Let's buy GIMPS an Opteron! - Page 7

Xyzzy · 2003-09-04, 23:29

Quote:

Originally Posted by Prime95

To those running timings on the opteron: I've started an mprime running. To pause it for benchmarking purposes, run the program /home/gw/pauser. Within 2 minutes mprime will stop running until you kill pauser. Please remember to kill pauser before you logout.

Are you running the client via nohup? You might want to so the process isn't killed if (when!) your connection is killed... I have a pretty reliable line but it is far from perfect...

Also, can the pauser program maybe kick back on if say 30 minutes of idle time goes by? I'd hate for someone to forget to kill it... (I check everything every 3-4 hours so I suppose I might catch it eventually!)

I'm curious, how much of an effect does having a program like mprime running at nice 19 have on running timings?

Do you think us having a revision B Opteron will be a problem? I'm half tempted to call AMD and complain, since our primary development focus deals with SSE2... Thoughts?

Prime95 · 2003-09-05, 00:37

Quote:

Originally Posted by Xyzzy

Are you running the client via nohup? You might want to so the process isn't killed if (when!) your connection is killed... I have a pretty reliable line but it is far from perfect...

Yeah, I've been dropped 4 times today. It seems if I'm idle for a while I get kicked off automatically.

Quote:

Also, can the pauser program maybe kick back on if say 30 minutes of idle time goes by? I'd hate for someone to forget to kill it

Good idea. I'll make pauser operate for just 1 hour.

Quote:

I'm curious, how much of an effect does having a program like mprime running at nice 19 have on running timings?

On my Celeron 950 it makes a big difference. Using FreeBSD on my Athlon it makes no difference. I just guessed its a Linux scheduler issue. Haven't tried it on the Opteron yet.

Quote:

Do you think us having a revision B Opteron will be a problem?

I doubt it. I suspect the changes are very, very minor. As soon as we get access to an Opteron C we will know for sure.

Prime95 · 2003-09-05, 02:36

BTW, would anyone like to volunteer to get some mlucas and glucas timings?

Xyzzy · 2003-09-05, 03:27

Quote:

Originally Posted by Prime95

Yeah, I've been dropped 4 times today. It seems if I'm idle for a while I get kicked off automatically.

Putty has a keep alive option...

Xyzzy · 2003-09-05, 03:33

Looks like P-1 time!

Code:

  PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME+  Command
 3112 gw        39  19  359m 359m  736 R 99.9 72.5 691:50.32 ./mprime

Dresdenboy · 2003-09-05, 07:51

Quote:

Originally Posted by GP2

I wonder if there is any other way to tell what revision it is, other than looking at the markings on the CPU, which is impractical after thermal compound and heat sink have been applied. Do the CPUID programs out there handle this yet?

I researched that topic a bit. The CPUID information is enough to find out about the revision. That should be available in most CPUID tools and also in /proc/cpuinfo.

Here's what I found out:
All Opterons have following values:
family 15
model 5

Revision B or earlier show stepping 0 or 1
Revision C shows stepping 8

(Athlon 64 1.8GHz shows family 15, model 4, stepping 8)

CPU-Z additionally shows a revision string, which is "SH7-B3" for some rev B CPUs. I also read "B0" somewhere. So it could be that "B" in this string really means revision B and that there are different B steppings too.

Some really new benchmark results (http://pcweb.mycom.co.jp/benchmarkla.../25/index.html) show that there is not just a little difference but a difference in the two digit percent range. I don't show SSE2 differences in Sandra again (since it's an synthetic benchmark and can suffer overproportionally from a few wrongly (or not at all) optimized instructions). Instead this one (especially MPEG-2 and DIVX) is more interesting - and a real world benchmark (taken from http://pcweb.mycom.co.jp/benchmarkla.../25/page5.html):
http://pcweb.mycom.co.jp/benchmarkla...images/g15.png

Remember: the clock difference between 240 (1.4GHz) and 144 (1.8GHz) is only 28.6%!

The 240 could be a CPU which was sold months ago while the 144 is really new.

Regards,
DDB

Xyzzy · 2003-09-05, 09:11

Code:

mv@opteron:/proc> cat cpuinfo
processor       : 0
vendor_id       : AuthenticAMD
cpu family      : 15
model           : 5
model name      : AMD Opteron(tm) Processor 140
stepping        : 1
cpu MHz         : 1396.059
cache size      : 1024 KB
fpu             : yes
fpu_exception   : yes
cpuid level     : 1
wp              : yes
flags           : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 syscall nx mmxext lm 3dnowext 3dnow
bogomips        : 2785.28
TLB size        : 1088 4K pages
clflush size    : 64
address sizes   : 40 bits physical, 48 bits virtual
power management: ts ttp

Xyzzy · 2003-09-05, 09:16

$925.13 in donations, $835.32 spent, and $89.81 left...

I just got around to ordering the CD-ROM... It was $16.99 plus $5 for delivery...

http://www.newegg.com/app/viewProduct.asp?description=27-101-204&depa=1

Bok · 2003-09-05, 13:56

Back in town today, this is my cpuinfo

Opteron64:~ # cat /proc/cpuinfo
processor : 0
vendor_id : AuthenticAMD
cpu family : 15
model : 5
model name : AMD Engineering Sample
stepping : 8
cpu MHz : 1800.028
cache size : 1024 KB
fpu : yes
fpu_exception : yes
cpuid level : 1
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 syscall nx mmxext lm 3dnowext 3dnow
bogomips : 3591.37
TLB size : 1088 4K pages
clflush size : 64
address sizes : 40 bits physical, 48 bits virtual
power management: ts ttp

so I guess it is a 'C'

I'll start running some tests if I can find the right software.

Dresdenboy · 2003-09-05, 14:12

Quote:

Originally Posted by Bok

cpu family : 15
model : 5
model name : AMD Engineering Sample
stepping : 8

Yep, looks like a "C" :)

Quote:

Originally Posted by Bok

I'll start running some tests if I can find the right software.

You could run Sandra and TMPGEnc on it. Both show a significant advantage for revision C and are available as test/demo versions. And everything else that uses SSE2 (and also SSE) could be tested - as long as it correctly detects the presence.

ScienceMark 2.0 (www.sciencemark.org) has some matrix multiply benchmarks (SGEMM, DGEMM) optimized for different architectures.

And most interesting: Prime95/mprime on a revision C! :) Plus the x87-only results (CpuSupportsSSE2=0).

Bok · 2003-09-05, 14:14

Quote:

Originally Posted by Prime95

To bok, who may have a revision C Opteron. If it is revision C run p95tst, do Advanced/Time 20000000, and post the strange timer results in results.txt.

I'm not sure whether I've run the right thing or not as I do not get a results.txt

I just ran mprime (what is p95tst ??) these are the results. Please let me know if this should be something different.

Main Menu

1. Test/Primenet
2. Test/User Information
3. Test/Vacation or Holiday
4. Test/Status
5. Test/Continue
6. Test/Exit
7. Advanced/Test
8. Advanced/Time
9. Advanced/P-1
10. Advanced/ECM
11. Advanced/Priority
12. Advanced/Manual Communication
13. Advanced/Unreserve Exponent
14. Advanced/Quit Gimps
15. Options/CPU
16. Options/Preferences
17. Options/Torture Test
18. Options/Benchmark
19. Help/About
20. Help/About PrimeNet Server
Your choice: 8

Exponent to time (10000000): 20000000
Number of Iterations (10):

Accept the answers above? (Y):
p: 20000000. Time: 76.813 ms.
p: 20000000. Time: 77.903 ms.
p: 20000000. Time: 78.012 ms.
p: 20000000. Time: 76.832 ms.
p: 20000000. Time: 76.813 ms.
p: 20000000. Time: 76.771 ms.
p: 20000000. Time: 76.808 ms.
p: 20000000. Time: 76.771 ms.
p: 20000000. Time: 76.889 ms.
p: 20000000. Time: 77.663 ms.
Iterations: 10. Total time: 0.771 sec.
Estimated time to complete this exponent: 17 days, 20 hours, 29 minutes.

Hit enter to continue:

Bok

Dresdenboy · 2003-09-05, 14:19

@bok:
You can find p95tst here: http://www.mersenne.org/gimps/p95tst.zip.

Bok · 2003-09-05, 14:24

yuk,

I'm only running linux I'm afraid, that's a win executable.

I've got a drive with XP (32bit) installed on it for the opteron, but as I'm at work (ssh'd ) I can't swap...

any linux tests I can do ??

Bok

Dresdenboy · 2003-09-05, 14:38

There are some small and quickly available benchmarks (like bytemark: http://www.tux.org/~mayer/linux/bmark.html) but they are compiler dependend.

Later you could run some tests on WinXP. It's not that urgent to know the results :)

I'm still at work, will go home soon (GMT +1). Then I'll have a look at some SSE2 stuff on Opteron.

gbvalor · 2003-09-05, 21:21

Quote:

BTW, would anyone like to volunteer to get some mlucas and glucas timings?

I've just made one in the opteron (our opteron) :)

These are the results for latest snapshot of Glucas v2.9.1
[code:1]
milsec/iter (user time)
FFT(k) round check
on / off
--- ---------
512 52/ 50
576 60/ 59
640 68/ 64
768 79/ 76
896 99/ 97
1024 108/105
1152 125/122
1280 140/135
1536 167/163
1792 206/205
2048 230/223
[/code:1]

The binary is made using SSE2 and the system compiler GCC 3.3.

You can try with a Revision C chip downloading my latest home snapshot

ftp://ftp.oxixares.com/glucas/glucas-2.9.1.tar.gz

Then a usual configure and make. To test, do a selftest

./Glucas -s p

And you will see the timings in 'selftest.out' file.

Guillermo

gbvalor · 2003-09-05, 21:45

Just for comparation, here are the results of my Athlon XP (Barton) BOX,

[code:1]
gbv@gauss:~/glucas/glucas-2.9.1> less selftest.out
You have new mail in /var/spool/mail/gbv
gbv@gauss:~/glucas/glucas-2.9.1> cat /proc/cpuinfo
processor : 0
vendor_id : AuthenticAMD
cpu family : 6
model : 10
model name : AMD Athlon(tm) XP 2500+
stepping : 0
cpu MHz : 1830.138
cache size : 512 KB
fdiv_bug : no
hlt_bug : no
f00f_bug : no
coma_bug : no
fpu : yes
fpu_exception : yes
cpuid level : 1
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 mmx fxsr sse syscall mmxext 3dnowext 3dnow
bogomips : 3643.80
[/code:1]

And the timings:
[code:1] milsec/iter (user time)
FFT(k) round check
on / off
--- ---------
512 53/ 50
576 63/ 60
640 74/ 70
768 88/ 84
896 107/100
1024 115/116
1152 138/130
1280 149/142
1536 181/172
1792 217/213
2048 238/228
[/code:1]

Guillermo

Bok · 2003-09-05, 23:16

Ok, ran that Glucas test (your link needs /pub/ btw)

I guess you cleaned up the results so I'll try and do the same, I think I'm interpreting it correctly.

milsec/iter (user time)
FFT(k) round check
on / off
--- ---------
512 43/40
576 52/48
640 56/51
768 66/62
896 80/76
1024 88/84
1152 107/100
1280 115/107
1536 137/130
1792 166/158
2048 183/175

I'll try making with -m64 -m128bit-long-double as well

Bok

Prime95 · 2003-09-05, 23:46

Bok, try this:

Run "mprime -m"
Choose 18
exit mprime
vi local.ini
add line "CpuSupportsSSE2=0"
repeat choice 18

Prime95 · 2003-09-05, 23:55

Bok, also please try this in a fresh directory

Get ftp://mersenne.org/gimps/mprtst.tar.gz

Run mprime -m
Choose 5
exit
delete worktodo.ini
Run mprime -m
Choose 8, exponent = 20000000, iterations = 10
exit

Post the results.txt file - full of weird timer numbers

Bok · 2003-09-06, 00:21

Quote:

Originally Posted by Prime95

Bok, try this:

Run "mprime -m"
Choose 18
exit mprime
vi local.ini
add line "CpuSupportsSSE2=0"
repeat choice 18

This is the results.txt for this one

AMD Engineering Sample
CPU speed: 1799.79 MHz
CPU features: RDTSC, CMOV, PREFETCH, 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 version 22.12, RdtscTiming=1
Best time for 256K FFT length: 16.347 ms.
Best time for 320K FFT length: 21.512 ms.
Best time for 384K FFT length: 25.947 ms.
Best time for 448K FFT length: 31.366 ms.
Best time for 512K FFT length: 34.787 ms.
Best time for 640K FFT length: 46.058 ms.
Best time for 768K FFT length: 56.882 ms.
Best time for 896K FFT length: 69.160 ms.
Best time for 1024K FFT length: 76.882 ms.
Best time for 1280K FFT length: 98.280 ms.
Best time for 1536K FFT length: 118.428 ms.
Best time for 1792K FFT length: 145.785 ms.

AMD Engineering Sample
CPU speed: 1800.20 MHz
CPU features: RDTSC, CMOV, PREFETCH, MMX, SSE
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 version 22.12, RdtscTiming=1
Best time for 256K FFT length: 17.196 ms.
Best time for 320K FFT length: 22.334 ms.
Best time for 384K FFT length: 29.183 ms.
Best time for 448K FFT length: 32.451 ms.
Best time for 512K FFT length: 35.374 ms.
Best time for 640K FFT length: 46.979 ms.
Best time for 768K FFT length: 56.599 ms.
Best time for 896K FFT length: 68.014 ms.
Best time for 1024K FFT length: 75.773 ms.
Best time for 1280K FFT length: 100.296 ms.
Best time for 1536K FFT length: 120.103 ms.
Best time for 1792K FFT length: 145.532 ms.

Bok · 2003-09-06, 00:24

Quote:

Originally Posted by Prime95

Bok, also please try this in a fresh directory

Get ftp://mersenne.org/gimps/mprtst.tar.gz

Run mprime -m
Choose 5
exit
delete worktodo.ini
Run mprime -m
Choose 8, exponent = 20000000, iterations = 10
exit

Post the results.txt file - full of weird timer numbers

And here is the results of this one....

Test 0: 0.000 sec. (121 clocks), avg: 0.000 sec. (127 clocks)
Test 1: 0.000 sec. (2131 clocks), avg: 0.000 sec. (2140 clocks)
Test 2: 0.000 sec. (654142 clocks), avg: 0.000 sec. (660652 clocks)
Test 3: 0.001 sec. (2623567 clocks), avg: 0.001 sec. (2685738 clocks)
Test 4: 0.002 sec. (2840212 clocks), avg: 0.002 sec. (2854997 clocks)
Test 1000: 0.000 sec. (528147 clocks), avg: 0.000 sec. (537882 clocks)
Test 1001: 0.001 sec. (1547728 clocks), avg: 0.001 sec. (1581690 clocks)
Test 1002: 0.001 sec. (2323676 clocks), avg: 0.001 sec. (2347559 clocks)
Test 1003: 0.000 sec. (525146 clocks), avg: 0.000 sec. (525155 clocks)
Test 1004: 0.001 sec. (2460613 clocks), avg: 0.001 sec. (2470975 clocks)
Test 1005: 0.002 sec. (4224755 clocks), avg: 0.002 sec. (4397632 clocks)
Test 1006: 0.000 sec. (528147 clocks), avg: 0.000 sec. (529015 clocks)
Test 1007: 0.001 sec. (2485652 clocks), avg: 0.001 sec. (2512956 clocks)
Test 1008: 0.002 sec. (3813135 clocks), avg: 0.002 sec. (4017296 clocks)
Test 1009: 0.001 sec. (1040145 clocks), avg: 0.001 sec. (1042419 clocks)
Test 1010: 0.003 sec. (4907879 clocks), avg: 0.003 sec. (4914776 clocks)
Test 1011: 0.003 sec. (4732079 clocks), avg: 0.003 sec. (4976143 clocks)
Test 1012: 0.000 sec. (504955 clocks), avg: 0.000 sec. (505812 clocks)
Test 1013: 0.000 sec. (831666 clocks), avg: 0.000 sec. (834674 clocks)
Test 1014: 0.001 sec. (1120767 clocks), avg: 0.001 sec. (1142366 clocks)
Test 1015: 0.001 sec. (1581311 clocks), avg: 0.001 sec. (1590955 clocks)
Test 1016: 0.001 sec. (1290771 clocks), avg: 0.001 sec. (1312691 clocks)
Test 1017: 0.001 sec. (1901221 clocks), avg: 0.001 sec. (1910335 clocks)
Test 1018: 0.001 sec. (1152645 clocks), avg: 0.001 sec. (1154434 clocks)
Test 1019: 0.001 sec. (2025428 clocks), avg: 0.001 sec. (2029746 clocks)
[Fri Sep 5 16:23:34 2003]
timer 0: 72921256
timer 1: 57044016
timer 2: 579204
timer 3: 72340708
timer 4: 10620074
timer 5: 6002809
timer 6: 13630747
timer 9: 12004685
timer 10: 13410368
timer 13: 7685964
timer 14: 7410400
timer 16: 9099429
timer 17: 5984708
timer 18: 8094904
timer 20: 12457308
timer 21: 13360638
timer 24: 7918447
timer 26: 18079
timer 27: 36327

Bok

GP2 · 2003-09-06, 03:14

Bok, perhaps you ought to try mprime version 23.5 instead of 22.12?

Bok · 2003-09-06, 04:42

ok, ran it again with version 23.5

AMD Engineering Sample
CPU speed: 1799.80 MHz
CPU features: RDTSC, CMOV, PREFETCH, 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 version 23.5, RdtscTiming=1
Best time for 384K FFT length: 25.615 ms.
Best time for 448K FFT length: 30.690 ms.
Best time for 512K FFT length: 34.733 ms.
Best time for 640K FFT length: 43.128 ms.
Best time for 768K FFT length: 52.786 ms.
Best time for 896K FFT length: 63.544 ms.
Best time for 1024K FFT length: 71.629 ms.
Best time for 1280K FFT length: 95.534 ms.
Best time for 1536K FFT length: 116.598 ms.
Best time for 1792K FFT length: 140.507 ms.
Best time for 2048K FFT length: 158.316 ms.
[Fri Sep 5 20:38:36 2003]
Compare your results to other computers at http://www.mersenne.org/bench.htm
That web page also contains instructions on how your results can be included.

AMD Engineering Sample
CPU speed: 1799.76 MHz
CPU features: RDTSC, CMOV, PREFETCH, MMX, SSE
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 version 23.5, RdtscTiming=1
Best time for 384K FFT length: 28.408 ms.
Best time for 448K FFT length: 32.031 ms.
Best time for 512K FFT length: 34.807 ms.
Best time for 640K FFT length: 47.125 ms.
Best time for 768K FFT length: 55.922 ms.
Best time for 896K FFT length: 67.866 ms.
Best time for 1024K FFT length: 74.954 ms.
Best time for 1280K FFT length: 96.482 ms.
Best time for 1536K FFT length: 117.145 ms.
Best time for 1792K FFT length: 139.925 ms.
Best time for 2048K FFT length: 156.999 ms.

Should it have been slower ???

Bok

GP2 · 2003-09-06, 05:39

Quote:

Originally Posted by Bok

ok, ran it again with version 23.5

Should it have been slower ???

Bok

When comparing results, note that version 22 output starts at 256K FFT and ends at 1792K, while version 23 output starts at 384K and ends at 2048K.

So your results show that version 23.5 is actually a bit faster than 22.12 for your Opteron.

gbvalor · 2003-09-06, 08:03

It seems there is no significative advantage using Revision C for Glucas. Actually, the timings are near the same. The version Bok tested is a bit slower because I supressed an optimization causing problems in other targets.

Other interesting thing is that Glucas and mprime are closest than ever

. Well, when magic George's hands touch again the code this will disapear :) .

And I also have work to do with Glucas ...

Quote:

I'll try making with -m64 -m128bit-long-double as well

Bok, I don't think it will run, at least it will not give any advantage.

Guillermo

2003-09-05, 03:33	#155
Xyzzy Aug 2002 8,563 Posts	Looks like P-1 time! Code: PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ Command 3112 gw 39 19 359m 359m 736 R 99.9 72.5 691:50.32 ./mprime

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Opteron is Hyperthreaded ?	bgbeuning	Information & Answers	3	2016-01-10 08:26
Opteron web server...	Xyzzy	Lounge	14	2003-11-05 23:07
Opteron Bottleneck??	Prime95	Hardware	31	2003-09-17 06:54
AMD Opteron	naclosagc	Software	27	2003-08-10 19:14
What will an AMD Opteron be classified as ?	dsouza123	Software	4	2003-08-02 14:29

2003-09-05, 02:36	#153
Prime95 P90 years forever! Aug 2002 Yeehaw, FL 2³·1,021 Posts	BTW, would anyone like to volunteer to get some mlucas and glucas timings?

2003-09-05, 09:16	#158
Xyzzy Aug 2002 8,563 Posts	$925.13 in donations, $835.32 spent, and $89.81 left... I just got around to ordering the CD-ROM... It was $16.99 plus $5 for delivery... http://www.newegg.com/app/viewProduct.asp?description=27-101-204&depa=1

2003-09-05, 13:56	#159
Bok Aug 2003 5² Posts	Back in town today, this is my cpuinfo Opteron64:~ # cat /proc/cpuinfo processor : 0 vendor_id : AuthenticAMD cpu family : 15 model : 5 model name : AMD Engineering Sample stepping : 8 cpu MHz : 1800.028 cache size : 1024 KB fpu : yes fpu_exception : yes cpuid level : 1 wp : yes flags : fpu vme de pse tsc msr pae mce cx8 sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 syscall nx mmxext lm 3dnowext 3dnow bogomips : 3591.37 TLB size : 1088 4K pages clflush size : 64 address sizes : 40 bits physical, 48 bits virtual power management: ts ttp so I guess it is a 'C' I'll start running some tests if I can find the right software.

2003-09-05, 14:19	#162
Dresdenboy Apr 2003 Berlin, Germany 19² Posts	@bok: You can find p95tst here: http://www.mersenne.org/gimps/p95tst.zip.

2003-09-05, 14:24	#163
Bok Aug 2003 19₁₆ Posts	yuk, I'm only running linux I'm afraid, that's a win executable. I've got a drive with XP (32bit) installed on it for the opteron, but as I'm at work (ssh'd ) I can't swap... any linux tests I can do ?? Bok

2003-09-05, 14:38	#164
Dresdenboy Apr 2003 Berlin, Germany 169₁₆ Posts	There are some small and quickly available benchmarks (like bytemark: http://www.tux.org/~mayer/linux/bmark.html) but they are compiler dependend. Later you could run some tests on WinXP. It's not that urgent to know the results :) I'm still at work, will go home soon (GMT +1). Then I'll have a look at some SSE2 stuff on Opteron.

2003-09-05, 21:45	#166
gbvalor Aug 2002 3×37 Posts	Just for comparation, here are the results of my Athlon XP (Barton) BOX, [code:1] gbv@gauss:~/glucas/glucas-2.9.1> less selftest.out You have new mail in /var/spool/mail/gbv gbv@gauss:~/glucas/glucas-2.9.1> cat /proc/cpuinfo processor : 0 vendor_id : AuthenticAMD cpu family : 6 model : 10 model name : AMD Athlon(tm) XP 2500+ stepping : 0 cpu MHz : 1830.138 cache size : 512 KB fdiv_bug : no hlt_bug : no f00f_bug : no coma_bug : no fpu : yes fpu_exception : yes cpuid level : 1 wp : yes flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 mmx fxsr sse syscall mmxext 3dnowext 3dnow bogomips : 3643.80 [/code:1] And the timings: [code:1] milsec/iter (user time) FFT(k) round check on / off --- --------- 512 53/ 50 576 63/ 60 640 74/ 70 768 88/ 84 896 107/100 1024 115/116 1152 138/130 1280 149/142 1536 181/172 1792 217/213 2048 238/228 [/code:1] Guillermo

2003-09-05, 23:16	#167
Bok Aug 2003 5² Posts	Ok, ran that Glucas test (your link needs /pub/ btw) I guess you cleaned up the results so I'll try and do the same, I think I'm interpreting it correctly. milsec/iter (user time) FFT(k) round check on / off --- --------- 512 43/40 576 52/48 640 56/51 768 66/62 896 80/76 1024 88/84 1152 107/100 1280 115/107 1536 137/130 1792 166/158 2048 183/175 I'll try making with -m64 -m128bit-long-double as well Bok

2003-09-05, 23:46	#168
Prime95 P90 years forever! Aug 2002 Yeehaw, FL 1FE8₁₆ Posts	Bok, try this: Run "mprime -m" Choose 18 exit mprime vi local.ini add line "CpuSupportsSSE2=0" repeat choice 18

2003-09-05, 23:55	#169
Prime95 P90 years forever! Aug 2002 Yeehaw, FL 8168₁₀ Posts	Bok, also please try this in a fresh directory Get ftp://mersenne.org/gimps/mprtst.tar.gz Run mprime -m Choose 5 exit delete worktodo.ini Run mprime -m Choose 8, exponent = 20000000, iterations = 10 exit Post the results.txt file - full of weird timer numbers

2003-09-06, 03:14	#172
GP2 Sep 2003 3×863 Posts	Bok, perhaps you ought to try mprime version 23.5 instead of 22.12?

2003-09-06, 04:42	#173
Bok Aug 2003 5² Posts	ok, ran it again with version 23.5 AMD Engineering Sample CPU speed: 1799.80 MHz CPU features: RDTSC, CMOV, PREFETCH, 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 version 23.5, RdtscTiming=1 Best time for 384K FFT length: 25.615 ms. Best time for 448K FFT length: 30.690 ms. Best time for 512K FFT length: 34.733 ms. Best time for 640K FFT length: 43.128 ms. Best time for 768K FFT length: 52.786 ms. Best time for 896K FFT length: 63.544 ms. Best time for 1024K FFT length: 71.629 ms. Best time for 1280K FFT length: 95.534 ms. Best time for 1536K FFT length: 116.598 ms. Best time for 1792K FFT length: 140.507 ms. Best time for 2048K FFT length: 158.316 ms. [Fri Sep 5 20:38:36 2003] Compare your results to other computers at http://www.mersenne.org/bench.htm That web page also contains instructions on how your results can be included. AMD Engineering Sample CPU speed: 1799.76 MHz CPU features: RDTSC, CMOV, PREFETCH, MMX, SSE 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 version 23.5, RdtscTiming=1 Best time for 384K FFT length: 28.408 ms. Best time for 448K FFT length: 32.031 ms. Best time for 512K FFT length: 34.807 ms. Best time for 640K FFT length: 47.125 ms. Best time for 768K FFT length: 55.922 ms. Best time for 896K FFT length: 67.866 ms. Best time for 1024K FFT length: 74.954 ms. Best time for 1280K FFT length: 96.482 ms. Best time for 1536K FFT length: 117.145 ms. Best time for 1792K FFT length: 139.925 ms. Best time for 2048K FFT length: 156.999 ms. Should it have been slower ??? Bok