mersenneforum.org  

Go Back   mersenneforum.org > Great Internet Mersenne Prime Search > Hardware

Reply
 
Thread Tools
Old 2006-03-23, 18:20   #210
Templus
 
Templus's Avatar
 
Jun 2004

2·53 Posts
Default

AMD Turion(tm) 64 Mobile Technology MT-34
CPU speed: 795.86 MHz
CPU features: RDTSC, CMOV, Prefetch, 3DNow!, MMX, SSE, SSE2
L1 cache size: 64 KB
L2 cache size: 1024 KB
L1 cache line size: 64 bytes
L2 cache line size: 64 bytes
L1 TLBS: 32
L2 TLBS: 512
Prime95 32-bit version 24.14, RdtscTiming=1
Best time for 512K FFT length: 26.696 ms.
Best time for 640K FFT length: 35.347 ms.
Best time for 768K FFT length: 43.540 ms.
Best time for 896K FFT length: 51.667 ms.
Best time for 1024K FFT length: 57.578 ms.
Best time for 1280K FFT length: 74.048 ms.
Best time for 1536K FFT length: 88.726 ms.
Best time for 1792K FFT length: 109.521 ms.
Best time for 2048K FFT length: 120.541 ms.
Best time for 2560K FFT length: 159.731 ms.
Best time for 3072K FFT length: 195.941 ms.
Best time for 3584K FFT length: 236.920 ms.
Best time for 4096K FFT length: 259.591 ms.
Best time for 58 bit trial factors: 6.720 ms.
Best time for 59 bit trial factors: 6.723 ms.
Best time for 60 bit trial factors: 6.761 ms.
Best time for 61 bit trial factors: 6.735 ms.
Best time for 62 bit trial factors: 12.100 ms.
Best time for 63 bit trial factors: 12.135 ms.
Best time for 64 bit trial factors: 15.686 ms.
Best time for 65 bit trial factors: 15.678 ms.
Best time for 66 bit trial factors: 15.534 ms.
Best time for 67 bit trial factors: 15.552 ms.
Templus is offline   Reply With Quote
Old 2006-03-24, 19:41   #211
TheJudger
 
TheJudger's Avatar
 
"Oliver"
Mar 2005
Germany

2×3×5×37 Posts
Default

hi Guido,

6-7% on a Pentium-D? For me it seems to be too much.

Personnally I own a 950 (3.4GHz, Dualcore, 2x 2MB L2), currently running @3,8GHz

Running one task of prime95 at 896k FFT: ~22.80 ms per iteration
Running two tasks of prime95 at 896k FFT: ~23.05 ms per iteration
(current doublechecks are 896k ;))
so the slowdown on my machine for 2 tasks of prime95 is arround 1% :)

896k FFT means 7MB of memory -> bigger than L2-cache
TheJudger is offline   Reply With Quote
Old 2006-03-29, 21:02   #212
E_tron
 
E_tron's Avatar
 
Sep 2002
Austin, TX

56110 Posts
Default

Quote:
Originally Posted by Templus
AMD Turion(tm) 64 Mobile Technology MT-34
CPU speed: 795.86 MHz
CPU features: RDTSC, CMOV, Prefetch, 3DNow!, MMX, SSE, SSE2
L1 cache size: 64 KB
L2 cache size: 1024 KB
L1 cache line size: 64 bytes
L2 cache line size: 64 bytes
L1 TLBS: 32
L2 TLBS: 512
Prime95 32-bit version 24.14, RdtscTiming=1
Best time for 512K FFT length: 26.696 ms.
Best time for 640K FFT length: 35.347 ms.
Best time for 768K FFT length: 43.540 ms.
Best time for 896K FFT length: 51.667 ms.
Best time for 1024K FFT length: 57.578 ms.
Best time for 1280K FFT length: 74.048 ms.
Best time for 1536K FFT length: 88.726 ms.
Best time for 1792K FFT length: 109.521 ms.
Best time for 2048K FFT length: 120.541 ms.
Best time for 2560K FFT length: 159.731 ms.
Best time for 3072K FFT length: 195.941 ms.
Best time for 3584K FFT length: 236.920 ms.
Best time for 4096K FFT length: 259.591 ms.
Best time for 58 bit trial factors: 6.720 ms.
Best time for 59 bit trial factors: 6.723 ms.
Best time for 60 bit trial factors: 6.761 ms.
Best time for 61 bit trial factors: 6.735 ms.
Best time for 62 bit trial factors: 12.100 ms.
Best time for 63 bit trial factors: 12.135 ms.
Best time for 64 bit trial factors: 15.686 ms.
Best time for 65 bit trial factors: 15.678 ms.
Best time for 66 bit trial factors: 15.534 ms.
Best time for 67 bit trial factors: 15.552 ms.

How fast is that turion? It looks like prime95 recorded its idle speed.
E_tron is offline   Reply With Quote
Old 2006-04-13, 17:38   #213
ewmayer
2ω=0
 
ewmayer's Avatar
 
Sep 2002
Rep├║blica de California

2D3B16 Posts
Default

Might I suggest moving to a more-transparent set of timing benchmarks for trial-factoring? The LL-test benchmark is transparent, and can easily be compared with non-Prime95 LL-test programs. The TF benchmark, on the other hand, is some internal measure of Prime95's factoring speed which tells me zilch about how long it would take to actually factor a given Mersenne number to a given bit depth. I suggest instead using a TF benchmark based on the time needed to factor an M(p) of right around 10Mdigits to various bit depths, say:

Time to TF M33219281:

From 1 to 64 bits;
From 64 to 65 bits;
From 65 to 66 bits;
From 66 to 67 bits;
From 67 to 68 bits;
From 68 to 69 bits;
From 69 to 70 bits.

These runs wouldn't need to be run to completion - the code could do (say) just one of the 16 factoring passes to the desired depth and then multiply the resulting timing by 16.

For example, the latest build of my Mfactor code running on a 1.4GHz Itanium 2 needs the following times for the above benchmark:

Time to TF M33219281:

From 1 to 64 bits: 3360 sec (i.e. 210 sec/pass)
From 64 to 65 bits: 7262 sec (i.e. ~2.2x performance penalty for q > 64-bit)
From 65 to 66 bits: 14368 sec

For Mfactor, the remaining timings can easily be extrapolated, since the code uses the same 96-bit modmul routines for any factor candidates > 64-bits. So the time to go from 65 to 66-bit is almost exactly double that from 64-to-65-bit, and so forth. This may not be true of factoring code that uses floating-point-based modmul routines (like Prime95), hence the bit-per-bit breakdown.

Similarly, for a 2.2GHz AMD64, we have

Time to TF M33219281:

From 1 to 64 bits: 2528 sec (i.e. 158 sec/pass)
From 64 to 65 bits: 7360 sec (i.e. ~3x performance penalty for q > 64-bit)
etc.

Last fiddled with by ewmayer on 2006-04-13 at 17:40
ewmayer is online now   Reply With Quote
Old 2006-06-06, 20:26   #214
markhl
 
Apr 2003
California

9210 Posts
Default

This is version 24.14, running on my new eMachines D5039 system. It's good to get a new system, but the fan is noisy when Prime95 runs!

Intel(R) Pentium(R) 4 CPU 3.06GHz
CPU speed: 3066.52 MHz
CPU features: RDTSC, CMOV, Prefetch, MMX, SSE, SSE2
L1 cache size: 16 KB
L2 cache size: 1024 KB
L1 cache line size: 64 bytes
L2 cache line size: 128 bytes
TLBS: 64
Prime95 32-bit version 24.14, RdtscTiming=1
Best time for 512K FFT length: 15.663 ms.
Best time for 640K FFT length: 19.980 ms.
Best time for 768K FFT length: 24.339 ms.
Best time for 896K FFT length: 29.447 ms.
Best time for 1024K FFT length: 33.746 ms.
Best time for 1280K FFT length: 41.514 ms.
Best time for 1536K FFT length: 50.402 ms.
Best time for 1792K FFT length: 60.032 ms.
Best time for 2048K FFT length: 67.210 ms.
Best time for 2560K FFT length: 89.048 ms.
Best time for 3072K FFT length: 107.702 ms.
Best time for 3584K FFT length: 129.165 ms.
Best time for 4096K FFT length: 145.755 ms.
Best time for 58 bit trial factors: 9.046 ms.
Best time for 59 bit trial factors: 9.103 ms.
Best time for 60 bit trial factors: 9.071 ms.
Best time for 61 bit trial factors: 9.107 ms.
Best time for 62 bit trial factors: 12.388 ms.
Best time for 63 bit trial factors: 12.442 ms.
Best time for 64 bit trial factors: 14.657 ms.
Best time for 65 bit trial factors: 14.557 ms.
Best time for 66 bit trial factors: 14.536 ms.
Best time for 67 bit trial factors: 14.568 ms.
markhl is offline   Reply With Quote
Old 2006-06-14, 18:37   #215
AntonVrba
 
AntonVrba's Avatar
 
Jun 2005

2×72 Posts
Default

System is a standard HP xw62000 workstation with 2 x Dual Xeon 3.6 GHZ

Hyperthreading is disabled.
(If Hyperthreading is enabled adds 3.2mS to 4096K FFT)

Best regards
Anton


Intel(R) Xeon(TM) CPU 3.60GHz
CPU speed: 3600.05 MHz
CPU features: RDTSC, CMOV, Prefetch, MMX, SSE, SSE2
L1 cache size: 16 KB
L2 cache size: 2048 KB
L1 cache line size: 64 bytes
L2 cache line size: 64 bytes
TLBS: 64
Prime95 32-bit version 24.14, RdtscTiming=1
Best time for 512K FFT length: 13.028 ms.
Best time for 640K FFT length: 16.685 ms.
Best time for 768K FFT length: 20.281 ms.
Best time for 896K FFT length: 24.665 ms.
Best time for 1024K FFT length: 28.338 ms.
Best time for 1280K FFT length: 35.071 ms.
Best time for 1536K FFT length: 42.401 ms.
Best time for 1792K FFT length: 50.570 ms.
Best time for 2048K FFT length: 56.613 ms.
Best time for 2560K FFT length: 74.864 ms.
Best time for 3072K FFT length: 90.841 ms.
Best time for 3584K FFT length: 108.321 ms.
Best time for 4096K FFT length: 121.698 ms.
Best time for 58 bit trial factors: 7.679 ms.
Best time for 59 bit trial factors: 7.684 ms.
Best time for 60 bit trial factors: 7.670 ms.
Best time for 61 bit trial factors: 7.671 ms.
Best time for 62 bit trial factors: 10.525 ms.
Best time for 63 bit trial factors: 10.537 ms.
Best time for 64 bit trial factors: 12.379 ms.
Best time for 65 bit trial factors: 12.303 ms.
Best time for 66 bit trial factors: 12.344 ms.
Best time for 67 bit trial factors: 12.289 ms.
AntonVrba is offline   Reply With Quote
Old 2006-08-04, 10:35   #216
T.Rex
 
T.Rex's Avatar
 
Feb 2004
France

16228 Posts
Default Woodrest 2.66 GHz under Linux 64bits 2.6.17 - 4 CPUs

Done with mprime 24.6 .
Alone on machine.

T.

# cat /proc/cpuinfo
processor : 0
vendor_id : GenuineIntel
cpu family : 6
model : 15
model name : Genuine Intel(R) CPU @ 2.66GHz
stepping : 4
cpu MHz : 2666.713
cache size : 4096 KB
physical id : 0
siblings : 2
core id : 0
cpu cores : 2
fpu : yes
fpu_exception : yes
cpuid level : 10
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm syscall lm constant_tsc pni monitor ds_cpl vmx est tm2 cx16 xtpr lahf_lm
bogomips : 5337.45
clflush size : 64
cache_alignment : 64
address sizes : 36 bits physical, 48 bits virtual
power management:

etc.



Timing 100 iterations at 4K FFT length. Best time: 0.049 ms.
Timing 100 iterations at 5K FFT length. Best time: 0.073 ms.
Timing 100 iterations at 6K FFT length. Best time: 0.087 ms.
Timing 100 iterations at 7K FFT length. Best time: 0.106 ms.
Timing 100 iterations at 8K FFT length. Best time: 0.110 ms.
Timing 100 iterations at 10K FFT length. Best time: 0.143 ms.
Timing 100 iterations at 12K FFT length. Best time: 0.175 ms.
Timing 100 iterations at 14K FFT length. Best time: 0.209 ms.
Timing 100 iterations at 16K FFT length. Best time: 0.232 ms.
Timing 100 iterations at 20K FFT length. Best time: 0.297 ms.
Timing 100 iterations at 24K FFT length. Best time: 0.364 ms.
Timing 100 iterations at 28K FFT length. Best time: 0.437 ms.
Timing 100 iterations at 32K FFT length. Best time: 0.480 ms.
Timing 100 iterations at 40K FFT length. Best time: 0.693 ms.
Timing 100 iterations at 48K FFT length. Best time: 0.846 ms.
Timing 100 iterations at 56K FFT length. Best time: 1.010 ms.
Timing 100 iterations at 64K FFT length. Best time: 1.117 ms.
Timing 100 iterations at 80K FFT length. Best time: 1.478 ms.
Timing 100 iterations at 96K FFT length. Best time: 1.800 ms.
Timing 100 iterations at 112K FFT length. Best time: 2.143 ms.
Timing 100 iterations at 128K FFT length. Best time: 2.380 ms.
Timing 100 iterations at 160K FFT length. Best time: 3.096 ms.
Timing 100 iterations at 192K FFT length. Best time: 3.755 ms.
Timing 100 iterations at 224K FFT length. Best time: 4.495 ms.
Timing 100 iterations at 256K FFT length. Best time: 5.006 ms.
Timing 100 iterations at 320K FFT length. Best time: 6.770 ms.
Timing 88 iterations at 384K FFT length. Best time: 8.353 ms.
Timing 76 iterations at 448K FFT length. Best time: 10.270 ms.
Timing 66 iterations at 512K FFT length. Best time: 11.315 ms.
Timing 53 iterations at 640K FFT length. Best time: 14.290 ms.
Timing 44 iterations at 768K FFT length. Best time: 17.647 ms.
Timing 38 iterations at 896K FFT length. Best time: 21.018 ms.
Timing 33 iterations at 1024K FFT length. Best time: 23.495 ms.
Timing 26 iterations at 1280K FFT length. Best time: 31.071 ms.
Timing 22 iterations at 1536K FFT length. Best time: 38.243 ms.
Timing 19 iterations at 1792K FFT length. Best time: 46.762 ms.
Timing 16 iterations at 2048K FFT length. Best time: 52.662 ms.
Timing 13 iterations at 2560K FFT length. Best time: 68.639 ms.
Timing 11 iterations at 3072K FFT length. Best time: 83.836 ms.
Timing 10 iterations at 3584K FFT length. Best time: 101.619 ms.
Timing 10 iterations at 4096K FFT length. Best time: 112.420 ms.

Last fiddled with by T.Rex on 2006-08-04 at 10:41
T.Rex is offline   Reply With Quote
Old 2006-08-04, 10:51   #217
T.Rex
 
T.Rex's Avatar
 
Feb 2004
France

91410 Posts
Default Same machine (Woodcrest 2.66 GHz) - mprime 24.14.2

Done with mprime 24.14.2 .

About 10 % better than previous version for 4096K FFT !!

T.


Timing 100 iterations at 4K FFT length. Best time: 0.048 ms.
Timing 100 iterations at 5K FFT length. Best time: 0.072 ms.
Timing 100 iterations at 6K FFT length. Best time: 0.086 ms.
Timing 100 iterations at 7K FFT length. Best time: 0.105 ms.
Timing 100 iterations at 8K FFT length. Best time: 0.109 ms.
Timing 100 iterations at 10K FFT length. Best time: 0.149 ms.
Timing 100 iterations at 12K FFT length. Best time: 0.182 ms.
Timing 100 iterations at 14K FFT length. Best time: 0.221 ms.
Timing 100 iterations at 16K FFT length. Best time: 0.232 ms.
Timing 100 iterations at 20K FFT length. Best time: 0.314 ms.
Timing 100 iterations at 24K FFT length. Best time: 0.382 ms.
Timing 100 iterations at 28K FFT length. Best time: 0.472 ms.
Timing 100 iterations at 32K FFT length. Best time: 0.493 ms.
Timing 100 iterations at 40K FFT length. Best time: 0.646 ms.
Timing 100 iterations at 48K FFT length. Best time: 0.785 ms.
Timing 100 iterations at 56K FFT length. Best time: 0.961 ms.
Timing 100 iterations at 64K FFT length. Best time: 1.026 ms.
Timing 100 iterations at 80K FFT length. Best time: 1.447 ms.
Timing 100 iterations at 96K FFT length. Best time: 1.761 ms.
Timing 100 iterations at 112K FFT length. Best time: 2.124 ms.
Timing 100 iterations at 128K FFT length. Best time: 2.274 ms.
Timing 100 iterations at 160K FFT length. Best time: 2.765 ms.
Timing 100 iterations at 192K FFT length. Best time: 3.410 ms.
Timing 100 iterations at 224K FFT length. Best time: 4.080 ms.
Timing 100 iterations at 256K FFT length. Best time: 4.533 ms.
Timing 100 iterations at 320K FFT length. Best time: 5.810 ms.
Timing 88 iterations at 384K FFT length. Best time: 7.191 ms.
Timing 76 iterations at 448K FFT length. Best time: 8.717 ms.
Timing 66 iterations at 512K FFT length. Best time: 9.777 ms.
Timing 53 iterations at 640K FFT length. Best time: 13.180 ms.
Timing 44 iterations at 768K FFT length. Best time: 16.255 ms.
Timing 38 iterations at 896K FFT length. Best time: 19.509 ms.
Timing 33 iterations at 1024K FFT length. Best time: 21.696 ms.
Timing 26 iterations at 1280K FFT length. Best time: 27.712 ms.
Timing 22 iterations at 1536K FFT length. Best time: 33.935 ms.
Timing 19 iterations at 1792K FFT length. Best time: 40.640 ms.
Timing 16 iterations at 2048K FFT length. Best time: 45.401 ms.
Timing 13 iterations at 2560K FFT length. Best time: 60.097 ms.
Timing 11 iterations at 3072K FFT length. Best time: 73.488 ms.
Timing 10 iterations at 3584K FFT length. Best time: 89.542 ms.
Timing 10 iterations at 4096K FFT length. Best time: 100.804 ms.
Timing 10 iterations at 5120K FFT length. Best time: 132.577 ms.
Timing 10 iterations at 6144K FFT length. Best time: 160.935 ms.
Timing 10 iterations at 7168K FFT length. Best time: 195.470 ms.
Timing 10 iterations at 8192K FFT length. Best time: 214.779 ms.
Timing 10 iterations at 10240K FFT length. Best time: 286.263 ms.
Timing 10 iterations at 12288K FFT length. Best time: 347.065 ms.
Timing 10 iterations at 14336K FFT length. Best time: 420.182 ms.
Timing 10 iterations at 16384K FFT length. Best time: 462.151 ms.
Timing 10 iterations at 20480K FFT length. Best time: 626.761 ms.
Timing 10 iterations at 24576K FFT length. Best time: 761.485 ms.
Timing 10 iterations at 28672K FFT length. Best time: 910.194 ms.
Timing 10 iterations at 32768K FFT length. Best time: 1004.484 ms.
Timing trial factoring of M35000011 with 58 bit length factors. Best time: 4.101 ms.
Timing trial factoring of M35000011 with 59 bit length factors. Best time: 4.098 ms.
Timing trial factoring of M35000011 with 60 bit length factors. Best time: 4.127 ms.
Timing trial factoring of M35000011 with 61 bit length factors. Best time: 4.088 ms.
Timing trial factoring of M35000011 with 62 bit length factors. Best time: 6.675 ms.
Timing trial factoring of M35000011 with 63 bit length factors. Best time: 6.675 ms.
Timing trial factoring of M35000011 with 64 bit length factors. Best time: 6.392 ms.
Timing trial factoring of M35000011 with 65 bit length factors. Best time: 6.347 ms.
Timing trial factoring of M35000011 with 66 bit length factors. Best time: 6.349 ms.
Timing trial factoring of M35000011 with 67 bit length factors. Best time: 6.319 ms.
Benchmark complete.
T.Rex is offline   Reply With Quote
Old 2006-08-15, 17:54   #218
BlueCatZ1
 
Aug 2006

2·3 Posts
Default Intel X6800 2.93 Extreme

Here is my brothers new Comp:
Running Windows Xp Media Center Edition


Intel(R) Core(TM)2 CPU X6800 @ 2.93GHz
CPU speed: 2933.32 MHz
CPU features: RDTSC, CMOV, Prefetch, MMX, SSE, SSE2
L1 cache size: 32 KB
L2 cache size: unknown
L1 cache line size: 64 bytes
L2 cache line size: unknown
Prime95 32-bit version 24.14, RdtscTiming=1
Best time for 512K FFT length: 8.797 ms.
Best time for 640K FFT length: 12.384 ms.
Best time for 768K FFT length: 14.971 ms.
Best time for 896K FFT length: 17.801 ms.
Best time for 1024K FFT length: 19.771 ms.
Best time for 1280K FFT length: 25.228 ms.
Best time for 1536K FFT length: 30.783 ms.
Best time for 1792K FFT length: 36.648 ms.
Best time for 2048K FFT length: 45.836 ms.
Best time for 2560K FFT length: 54.814 ms.
Best time for 3072K FFT length: 66.701 ms.
Best time for 3584K FFT length: 80.777 ms.
Best time for 4096K FFT length: 89.120 ms.
Best time for 58 bit trial factors: 3.807 ms.
Best time for 59 bit trial factors: 3.815 ms.
Best time for 60 bit trial factors: 3.818 ms.
Best time for 61 bit trial factors: 3.812 ms.
Best time for 62 bit trial factors: 6.082 ms.
Best time for 63 bit trial factors: 6.083 ms.
Best time for 64 bit trial factors: 5.827 ms.
Best time for 65 bit trial factors: 5.786 ms.
Best time for 66 bit trial factors: 5.789 ms.
Best time for 67 bit trial factors: 5.778 ms.

Would I see any increase with Winxp 64 bit?
BlueCatZ1 is offline   Reply With Quote
Old 2006-08-18, 01:55   #219
James Heinrich
 
James Heinrich's Avatar
 
"James Heinrich"
May 2004
ex-Northern Ontario

1100101001102 Posts
Default

Quote:
Originally Posted by BlueCatZ1 View Post
Would I see any increase with Winxp 64 bit?
Quite possibly, at least for trial factoring. In this thread I show the comparison between my Opteron165 under WinXP32 and Vista64 and TF in 64-bit mode is 33-55% faster. I'm not sure how that applies to Core2, but I would imagine it would be somewhat similar.
James Heinrich is offline   Reply With Quote
Old 2006-08-20, 22:26   #220
James Heinrich
 
James Heinrich's Avatar
 
"James Heinrich"
May 2004
ex-Northern Ontario

2·1,619 Posts
Default

I've taken all the results posted here, along with some of my own benchmarks, and compiled it into a database that allows you to sort by speed in any FFT size or TF bitdepth, or by CPU type, L2 cache size, MHz, etc:

http://mersenne-aries.sili.net/bench.php

If you want your results included in there, just post in this thread and I'll include them within a day or so.
James Heinrich is offline   Reply With Quote
Reply

Thread Tools


Similar Threads
Thread Thread Starter Forum Replies Last Post
Perpetual "interesting video" thread... Xyzzy Lounge 14 2021-01-15 07:44
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

All times are UTC. The time now is 21:59.

Sat Jan 16 21:59:52 UTC 2021 up 44 days, 18:11, 0 users, load averages: 2.62, 2.22, 1.88

Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2021, Jelsoft Enterprises Ltd.

This forum has received and complied with 0 (zero) government requests for information.

Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.2 or any later version published by the Free Software Foundation.
A copy of the license is included in the FAQ.