Reproducable hardware errors?

cmokruhl · 2002-09-13, 17:53

I've gotten this sort of message on and off for a while. I've always run the most up-to-date client and am currently running 22.9. The systems that seem to have this problem are P3-933s running RedHat 6.2 (2.2.14-5.0).

[code:1]Resuming primality test of M15805927 at iteration 10193505 [64.491%]
Iteration: 10193505/15805927, ERROR: SUM(INPUTS) != SUM(OUTPUTS), -2.152232271144956e+17 != -2.152232271144956e+17
Possible hardware failure, consult the readme.txt file.
Continuing from last save file.
Waiting five minutes before restarting.
Resuming primality test of M15805927 at iteration 10193505 [64.491%]
Iteration: 10193505/15805927, ERROR: SUM(INPUTS) != SUM(OUTPUTS), 1.619612012313178e+19 != 1.619612012313178e+19
Possible hardware failure, consult the readme.txt file.
Continuing from last save file.
Waiting five minutes before restarting.
Resuming primality test of M15805927 at iteration 10193505 [64.491%]
Disregard last error. Result is reproducible and thus not a hardware problem.
Disregard last error. Result is reproducible and thus not a hardware problem.
For added safety, redoing iteration using a slower, more reliable method.
Continuing from last save file.
Resuming primality test of M15805927 at iteration 10193505 [64.491%]
Iteration: 10193505/15805927, ERROR: SUM(INPUTS) != SUM(OUTPUTS), -2.152232271144956e+17 != -2.152232271144956e+17
Possible hardware failure, consult the readme.txt file.
[/code:1]
Mprime gets stuck at this iteration. My only solution in the past was to unreserve the exponent and try a new one, but this obviously results in wasted CPU time. Any ideas?

PageFault · 2002-09-17, 17:28

It would appear that you have a hardware problem. The " reproducible hardware problem " is something that was found with SSE2 being too close to a FFT crossover point. It was added specifically for those, which are not a real problem.

You are getting all sorts of error messages. Can you copy the checkpoint file to a known good machine and try it there?

Are you overclocking? That would put you really close to the limit of the coppermines useful frequency. Can you do some torture testing and pass a days worth? How about trying an earlier version of the client?

Prime95 · 2002-09-17, 19:04

Did you save the pF805927 file? If so, can you email the file to me? Are you using a P4?

2002-09-13, 17:53	#1
cmokruhl Sep 2002 2·3 Posts	Reproducable hardware errors? I've gotten this sort of message on and off for a while. I've always run the most up-to-date client and am currently running 22.9. The systems that seem to have this problem are P3-933s running RedHat 6.2 (2.2.14-5.0). [code:1]Resuming primality test of M15805927 at iteration 10193505 [64.491%] Iteration: 10193505/15805927, ERROR: SUM(INPUTS) != SUM(OUTPUTS), -2.152232271144956e+17 != -2.152232271144956e+17 Possible hardware failure, consult the readme.txt file. Continuing from last save file. Waiting five minutes before restarting. Resuming primality test of M15805927 at iteration 10193505 [64.491%] Iteration: 10193505/15805927, ERROR: SUM(INPUTS) != SUM(OUTPUTS), 1.619612012313178e+19 != 1.619612012313178e+19 Possible hardware failure, consult the readme.txt file. Continuing from last save file. Waiting five minutes before restarting. Resuming primality test of M15805927 at iteration 10193505 [64.491%] Disregard last error. Result is reproducible and thus not a hardware problem. Disregard last error. Result is reproducible and thus not a hardware problem. For added safety, redoing iteration using a slower, more reliable method. Continuing from last save file. Resuming primality test of M15805927 at iteration 10193505 [64.491%] Iteration: 10193505/15805927, ERROR: SUM(INPUTS) != SUM(OUTPUTS), -2.152232271144956e+17 != -2.152232271144956e+17 Possible hardware failure, consult the readme.txt file. [/code:1] Mprime gets stuck at this iteration. My only solution in the past was to unreserve the exponent and try a new one, but this obviously results in wasted CPU time. Any ideas?

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Possible hardware errors have occurred during the test! 1 ROUNDOFF > 0.4.	Xyzzy	Software	7	2016-12-20 00:01
Possible hardware errors...	SverreMunthe	Hardware	16	2013-08-19 14:39
Warning of "possible hardware errors"	GARYP166	Information & Answers	4	2010-11-17 17:16
Hardware, FFT limits and round off errors	ewergela	Hardware	9	2005-09-01 14:51
more about hardware errors	graeme	Hardware	4	2003-07-08 09:14

2002-09-17, 17:28	#2
PageFault Aug 2002 Dawn of the Dead 5·47 Posts	It would appear that you have a hardware problem. The " reproducible hardware problem " is something that was found with SSE2 being too close to a FFT crossover point. It was added specifically for those, which are not a real problem. You are getting all sorts of error messages. Can you copy the checkpoint file to a known good machine and try it there? Are you overclocking? That would put you really close to the limit of the coppermines useful frequency. Can you do some torture testing and pass a days worth? How about trying an earlier version of the client?

2002-09-17, 19:04	#3
Prime95 P90 years forever! Aug 2002 Yeehaw, FL 2²·2,089 Posts	Did you save the pF805927 file? If so, can you email the file to me? Are you using a P4?