mersenneforum.org

mersenneforum.org (https://www.mersenneforum.org/index.php)
-   3*2^n-1 Search (https://www.mersenneforum.org/forumdisplay.php?f=14)
-   -   Can not be a good thing :( (https://www.mersenneforum.org/showthread.php?t=2337)

SB2 2004-04-07 21:18

Can not be a good thing :(
 
Started getting these errors this morning, appears to happen every 3rd or 4th number.

Wed Apr 07 05:01:03 2004]
Using IBDWT : Mersenne fftlen = 40960, Proth fftlen = 98304, Used fftlen = 49152
fftlen seems to be too small, using next fftlen...
Using IBDWT : Mersenne fftlen = 40960, Proth fftlen = 98304, Used fftlen = 57344
[Wed Apr 07 05:06:58 2004]
fftlen seems to be too small, using next fftlen...
Using IBDWT : Mersenne fftlen = 40960, Proth fftlen = 98304, Used fftlen = 65536
[Wed Apr 07 05:20:20 2004]
Iter: 145688/842577, ERROR: SUM(INPUTS) != SUM(OUTPUTS), 1954302673964.291 != 1954302673956.29
Possible hardware failure, consult the readme file.
Continuing from last save file.
[Wed Apr 07 05:25:21 2004]
Using IBDWT : Mersenne fftlen = 40960, Proth fftlen = 98304, Used fftlen = 49152
[Wed Apr 07 06:20:42 2004]
3*2^842578-1 is not prime. Res64: 20621E388A72C72E Time : 3321.330 sec.
Using IBDWT : Mersenne fftlen = 40960, Proth fftlen = 98304, Used fftlen = 49152
[Wed Apr 07 07:16:24 2004]
3*2^842606-1 is not prime. Res64: DBB18BC928621BF8 Time : 3342.324 sec.
Using IBDWT : Mersenne fftlen = 40960, Proth fftlen = 98304, Used fftlen = 49152
[Wed Apr 07 08:11:49 2004]
3*2^842616-1 is not prime. Res64: 1A3D147214EA304E Time : 3325.003 sec.
Using IBDWT : Mersenne fftlen = 40960, Proth fftlen = 98304, Used fftlen = 49152
[Wed Apr 07 08:49:17 2004]
fftlen seems to be too small, using next fftlen...
Using IBDWT : Mersenne fftlen = 40960, Proth fftlen = 98304, Used fftlen = 57344
[Wed Apr 07 08:58:59 2004]
fftlen seems to be too small, using next fftlen...
Using IBDWT : Mersenne fftlen = 40960, Proth fftlen = 98304, Used fftlen = 65536
[Wed Apr 07 09:08:47 2004]
Iter: 106548/842658, ERROR: SUM(INPUTS) != SUM(OUTPUTS), 175223027041.6848 != 175223027169.6853
Possible hardware failure, consult the readme file.
Continuing from last save file.

paulunderwood 2004-04-08 00:09

This look like overheating of your CPU.

SB2 2004-04-08 02:18

[QUOTE=paulunderwood]This look like overheating of your CPU.[/QUOTE]

Darn. :sad:

Prime95 2004-04-08 03:00

You may be all right on the errors. Wait for Jean to weigh in. If he changed the DWT weights and did not change the error reporting threshold, then the program may just be giving you spurious messages. The numbers are pretty close.

SB2 2004-04-08 10:21

[QUOTE=Prime95]You may be all right on the errors. Wait for Jean to weigh in. If he changed the DWT weights and did not change the error reporting threshold, then the program may just be giving you spurious messages. The numbers are pretty close.[/QUOTE]

Thanks for the reply George. I opened the case to let a little more air circulate around the case and the errors seem to have disappeared (keeping fingers crossed for now), so perhaps it was an over-heating issue. I will keep a close watch on it for a few days.

paulunderwood 2004-06-25 10:17

I have been doing some double checking of files that contain "SUM(INPUTS)!=SUM(OUTPUTS)". These are likely due to hardware failures.

Firstly, I found 6 consecutive erroneous Res64's in James' bad file. It looks like James cured a overheating problem. :furious:

Next, I checked a file which was run on a very dodgy computer ( Nuno Simões ) in a hot Portugese summer last year. This file contained many SUM errors and it was not really supprising that about half of the Res64's were wrong :sick:

Three SUM errors checked correct from three separate files from Luigi Morelli :innocent:

A SUM error occurred in a file by yours truely which did not match -- I could not find any more error in a sample from this file. :ermm:

Markus Frind's last file contained 4 SUM errors -- 3 matched and 1 didn't -- I will recheck this file completely at future date. :geek:

paulunderwood 2004-09-22 21:37

I checked Markus' last file and found several mismatches. Markus has just informed me that it was "bad memory".

While looking at Markus' results I discovered a new type of ERROR: ILLEGAL SUMOUT. I found this error in one other file which was done by me. I am rechecking this file...

paulunderwood 2004-09-23 08:48

I have checked the ILLEGAL SUMOUT error and there was a mismatch. There were no more errors from a sample of the affected file :ermm:

I also found a couple of errors in the n=1-2m work, both by me, in the same file. I checked these and the residues match. I do not expect any more errors from the computer that caused it because shortly after the errors occurred I replaced the DDR400 memory for much more stable DDR333. :grin:


All times are UTC. The time now is 05:56.

Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2022, Jelsoft Enterprises Ltd.