FFT length: 25.675 ms., avg: 26.112 ms.
Best time for 1536K FFT length: 31.820 ms., avg: 32.881 ms.
Best time for 1792K FFT length: 37.585 ms., avg: 38.641 ms.
Best time for 2048K FFT length: 41.521 ms., avg: 42.125 ms.
Best time for 2560K FFT length: 53.452 ms., avg: 53.775 ms.
Best time for 3072K FFT length: 67.550 ms., avg: 68.831 ms.
Best time for 3584K FFT length: 80.515 ms., avg: 82.001 ms.
Best time for 4096K FFT length: 86.909 ms., avg: 87.890 ms.
Best time for 5120K FFT length: 113.483 ms., avg: 114.957 ms.
Best time for 6144K FFT length: 139.527 ms., avg: 141.252 ms.
Best time for 7168K FFT length: 164.358 ms., avg: 166.436 ms.
Best time for 8192K FFT length: 181.695 ms., avg: 184.009 ms.
Timing FFTs using 2 threads.
Best time for 768K FFT length: 8.769 ms., avg: 9.176 ms.
Best time for 896K FFT length: 10.325 ms., avg: 11.153 ms.
Best time for 1024K FFT length: 11.824 ms., avg: 12.181 ms.
Best time for 1280K FFT length: 15.407 ms., avg: 15.716 ms.
Best time for 1536K FFT length: 18.871 ms., avg: 19.727 ms.
Best time for 1792K FFT length: 23.503 ms., avg: 24.262 ms.
Best time for 2048K FFT length: 25.426 ms., avg: 29.771 ms.
Best time for 2560K FFT length: 32.234 ms., avg: 33.768 ms.
Best time for 3072K FFT length: 46.709 ms., avg: 48.348 ms.
Best time for 3584K FFT length: 57.766 ms., avg: 58.543 ms.
Best time for 4096K FFT length: 54.166 ms., avg: 55.867 ms.
Best time for 5120K FFT length: 71.589 ms., avg: 73.280 ms.
Best time for 6144K FFT length: 89.730 ms., avg: 91.996 ms.
Best time for 7168K FFT length: 104.076 ms., avg: 106.812 ms.
Best time for 8192K FFT length: 117.192 ms., avg: 118.672 ms.
Best time for 61 bit trial factors: 4.745 ms.
Best time for 62 bit trial factors: 4.763 ms.
Best time for 63 bit trial factors: 7.660 ms.
Best time for 64 bit trial factors: 7.684 ms.
Best time for 65 bit trial factors: 7.138 ms.
Best time for 66 bit trial factors: 7.125 ms.
Best time for 67 bit trial factors: 7.095 ms.
Best time for 75 bit trial factors: 7.555 ms.
Best time for 76 bit trial factors: 7.592 ms.
Best time for 77 bit trial factors: 7.565 ms.

AMD Sempron LE-1300

Code:

Compare your results to other computers at http://www.mersenne.org/report_benchmarks
AMD Sempron(tm) Processor LE-1300
CPU speed: 2310.49 MHz
CPU features: 3DNow!, SSE, SSE2
L1 cache size: 64 KB
L2 cache size: 512 KB
L1 cache line size: 64 bytes
L2 cache line size: 64 bytes
L1 TLBS: 32
L2 TLBS: 512
Prime95 64-bit version 27.7, RdtscTiming=1
Best time for 768K FFT length: 30.051 ms., avg: 32.553 ms.
Best time for 896K FFT length: 38.671 ms., avg: 39.481 ms.
Best time for 1024K FFT length: 40.193 ms., avg: 41.064 ms.
Best time for 1280K FFT length: 52.859 ms., avg: 57.015 ms.
Best time for 1536K FFT length: 63.492 ms., avg: 66.221 ms.
Best time for 1792K FFT length: 78.951 ms., avg: 80.769 ms.
Best time for 2048K FFT length: 86.857 ms., avg: 89.129 ms.
Best time for 2560K FFT length: 114.682 ms., avg: 118.147 ms.
Best time for 3072K FFT length: 136.675 ms., avg: 139.415 ms.
Best time for 3584K FFT length: 173.936 ms., avg: 179.932 ms.
Best time for 4096K FFT length: 187.139 ms., avg: 191.833 ms.
Best time for 5120K FFT length: 253.838 ms., avg: 259.152 ms.
Best time for 6144K FFT length: 324.138 ms., avg: 330.756 ms.
Best time for 7168K FFT length: 442.552 ms., avg: 489.383 ms.
Best time for 8192K FFT length: 481.176 ms., avg: 490.415 ms.
Best time for 61 bit trial factors: 3.552 ms.
Best time for 62 bit trial factors: 3.619 ms.
Best time for 63 bit trial factors: 4.277 ms.
Best time for 64 bit trial factors: 4.869 ms.
Best time for 65 bit trial factors: 5.932 ms.
Best time for 66 bit trial factors: 6.946 ms.
Best time for 67 bit trial factors: 6.894 ms.
Best time for 75 bit trial factors: 6.760 ms.
Best time for 76 bit trial factors: 6.763 ms.
Best time for 77 bit trial factors: 6.745 ms.

Intel Pentium E5200

Code:

Compare your results to other computers at http://www.mersenne.org/report_benchmarks
Pentium(R) Dual-Core  CPU      E5200  @ 2.50GHz
CPU speed: 2493.80 MHz, 2 cores
CPU features: Prefetch, SSE, SSE2
L1 cache size: 32 KB
L2 cache size: 2 MB
L1 cache line size: 64 bytes
L2 cache line size: 64 bytes
TLBS: 256
Prime95 32-bit version 27.7, RdtscTiming=1
Best time for 768K FFT length: 14.414 ms., avg: 15.317 ms.
Best time for 896K FFT length: 17.264 ms., avg: 18.295 ms.
Best time for 1024K FFT length: 18.934 ms., avg: 21.078 ms.
Best time for 1280K FFT length: 24.630 ms., avg: 25.874 ms.
Best time for 1536K FFT length: 30.345 ms., avg: 31.445 ms.
Best time for 1792K FFT length: 36.030 ms., avg: 38.074 ms.
Best time for 2048K FFT length: 39.516 ms., avg: 41.770 ms.
Best time for 2560K FFT length: 51.150 ms., avg: 53.263 ms.
Best time for 3072K FFT length: 67.773 ms., avg: 70.705 ms.
Best time for 3584K FFT length: 86.234 ms., avg: 101.937 ms.
Best time for 4096K FFT length: 82.980 ms., avg: 87.732 ms.
Best time for 5120K FFT length: 108.831 ms., avg: 113.078 ms.
Best time for 6144K FFT length: 133.649 ms., avg: 142.639 ms.
Best time for 7168K FFT length: 159.742 ms., avg: 165.055 ms.
Best time for 8192K FFT length: 176.228 ms., avg: 183.923 ms.
Timing FFTs using 2 threads.
Best time for 768K FFT length: 9.809 ms., avg: 11.586 ms.
Best time for 896K FFT length: 10.710 ms., avg: 11.806 ms.
Best time for 1024K FFT length: 12.153 ms., avg: 15.580 ms.
Best time for 1280K FFT length: 15.622 ms., avg: 16.616 ms.
Best time for 1536K FFT length: 18.552 ms., avg: 20.037 ms.
Best time for 1792K FFT length: 25.220 ms., avg: 27.367 ms.
Best time for 2048K FFT length: 24.633 ms., avg: 26.740 ms.
Best time for 2560K FFT length: 31.070 ms., avg: 33.678 ms.
Best time for 3072K FFT length: 62.606 ms., avg: 66.542 ms.
Best time for 3584K FFT length: 81.114 ms., avg: 86.333 ms.
Best time for 4096K FFT length: 52.849 ms., avg: 57.065 ms.
Best time for 5120K FFT length: 71.329 ms., avg: 74.952 ms.
Best time for 6144K FFT length: 95.029 ms., avg: 98.523 ms.
Best time for 7168K FFT length: 111.757 ms., avg: 117.066 ms.
Best time for 8192K FFT length: 125.025 ms., avg: 131.302 ms.
Best time for 61 bit trial factors: 4.172 ms.
Best time for 62 bit trial factors: 4.198 ms.
Best time for 63 bit trial factors: 7.103 ms.
Best time for 64 bit trial factors: 7.100 ms.
Best time for 65 bit trial factors: 6.563 ms.
Best time for 66 bit trial factors: 6.562 ms.
Best time for 67 bit trial factors: 6.559 ms.
Best time for 75 bit trial factors: 6.940 ms.
Best time for 76 bit trial factors: 6.996 ms.
Best time for 77 bit trial factors: 6.923 ms.

kracker · 2012-11-08, 01:47

Intel Pentium T4400

Code:

Compare your results to other computers at http://www.mersenne.org/report_benchmarks
Pentium(R) Dual-Core CPU       T4400  @ 2.20GHz
CPU speed: 2195.21 MHz, 2 cores
CPU features: Prefetch, SSE, SSE2
L1 cache size: 32 KB
L2 cache size: 1 MB
L1 cache line size: 64 bytes
L2 cache line size: 64 bytes
TLBS: 256
Prime95 64-bit version 27.7, RdtscTiming=1
Best time for 768K FFT length: 14.009 ms., avg: 15.831 ms.
Best time for 896K FFT length: 17.680 ms., avg: 18.787 ms.
Best time for 1024K FFT length: 18.508 ms., avg: 19.397 ms.
Best time for 1280K FFT length: 23.767 ms., avg: 24.156 ms.
Best time for 1536K FFT length: 29.621 ms., avg: 31.942 ms.
Best time for 1792K FFT length: 36.423 ms., avg: 41.326 ms.
Best time for 2048K FFT length: 39.637 ms., avg: 43.957 ms.
Best time for 2560K FFT length: 50.019 ms., avg: 55.341 ms.
Best time for 3072K FFT length: 62.643 ms., avg: 71.071 ms.
Best time for 3584K FFT length: 74.914 ms., avg: 77.373 ms.
Best time for 4096K FFT length: 88.573 ms., avg: 112.924 ms.
Best time for 5120K FFT length: 106.955 ms., avg: 111.119 ms.
Best time for 6144K FFT length: 134.387 ms., avg: 140.302 ms.
Best time for 7168K FFT length: 184.859 ms., avg: 255.840 ms.
Best time for 8192K FFT length: 193.791 ms., avg: 217.269 ms.
Timing FFTs using 2 threads.
Best time for 768K FFT length: 8.663 ms., avg: 9.434 ms.
Best time for 896K FFT length: 10.602 ms., avg: 11.403 ms.
Best time for 1024K FFT length: 11.546 ms., avg: 12.258 ms.
Best time for 1280K FFT length: 14.641 ms., avg: 15.436 ms.
Best time for 1536K FFT length: 18.156 ms., avg: 19.443 ms.
Best time for 1792K FFT length: 21.515 ms., avg: 27.154 ms.
Best time for 2048K FFT length: 23.877 ms., avg: 31.405 ms.
Best time for 2560K FFT length: 30.561 ms., avg: 32.511 ms.
Best time for 3072K FFT length: 37.583 ms., avg: 39.127 ms.
Best time for 3584K FFT length: 47.638 ms., avg: 50.504 ms.
Best time for 4096K FFT length: 54.596 ms., avg: 57.309 ms.
Best time for 5120K FFT length: 77.799 ms., avg: 81.655 ms.
Best time for 6144K FFT length: 105.528 ms., avg: 109.316 ms.
Best time for 7168K FFT length: 174.678 ms., avg: 183.644 ms.
Best time for 8192K FFT length: 262.170 ms., avg: 270.548 ms.
Best time for 61 bit trial factors: 4.332 ms.
Best time for 62 bit trial factors: 4.589 ms.
Best time for 63 bit trial factors: 5.300 ms.
Best time for 64 bit trial factors: 6.875 ms.
Best time for 65 bit trial factors: 7.300 ms.
Best time for 66 bit trial factors: 8.067 ms.
Best time for 67 bit trial factors: 8.072 ms.
Best time for 75 bit trial factors: 7.774 ms.
Best time for 76 bit trial factors: 7.813 ms.
Best time for 77 bit trial factors: 7.789 ms.

Kyle · 2012-11-12, 13:49

4x AMD Opteron 6176 (4x 12 cores - 4x 32 Go DDR3)

http://pastebin.com/MZdGjv8A (too long to post it here)

I get every time a Segmentation fault at 8K FFT on 17 threads.
Does anyone have the same issue?

henryzz · 2012-11-14, 14:16

Quote:

Originally Posted by Kyle

4x AMD Opteron 6176 (4x 12 cores - 4x 32 Go DDR3)

http://pastebin.com/MZdGjv8A (too long to post it here)

I get every time a Segmentation fault at 8K FFT on 17 threads.
Does anyone have the same issue?

Is it possible that the way the ffts are divided something as small as 8K can only be divided into <=16?
Does anything smaller work with 17 threads?

Kyle · 2012-11-15, 11:36

Quote:

Originally Posted by henryzz

Is it possible that the way the ffts are divided something as small as 8K can only be divided into <=16?
Does anything smaller work with 17 threads?

As you can see 4K, 5K and 6K worked without any problem...

2011-08-28, 14:51	#557
ciic May 2011 FR Germany - Berlin 2·7 Posts	MPrime perf data :: Intel Pentium G620T @ 2.20GHz Hi ! ------------------------------------------------------------------------------------------------------------------- Here's just some more perf data on the new Sandy Bridge Pentium: Intel(R) Pentium(R) CPU G620T @ 2.20GHz 35W TDP (System ~45-55W) [Thu Aug 4 04:25:30 2011] Compare your results to other computers at http://www.mersenne.org/report_benchmarks Intel(R) Pentium(R) CPU G620T @ 2.20GHz CPU speed: 2160.98 MHz, 2 cores CPU features: Prefetch, MMX, SSE, SSE2, SSE4 L1 cache size: 32 KB L2 cache size: 256 KB, L3 cache size: 3 MB L1 cache line size: 64 bytes L2 cache line size: 64 bytes TLBS: 64 Prime95 64-bit version 26.6, RdtscTiming=1 Best time for 768K FFT length: 11.801 ms., avg: 11.817 ms. Best time for 896K FFT length: 14.451 ms., avg: 14.473 ms. Best time for 1024K FFT length: 15.775 ms., avg: 15.802 ms. Best time for 1280K FFT length: 20.394 ms., avg: 20.422 ms. Best time for 1536K FFT length: 24.889 ms., avg: 24.912 ms. Best time for 1792K FFT length: 30.053 ms., avg: 30.067 ms. Best time for 2048K FFT length: 33.272 ms., avg: 33.296 ms. Best time for 2560K FFT length: 42.139 ms., avg: 42.156 ms. Best time for 3072K FFT length: 51.821 ms., avg: 51.844 ms. Best time for 3584K FFT length: 62.475 ms., avg: 62.494 ms. Best time for 4096K FFT length: 70.048 ms., avg: 70.056 ms. Best time for 5120K FFT length: 90.898 ms., avg: 90.910 ms. Best time for 6144K FFT length: 112.647 ms., avg: 112.688 ms. Best time for 7168K FFT length: 137.351 ms., avg: 137.421 ms. Best time for 8192K FFT length: 148.731 ms., avg: 148.790 ms. Timing FFTs using 2 threads. Best time for 768K FFT length: 6.337 ms., avg: 6.366 ms. Best time for 896K FFT length: 7.789 ms., avg: 7.807 ms. Best time for 1024K FFT length: 8.484 ms., avg: 8.528 ms. Best time for 1280K FFT length: 10.847 ms., avg: 10.868 ms. Best time for 1536K FFT length: 13.067 ms., avg: 13.080 ms. Best time for 1792K FFT length: 16.338 ms., avg: 16.372 ms. Best time for 2048K FFT length: 17.974 ms., avg: 18.129 ms. Best time for 2560K FFT length: 22.592 ms., avg: 22.624 ms. Best time for 3072K FFT length: 27.876 ms., avg: 27.908 ms. Best time for 3584K FFT length: 33.386 ms., avg: 33.512 ms. Best time for 4096K FFT length: 36.597 ms., avg: 36.630 ms. Best time for 5120K FFT length: 47.321 ms., avg: 47.347 ms. Best time for 6144K FFT length: 59.132 ms., avg: 59.219 ms. Best time for 7168K FFT length: 77.364 ms., avg: 77.638 ms. Best time for 8192K FFT length: 78.744 ms., avg: 79.095 ms. Best time for 61 bit trial factors: 3.599 ms. Best time for 62 bit trial factors: 3.640 ms. Best time for 63 bit trial factors: 4.084 ms. Best time for 64 bit trial factors: 4.256 ms. Best time for 65 bit trial factors: 4.954 ms. Best time for 66 bit trial factors: 5.820 ms. Best time for 67 bit trial factors: 5.773 ms. Best time for 75 bit trial factors: 5.624 ms. Best time for 76 bit trial factors: 5.610 ms. Best time for 77 bit trial factors: 5.607 ms. [Thu Aug 4 04:45:53 2011] Compare your results to other computers at http://www.mersenne.org/report_benchmarks Intel(R) Pentium(R) CPU G620T @ 2.20GHz CPU speed: 2159.71 MHz, 2 cores CPU features: Prefetch, MMX, SSE, SSE2, SSE4 L1 cache size: 32 KB L2 cache size: 256 KB, L3 cache size: 3 MB L1 cache line size: 64 bytes L2 cache line size: 64 bytes TLBS: 64 Prime95 64-bit version 26.6, RdtscTiming=1 Best time for 768K FFT length: 11.781 ms., avg: 11.802 ms. Best time for 896K FFT length: 14.514 ms., avg: 14.542 ms. Best time for 1024K FFT length: 15.906 ms., avg: 15.929 ms. Best time for 1280K FFT length: 20.630 ms., avg: 20.656 ms. Best time for 1536K FFT length: 25.333 ms., avg: 25.360 ms. Best time for 1792K FFT length: 30.390 ms., avg: 30.434 ms. Best time for 2048K FFT length: 33.557 ms., avg: 33.578 ms. Best time for 2560K FFT length: 42.977 ms., avg: 43.029 ms. Best time for 3072K FFT length: 52.514 ms., avg: 52.534 ms. Best time for 3584K FFT length: 63.344 ms., avg: 63.374 ms. Best time for 4096K FFT length: 70.448 ms., avg: 70.475 ms. Best time for 5120K FFT length: 91.047 ms., avg: 91.087 ms. Best time for 6144K FFT length: 112.786 ms., avg: 112.865 ms. Best time for 7168K FFT length: 136.480 ms., avg: 136.524 ms. Best time for 8192K FFT length: 148.853 ms., avg: 148.892 ms. Timing FFTs using 2 threads. Best time for 768K FFT length: 6.298 ms., avg: 6.368 ms. Best time for 896K FFT length: 7.776 ms., avg: 7.803 ms. Best time for 1024K FFT length: 8.452 ms., avg: 8.480 ms. Best time for 1280K FFT length: 10.881 ms., avg: 10.893 ms. Best time for 1536K FFT length: 13.060 ms., avg: 13.086 ms. Best time for 1792K FFT length: 16.126 ms., avg: 16.152 ms. Best time for 2048K FFT length: 18.057 ms., avg: 18.085 ms. Best time for 2560K FFT length: 22.595 ms., avg: 22.654 ms. Best time for 3072K FFT length: 27.772 ms., avg: 27.859 ms. Best time for 3584K FFT length: 33.517 ms., avg: 33.549 ms. Best time for 4096K FFT length: 37.300 ms., avg: 37.351 ms. Best time for 5120K FFT length: 48.189 ms., avg: 48.293 ms. Best time for 6144K FFT length: 58.756 ms., avg: 58.806 ms. Best time for 7168K FFT length: 80.770 ms., avg: 80.839 ms. Best time for 8192K FFT length: 80.001 ms., avg: 80.046 ms. Best time for 61 bit trial factors: 3.597 ms. Best time for 62 bit trial factors: 3.638 ms. Best time for 63 bit trial factors: 4.083 ms. Best time for 64 bit trial factors: 4.256 ms. Best time for 65 bit trial factors: 4.955 ms. Best time for 66 bit trial factors: 5.817 ms. Best time for 67 bit trial factors: 5.772 ms. Best time for 75 bit trial factors: 5.622 ms. Best time for 76 bit trial factors: 5.606 ms. Best time for 77 bit trial factors: 5.616 ms. [Thu Aug 4 04:54:47 2011] Compare your results to other computers at http://www.mersenne.org/report_benchmarks Intel(R) Pentium(R) CPU G620T @ 2.20GHz CPU speed: 2160.56 MHz, 2 cores CPU features: Prefetch, MMX, SSE, SSE2, SSE4 L1 cache size: 32 KB L2 cache size: 256 KB, L3 cache size: 3 MB L1 cache line size: 64 bytes L2 cache line size: 64 bytes TLBS: 64 Prime95 64-bit version 26.6, RdtscTiming=1 Best time for 768K FFT length: 11.771 ms., avg: 11.796 ms. Best time for 896K FFT length: 14.500 ms., avg: 14.523 ms. Best time for 1024K FFT length: 15.912 ms., avg: 15.938 ms. Best time for 1280K FFT length: 20.677 ms., avg: 20.692 ms. Best time for 1536K FFT length: 25.299 ms., avg: 25.323 ms. Best time for 1792K FFT length: 30.449 ms., avg: 30.471 ms. Best time for 2048K FFT length: 33.537 ms., avg: 33.556 ms. Best time for 2560K FFT length: 43.151 ms., avg: 43.161 ms. Best time for 3072K FFT length: 52.367 ms., avg: 52.384 ms. Best time for 3584K FFT length: 63.243 ms., avg: 63.290 ms. Best time for 4096K FFT length: 70.376 ms., avg: 70.398 ms. Best time for 5120K FFT length: 91.385 ms., avg: 91.410 ms. Best time for 6144K FFT length: 111.428 ms., avg: 111.459 ms. Best time for 7168K FFT length: 136.892 ms., avg: 136.968 ms. Best time for 8192K FFT length: 148.951 ms., avg: 149.047 ms. Timing FFTs using 2 threads. Best time for 768K FFT length: 6.336 ms., avg: 6.359 ms. Best time for 896K FFT length: 7.778 ms., avg: 7.827 ms. Best time for 1024K FFT length: 8.479 ms., avg: 8.515 ms. Best time for 1280K FFT length: 10.831 ms., avg: 10.948 ms. Best time for 1536K FFT length: 13.049 ms., avg: 13.069 ms. Best time for 1792K FFT length: 16.066 ms., avg: 16.089 ms. Best time for 2048K FFT length: 17.947 ms., avg: 17.961 ms. Best time for 2560K FFT length: 22.665 ms., avg: 22.689 ms. Best time for 3072K FFT length: 27.663 ms., avg: 27.783 ms. Best time for 3584K FFT length: 33.524 ms., avg: 33.572 ms. Best time for 4096K FFT length: 36.449 ms., avg: 36.526 ms. Best time for 5120K FFT length: 47.151 ms., avg: 47.301 ms. Best time for 6144K FFT length: 59.534 ms., avg: 59.658 ms. Best time for 7168K FFT length: 77.943 ms., avg: 77.981 ms. Best time for 8192K FFT length: 79.611 ms., avg: 79.806 ms. Best time for 61 bit trial factors: 3.599 ms. Best time for 62 bit trial factors: 3.637 ms. Best time for 63 bit trial factors: 4.083 ms. Best time for 64 bit trial factors: 4.256 ms. Best time for 65 bit trial factors: 4.956 ms. Best time for 66 bit trial factors: 5.815 ms. Best time for 67 bit trial factors: 5.773 ms. Best time for 75 bit trial factors: 5.621 ms. Best time for 76 bit trial factors: 5.610 ms. Best time for 77 bit trial factors: 5.608 ms. Think my benchmarks seem a bit faster than that of the mersenne-aries.sili.net/throughput.php webpage?? Cheers.. ciic

2011-08-28, 15:02	#558
ciic May 2011 FR Germany - Berlin 2·7 Posts	Wikipedia Edits: Prime95 Hi! --------------------------------------------------- Found some new Wikipedia-Edits: * http://en.wikipedia.org/wiki/Prime95 and * http://de.wikipedia.org/wiki/Prime95 Would like some of you to comment - constructively - on them (accuracy, collum/row-titles, physical quatities & units of measurements) or even would like to appeal you for contributional edits - please. Cheers, ciic

2011-12-16, 22:37	#561
Dubslow Basketry That Evening! "Bunslow the Bold" Jun 2011 40<A<43 -89<O<-88 3·29·83 Posts	It still seems that small FFT's do not parallelize well in AVX. (Someone else posted the same in the v27 thread.) Last fiddled with by Dubslow on 2011-12-16 at 22:37

2012-06-12, 14:57	#563
petrw1 1976 Toyota Corona years forever! "Wayne" Nov 2006 Saskatchewan, Canada 11127₈ Posts	DC timing anomaly This may not exactly fit in this thread but I didn't want to start yet another thread for a minor topic. My i5-750 is OC'd at 3.2 Ghz. It has been doing DC for all of 2012 so far. The per iteration timings in the 1344K FFT range have been a very consistent 20Ms on all 4 cores. Just recently it started a 1440K FFT on 1 core and 1344K FFT on the other 3 cores. The 1440K FFT timing went to 20.25Ms; about a 1.25% increase as expected. The timings on the 3 1344K FFT cores DROPPED to 19.8; 19.55 and 19.45; decreases of about 1%, 2.25% and 2.75%. Could I be correct in extrapolating from this that LL/DC performance is slightly better if you mix FFT sizes across the cores.

2012-11-12, 13:49	#573
Kyle Nov 2012 2·3² Posts	4x AMD Opteron 6176 4x AMD Opteron 6176 (4x 12 cores - 4x 32 Go DDR3) http://pastebin.com/MZdGjv8A (too long to post it here) I get every time a Segmentation fault at 8K FFT on 17 threads. Does anyone have the same issue?

2012-10-06, 12:39	#567
henryzz Just call me Henry "David" Sep 2007 Cambridge (GMT/BST) 7×29² Posts	It looks to me like you might get better results if you turn off hyperthreading (at least in prime95. Everything else will be more responsive to normal usage if it is on while prime95 is running). 2 threads on one physical core is slower than 1 thread in your benchmark.

2012-10-28, 05:45	#568
dbaugh Aug 2005 166₈ Posts	[Sat Oct 27 22:11:02 2012] Compare your results to other computers at http://www.mersenne.org/report_benchmarks Intel(R) Core(TM) i7-3960X CPU @ 3.30GHz CPU speed: 3990.97 MHz, 6 hyperthreaded cores CPU features: Prefetch, SSE, SSE2, SSE4, AVX L1 cache size: 32 KB L2 cache size: 256 KB, L3 cache size: 15 MB L1 cache line size: 64 bytes L2 cache line size: 64 bytes TLBS: 64 Prime95 64-bit version 27.7, RdtscTiming=1 Best time for 768K FFT length: 3.930 ms., avg: 3.949 ms. Best time for 896K FFT length: 4.658 ms., avg: 4.667 ms. Best time for 1024K FFT length: 5.276 ms., avg: 5.287 ms. Best time for 1280K FFT length: 6.744 ms., avg: 6.760 ms. Best time for 1536K FFT length: 8.267 ms., avg: 8.290 ms. Best time for 1792K FFT length: 9.898 ms., avg: 9.937 ms. Best time for 2048K FFT length: 11.163 ms., avg: 11.196 ms. Best time for 2560K FFT length: 14.344 ms., avg: 14.386 ms. Best time for 3072K FFT length: 17.935 ms., avg: 17.960 ms. Best time for 3584K FFT length: 21.781 ms., avg: 21.813 ms. Best time for 4096K FFT length: 24.356 ms., avg: 24.418 ms. Best time for 5120K FFT length: 31.909 ms., avg: 31.934 ms. Best time for 6144K FFT length: 38.349 ms., avg: 38.406 ms. Best time for 7168K FFT length: 46.684 ms., avg: 46.772 ms. Best time for 8192K FFT length: 54.303 ms., avg: 54.359 ms. Timing FFTs using 2 threads on 1 physical CPUs. Best time for 768K FFT length: 4.081 ms., avg: 4.144 ms. Best time for 896K FFT length: 4.745 ms., avg: 4.771 ms. Best time for 1024K FFT length: 5.486 ms., avg: 5.552 ms. Best time for 1280K FFT length: 7.089 ms., avg: 7.202 ms. Best time for 1536K FFT length: 8.838 ms., avg: 8.958 ms. Best time for 1792K FFT length: 10.206 ms., avg: 10.357 ms. Best time for 2048K FFT length: 11.821 ms., avg: 12.070 ms. Best time for 2560K FFT length: 15.299 ms., avg: 15.692 ms. Best time for 3072K FFT length: 18.920 ms., avg: 19.217 ms. Best time for 3584K FFT length: 22.490 ms., avg: 22.588 ms. Best time for 4096K FFT length: 25.571 ms., avg: 25.949 ms. Best time for 5120K FFT length: 34.642 ms., avg: 35.489 ms. Best time for 6144K FFT length: 43.871 ms., avg: 44.261 ms. Best time for 7168K FFT length: 52.427 ms., avg: 53.362 ms. Best time for 8192K FFT length: 61.596 ms., avg: 61.833 ms. Timing FFTs using 4 threads on 2 physical CPUs. Best time for 768K FFT length: 2.089 ms., avg: 2.115 ms. Best time for 896K FFT length: 2.414 ms., avg: 2.428 ms. Best time for 1024K FFT length: 2.794 ms., avg: 2.841 ms. Best time for 1280K FFT length: 3.563 ms., avg: 3.602 ms. Best time for 1536K FFT length: 4.505 ms., avg: 4.600 ms. Best time for 1792K FFT length: 5.202 ms., avg: 5.283 ms. Best time for 2048K FFT length: 6.041 ms., avg: 6.165 ms. Best time for 2560K FFT length: 7.750 ms., avg: 7.857 ms. Best time for 3072K FFT length: 9.603 ms., avg: 9.719 ms. Best time for 3584K FFT length: 11.424 ms., avg: 11.644 ms. Best time for 4096K FFT length: 13.042 ms., avg: 13.181 ms. Best time for 5120K FFT length: 17.650 ms., avg: 18.130 ms. Best time for 6144K FFT length: 22.769 ms., avg: 23.073 ms. Best time for 7168K FFT length: 27.653 ms., avg: 27.968 ms. Best time for 8192K FFT length: 30.269 ms., avg: 30.349 ms. Timing FFTs using 6 threads on 3 physical CPUs. Best time for 768K FFT length: 1.428 ms., avg: 1.451 ms. Best time for 896K FFT length: 1.630 ms., avg: 1.714 ms. Best time for 1024K FFT length: 1.893 ms., avg: 1.926 ms. Best time for 1280K FFT length: 2.397 ms., avg: 2.419 ms. Best time for 1536K FFT length: 3.042 ms., avg: 3.069 ms. Best time for 1792K FFT length: 3.539 ms., avg: 3.570 ms. Best time for 2048K FFT length: 4.084 ms., avg: 4.147 ms. Best time for 2560K FFT length: 5.260 ms., avg: 5.319 ms. Best time for 3072K FFT length: 6.517 ms., avg: 6.589 ms. Best time for 3584K FFT length: 7.791 ms., avg: 7.846 ms. Best time for 4096K FFT length: 8.824 ms., avg: 8.919 ms. Best time for 5120K FFT length: 12.011 ms., avg: 12.308 ms. Best time for 6144K FFT length: 15.231 ms., avg: 15.625 ms. Best time for 7168K FFT length: 18.785 ms., avg: 19.083 ms. Best time for 8192K FFT length: 21.429 ms., avg: 21.449 ms. Timing FFTs using 8 threads on 4 physical CPUs. Best time for 768K FFT length: 1.088 ms., avg: 1.109 ms. Best time for 896K FFT length: 1.237 ms., avg: 1.247 ms. Best time for 1024K FFT length: 1.432 ms., avg: 1.457 ms. Best time for 1280K FFT length: 1.809 ms., avg: 1.834 ms. Best time for 1536K FFT length: 2.282 ms., avg: 2.301 ms. Best time for 1792K FFT length: 2.657 ms., avg: 2.688 ms. Best time for 2048K FFT length: 3.082 ms., avg: 3.119 ms. Best time for 2560K FFT length: 3.989 ms., avg: 4.028 ms. Best time for 3072K FFT length: 4.997 ms., avg: 5.049 ms. Best time for 3584K FFT length: 5.960 ms., avg: 5.998 ms. Best time for 4096K FFT length: 6.777 ms., avg: 6.853 ms. Best time for 5120K FFT length: 9.183 ms., avg: 9.339 ms. Best time for 6144K FFT length: 11.694 ms., avg: 11.898 ms. Best time for 7168K FFT length: 14.237 ms., avg: 14.608 ms. Best time for 8192K FFT length: 15.414 ms., avg: 15.438 ms. Timing FFTs using 10 threads on 5 physical CPUs. Best time for 768K FFT length: 0.889 ms., avg: 0.900 ms. Best time for 896K FFT length: 1.008 ms., avg: 1.018 ms. Best time for 1024K FFT length: 1.169 ms., avg: 1.184 ms. Best time for 1280K FFT length: 1.480 ms., avg: 1.505 ms. Best time for 1536K FFT length: 1.845 ms., avg: 1.865 ms. Best time for 1792K FFT length: 2.192 ms., avg: 2.210 ms. Best time for 2048K FFT length: 2.529 ms., avg: 2.558 ms. Best time for 2560K FFT length: 3.273 ms., avg: 3.319 ms. Best time for 3072K FFT length: 4.100 ms., avg: 4.149 ms. Best time for 3584K FFT length: 4.896 ms., avg: 4.966 ms. Best time for 4096K FFT length: 5.600 ms., avg: 5.668 ms. Best time for 5120K FFT length: 7.549 ms., avg: 7.661 ms. Best time for 6144K FFT length: 9.534 ms., avg: 9.687 ms. Best time for 7168K FFT length: 11.499 ms., avg: 11.742 ms. Best time for 8192K FFT length: 13.041 ms., avg: 13.064 ms. Timing FFTs using 12 threads on 6 physical CPUs. Best time for 768K FFT length: 0.756 ms., avg: 0.767 ms. Best time for 896K FFT length: 0.858 ms., avg: 0.870 ms. Best time for 1024K FFT length: 0.992 ms., avg: 1.006 ms. Best time for 1280K FFT length: 1.264 ms., avg: 1.285 ms. Best time for 1536K FFT length: 1.549 ms., avg: 1.576 ms. Best time for 1792K FFT length: 1.836 ms., avg: 1.860 ms. Best time for 2048K FFT length: 2.142 ms., avg: 2.171 ms. Best time for 2560K FFT length: 2.825 ms., avg: 2.865 ms. Best time for 3072K FFT length: 3.589 ms., avg: 3.624 ms. Best time for 3584K FFT length: 4.316 ms., avg: 4.348 ms. Best time for 4096K FFT length: 4.937 ms., avg: 4.977 ms. Best time for 5120K FFT length: 6.463 ms., avg: 6.667 ms. Best time for 6144K FFT length: 8.084 ms., avg: 8.229 ms. Best time for 7168K FFT length: 9.734 ms., avg: 9.919 ms. Best time for 8192K FFT length: 11.081 ms., avg: 11.130 ms. Best time for 61 bit trial factors: 1.920 ms. Best time for 62 bit trial factors: 1.934 ms. Best time for 63 bit trial factors: 2.190 ms. Best time for 64 bit trial factors: 2.269 ms. Best time for 65 bit trial factors: 2.656 ms. Best time for 66 bit trial factors: 3.137 ms. Best time for 67 bit trial factors: 3.108 ms. Best time for 75 bit trial factors: 3.025 ms. Best time for 76 bit trial factors: 3.019 ms. Best time for 77 bit trial factors: 3.021 ms.

2012-10-31, 13:32	#569
henryzz Just call me Henry "David" Sep 2007 Cambridge (GMT/BST) 1011011111111₂ Posts	Looks to me like 6 threads on 6 physical cpus might be better. Basically this means turning hyperthreading off.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Perpetual "interesting video" thread...	Xyzzy	Lounge	43	2021-07-17 00:00
LLR benchmark thread	Oddball	Riesel Prime Search	5	2010-08-02 00:11
Perpetual I'm pi**ed off thread	rogue	Soap Box	19	2009-10-28 19:17
Perpetual autostereogram thread...	Xyzzy	Lounge	10	2006-09-28 00:36
Perpetual ECM factoring challenge thread...	Xyzzy	Factoring	65	2005-09-05 08:16