Prime95 version 26.3

Prime95 · 2010-10-13, 03:04

Prime95 · 2010-10-13, 03:04

Known bugs and fixes:

1) The length 160K x87 FFT crashes. Fixed in 26.3 build 2.
2) The length 4M FFT crashes on Northwood and Willamette Pentium 4. Fixed in 26.3 build 3.
3) On a brand new install, the default on hyperthreaded machines was 2 threads per worker. This was slower for most users. In the next release, the default will be 1 thread per worker.
4) Testing 522*928^24576-1 in 64-bit executables crashes. Fixed in 26.4.
5) You cannot edit the affinity settings from the Linux menus. Fixed in 26.4.
6) Running a multi-threaded torture test on AMD machines failed. Fixed in 26.4.
7) Setting a specific FFT length in worktodo.ini sometimes selects the next larger FFT length. This would most commonly occur when an exponent being tested is near the FFT crossover point and a specific FFT length is automatically chosen based on the roundoff error in the first 1000 iterations. Fixed in 26.4.
8) Multi-line additional information from the Primenet server was not parse properly. Only the first line was parsed. Thus, when successfully completing a double-check only the successful completion was written to prime.log. The cpu credit amount was not written to prime.log. Fixed in 26.4.
9) When doing PRP on non-base-2 numbers the time estimate in Test/Status was off by a factor of log2(base). Fixed in 26.4.
10) On a 32-bit Mac, some FFT implementations were missing. Fixed in next release.

harlee · 2010-10-13, 22:41

Benchmark still crashing, now when starting 4096K FFT Length:

[Wed Oct 13 18:37:43 2010]
Compare your results to other computers at http://www.mersenne.org/report_benchmarks
Intel(R) Pentium(R) 4 CPU 2.60GHz
CPU speed: 2593.88 MHz, with hyperthreading
CPU features: RDTSC, CMOV, Prefetch, MMX, SSE, SSE2
L1 cache size: 8 KB
L2 cache size: 512 KB
L1 cache line size: 64 bytes
L2 cache line size: 128 bytes
TLBS: 64
Prime95 32-bit version 26.3, RdtscTiming=1
Best time for 768K FFT length: 25.541 ms., avg: 26.312 ms.
Best time for 896K FFT length: 31.136 ms., avg: 31.494 ms.
Best time for 1024K FFT length: 35.015 ms., avg: 35.479 ms.
Best time for 1280K FFT length: 44.310 ms., avg: 44.916 ms.
Best time for 1536K FFT length: 54.100 ms., avg: 54.545 ms.
Best time for 1792K FFT length: 65.006 ms., avg: 65.448 ms.
Best time for 2048K FFT length: 72.573 ms., avg: 73.009 ms.
Best time for 2560K FFT length: 95.657 ms., avg: 96.104 ms.
Best time for 3072K FFT length: 118.269 ms., avg: 119.040 ms.
Best time for 3584K FFT length: 149.191 ms., avg: 150.035 ms.

CCol · 2010-10-14, 01:33

I've downloaded and installed this version because it fixes the ubuntu trial factoring problem, however primenet says this version is untrusted. I get an error 9. untrusted version How do I convince primenet this version is ok? TIA

Prime95 · 2010-10-14, 02:11

Quote:

Originally Posted by CCol

I've downloaded and installed this version because it fixes the ubuntu trial factoring problem, however primenet says this version is untrusted. I get an error 9. untrusted version How do I convince primenet this version is ok? TIA

That problem goes away eventually. I've no idea why it happens.

Prime95 · 2010-10-14, 03:30

Quote:

Originally Posted by harlee

Benchmark still crashing, now when starting 4096K FFT Length:

Fixed in build 3 of 26.3. I've uploaded Windows version for you.

CCol · 2010-10-14, 10:38

26.3 has found a factor, so the problem with it not reporting found factors to the server under ubuntu 10.04 is fixed.

However I had a i7 920 and mprime 26.3 will only recognise 4 workers, whereas the old 25.11 had no problem recognising all 8 threads.

I read the undocumented.txt file and added the number of cpus and hyperthreading=2 lines to the local.txt file, but it has no effect.

mprime adds the line oldcpuspeed=2699 for my processor as well, even if I already have the cpuspeed line in local.txt

the line computerID= was blank, so I tried to add a computer name so it would show up in the reports, but this has also had no effect.

When I upgraded to 26.3, I kept the old prime.txt file, but all the other files are new either from the zip file or created by 26.3.

Any help appreciated.

petrw1 · 2010-10-14, 14:39

Quote:

Originally Posted by CCol

However I had a i7 920 and mprime 26.3 will only recognise 4 workers, whereas the old 25.11 had no problem recognising all 8 threads.

the line computerID= was blank, so I tried to add a computer name so it would show up in the reports, but this has also had no effect.

Any help appreciated.

1. All that I have read here and experienced suggests that hyperthreading does NOT really double your number of CPUs and it is almost always best to only use 4 workers and then experiment with whether or not you want to use 2 CPUs with any of the workers.

2. I have only ever set the ComputerId= using the GUI
- Test
- Primenet...
- Optional Computer Name

ixfd64 · 2010-10-14, 17:45

Does anyone know if the 32-core limitation has been removed?

harlee · 2010-10-14, 19:29

Quote:

Originally Posted by Prime95

Fixed in build 3 of 26.3. I've uploaded Windows version for you.

Thanks, I've ran the benchmark without any problems. Need to run it a little later when I can stop doing everything to get some better timings.

joblack · 2010-10-14, 21:22

Quote:

Originally Posted by ixfd64

Does anyone know if the 32-core limitation has been removed?

I don't think so but I would be interested in that as well.

2010-10-13, 03:04	#1
Prime95 P90 years forever! Aug 2002 Yeehaw, FL 2²·17·109 Posts	Prime95 version 26.3 Version 26.3 is ready for testing. This version should be safe to use. This version fixes the few serious bugs that were reported for 26.2. Download links: Windows: ftp://mersenne.org/gimps/p95v263.zip Windows 64-bit: ftp://mersenne.org/gimps/p64v263.zip Linux: ftp://mersenne.org/gimps/mprime263.tar.gz Linux 64-bit: ftp://mersenne.org/gimps/mprime263-linux64.tar.gz Mac OS X: ftp://mersenne.org/gimps/Prime95-MacOSX-263.zip FreeBSD: ftp://mersenne.org/gimps/mprime263-FreeBSD.tar.gz FreeBSD 64-bit: ftp://mersenne.org/gimps/mprime263-FreeBSD64.tar.gz Windows NT service: ftp://mersenne.org/gimps/winnt263.zip Windows NT service 64-bit: ftp://mersenne.org/gimps/win64nt263.zip Source: ftp://mersenne.org/gimps/source263.zip Bug fixes are described here: http://www.mersenneforum.org/showpos...81&postcount=2 Last fiddled with by Prime95 on 2010-10-13 at 20:35

2010-10-13, 03:04	#2
Prime95 P90 years forever! Aug 2002 Yeehaw, FL 2²·17·109 Posts	Known bugs and fixes: 1) The length 160K x87 FFT crashes. Fixed in 26.3 build 2. 2) The length 4M FFT crashes on Northwood and Willamette Pentium 4. Fixed in 26.3 build 3. 3) On a brand new install, the default on hyperthreaded machines was 2 threads per worker. This was slower for most users. In the next release, the default will be 1 thread per worker. 4) Testing 522928^24576-1 in 64-bit executables crashes. Fixed in 26.4. 5) You cannot edit the affinity settings from the Linux menus. Fixed in 26.4. 6) Running a multi-threaded torture test on AMD machines failed. Fixed in 26.4. 7) Setting a specific FFT length in worktodo.ini sometimes selects the next larger FFT length. This would most commonly occur when an exponent being tested is near the FFT crossover point and a specific FFT length is automatically chosen based on the roundoff error in the first 1000 iterations. Fixed in 26.4. 8) Multi-line additional information from the Primenet server was not parse properly. Only the first line was parsed. Thus, when successfully completing a double-check only the successful completion was written to prime.log. The cpu credit amount was not written to prime.log. Fixed in 26.4. 9) When doing PRP on non-base-2 numbers the time estimate in Test/Status was off by a factor of log2(base). Fixed in 26.4. 10) On a 32-bit Mac, some FFT implementations were missing. Fixed in next release. Last fiddled with by Prime95 on 2010-11-08 at 14:50*

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Prime95 version 29.4	Prime95	Software	441	2020-02-16 15:18
Prime95 version 27.3	Prime95	Software	148	2012-03-18 19:24
Prime95 version 25.5	Prime95	PrimeNet	369	2008-02-26 05:21
Prime95 version 25.4	Prime95	PrimeNet	143	2007-09-24 21:01
When the next prime95 version ?	pacionet	Software	74	2006-12-07 20:30

2010-10-13, 22:41	#3
harlee Sep 2006 Odenton, MD, USA 2²×41 Posts	Benchmark still crashing, now when starting 4096K FFT Length: [Wed Oct 13 18:37:43 2010] Compare your results to other computers at http://www.mersenne.org/report_benchmarks Intel(R) Pentium(R) 4 CPU 2.60GHz CPU speed: 2593.88 MHz, with hyperthreading CPU features: RDTSC, CMOV, Prefetch, MMX, SSE, SSE2 L1 cache size: 8 KB L2 cache size: 512 KB L1 cache line size: 64 bytes L2 cache line size: 128 bytes TLBS: 64 Prime95 32-bit version 26.3, RdtscTiming=1 Best time for 768K FFT length: 25.541 ms., avg: 26.312 ms. Best time for 896K FFT length: 31.136 ms., avg: 31.494 ms. Best time for 1024K FFT length: 35.015 ms., avg: 35.479 ms. Best time for 1280K FFT length: 44.310 ms., avg: 44.916 ms. Best time for 1536K FFT length: 54.100 ms., avg: 54.545 ms. Best time for 1792K FFT length: 65.006 ms., avg: 65.448 ms. Best time for 2048K FFT length: 72.573 ms., avg: 73.009 ms. Best time for 2560K FFT length: 95.657 ms., avg: 96.104 ms. Best time for 3072K FFT length: 118.269 ms., avg: 119.040 ms. Best time for 3584K FFT length: 149.191 ms., avg: 150.035 ms.

2010-10-14, 01:33	#4
CCol Jul 2007 43₈ Posts	I've downloaded and installed this version because it fixes the ubuntu trial factoring problem, however primenet says this version is untrusted. I get an error 9. untrusted version How do I convince primenet this version is ok? TIA

2010-10-14, 10:38	#7
CCol Jul 2007 5×7 Posts	26.3 has found a factor, so the problem with it not reporting found factors to the server under ubuntu 10.04 is fixed. However I had a i7 920 and mprime 26.3 will only recognise 4 workers, whereas the old 25.11 had no problem recognising all 8 threads. I read the undocumented.txt file and added the number of cpus and hyperthreading=2 lines to the local.txt file, but it has no effect. mprime adds the line oldcpuspeed=2699 for my processor as well, even if I already have the cpuspeed line in local.txt the line computerID= was blank, so I tried to add a computer name so it would show up in the reports, but this has also had no effect. When I upgraded to 26.3, I kept the old prime.txt file, but all the other files are new either from the zip file or created by 26.3. Any help appreciated.

2010-10-14, 17:45	#9
ixfd64 Bemusing Prompter "Danny" Dec 2002 California 2,377 Posts	Does anyone know if the 32-core limitation has been removed?