mersenneforum.org  

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

Reply
 
Thread Tools
Old 2017-03-18, 04:00   #1
stebbo
 
"Chris"
Jul 2016
Melbourne

29 Posts
Default Suspect result

Hi All,

one of my machines has just returned a C-Suspect result on 78077873.

The log simply says "Computer hardware check recommended. Possible hardware errors occured during LL test".

I did reimage this machine during the course of the test and restored the test from backup and resumed.

What is the advised course of action here?

Cheers,
Chris.
stebbo is offline   Reply With Quote
Old 2017-03-18, 05:17   #2
Prime95
P90 years forever!
 
Prime95's Avatar
 
Aug 2002
Yeehaw, FL

7×1,019 Posts
Default

What are the relevant lines from results.txt?
Prime95 is offline   Reply With Quote
Old 2017-03-18, 22:45   #3
stebbo
 
"Chris"
Jul 2016
Melbourne

2910 Posts
Default

[Wed Jan 18 23:36:33 2017]
Iteration: 79644781/79644781, ERROR: Illegal double encountered.
Possible hardware failure, consult readme.txt file.
Continuing from last save file.
[Wed Jan 18 23:46:46 2017]
UID: M79644781 is not prime. Res64: 4CCC9398DFF06D__. We5: 7A865DC6,32379349,00010000, AID: ...
[Mon Feb 06 00:46:26 2017]
UID: M79706987 is not prime. Res64: 6DA1D6F732701D__. We5: EC0BD406,48100621,00000000, AID: ...
Trying 1000 iterations for exponent 78077873 using 4096K FFT.
If average roundoff error is above 0.243, then a larger FFT will be used.
Final average roundoff error is 0.23104, using 4096K FFT for exponent 78077873.
[Mon Feb 13 04:07:57 2017]
Iteration: 79748159/79748159, ERROR: Illegal double encountered.
Possible hardware failure, consult readme.txt file.
Continuing from last save file.
[Mon Feb 13 04:19:30 2017]
UID: M79748159 is not prime. Res64: 7F524C2E19928B__. We5: 4EB9D286,40226708,00010000, AID: __
[Sun Feb 26 09:17:28 2017]
Iteration 59749240 / 78077873
Iteration 41587931 / 78195713
[Sun Feb 26 20:01:24 2017]
Iteration: 61002499/78077873, Possible error: round off (0.4966702308) > 0.40625
Continuing from last save file.
[Sun Mar 12 11:02:24 2017]
Iteration 45392504 / 78195713
Iteration 63076779 / 78077873
[Tue Mar 14 15:17:49 2017]
Iteration 53147638 / 78195713
Iteration 69742153 / 78077873
[Tue Mar 14 15:30:14 2017]
Iteration 69764057 / 78077873
Iteration 53172524 / 78195713
[Fri Mar 17 09:35:10 2017]
UID: M78077873 is not prime. Res64: 6BB8447D8C41DF__. We5: 098710A2,63901602,00000100, AID: __

Last fiddled with by Batalov on 2017-03-19 at 00:14 Reason: masked last bits of Res64
stebbo is offline   Reply With Quote
Old 2017-03-18, 22:57   #4
VictordeHolland
 
VictordeHolland's Avatar
 
"Victor de Hollander"
Aug 2011
the Netherlands

2·587 Posts
Default

Somebody please obscure the residues, just in case.
VictordeHolland is offline   Reply With Quote
Old 2017-03-19, 00:25   #5
Prime95
P90 years forever!
 
Prime95's Avatar
 
Aug 2002
Yeehaw, FL

7×1,019 Posts
Default

The Illegal double errors are very rare - I've not seen them before! They can only happen on the last iteration. The 0.4966 roundoff error is almost definitely a hardware glitch.

Standard procedure from here is to check your memory with memtest. Run torture tests. Reduce overclocking. Increase voltages. Check temperatures. Etc. There's no way to tell at this point what the hardware problem is.
Prime95 is offline   Reply With Quote
Old 2017-03-19, 02:45   #6
Mark Rose
 
Mark Rose's Avatar
 
"/X\(‘-‘)/X\"
Jan 2013

23×359 Posts
Default

I recently had a machine produce bad results and I fixed it by reseating memory.
Mark Rose is offline   Reply With Quote
Old 2017-03-19, 06:55   #7
stebbo
 
"Chris"
Jul 2016
Melbourne

29 Posts
Default

Thanks for the comments, I will do some tests. This is a brand new machine (about 6 months ago) and one of it's first results ended up being bad.

I will see if I can make it fail and will talk to the supplier.

Cheers,
Chris.

P.S. Just out of curiosity, what's the deal with obscuring the residues? I have no idea what the issues are there?
stebbo is offline   Reply With Quote
Old 2017-03-19, 09:18   #8
henryzz
Just call me Henry
 
henryzz's Avatar
 
"David"
Sep 2007
Cambridge (GMT/BST)

22·1,433 Posts
Default

Quote:
Originally Posted by stebbo View Post

P.S. Just out of curiosity, what's the deal with obscuring the residues? I have no idea what the issues are there?
If we don't then it is possible for someone else to submit a matching result without any work giving the illusion that the number has been double checked.
henryzz is offline   Reply With Quote
Old 2017-03-19, 14:10   #9
ATH
Einyen
 
ATH's Avatar
 
Dec 2003
Denmark

3·23·43 Posts
Default

Is this a Skylake processor?

If yes then make sure your BIOS is upgraded to the latest version and maybe even try this microcode "fix". See post #62 and #67 here:
http://mersenneforum.org/showthread.php?t=21485
ATH is online now   Reply With Quote
Old 2017-03-20, 23:55   #10
stebbo
 
"Chris"
Jul 2016
Melbourne

2910 Posts
Default

Hi ATH,

yes, it's a skylake processor. I will try those fixes you suggested.

Out of interest, I have just run 20 hours of torture tests (4 threads) with 0 errors and 0 warnings.

Cheers,
Chris.
stebbo is offline   Reply With Quote
Old 2017-03-21, 23:11   #11
henryzz
Just call me Henry
 
henryzz's Avatar
 
"David"
Sep 2007
Cambridge (GMT/BST)

131448 Posts
Default

Quote:
Originally Posted by stebbo View Post
Hi ATH,

yes, it's a skylake processor. I will try those fixes you suggested.

Out of interest, I have just run 20 hours of torture tests (4 threads) with 0 errors and 0 warnings.

Cheers,
Chris.
My experience with it was that it can go a very long time without an error then you get a few in an hour or two and then none for ages.
The time between errors is random.
The latest BIOS helped me but didn't fix. I think turning off hyperthreading might have done but it has had months between errors before.
henryzz is offline   Reply With Quote
Reply

Thread Tools


Similar Threads
Thread Thread Starter Forum Replies Last Post
Suspect results koekie Data 7 2018-02-17 00:06
Suspect results bgbeuning PrimeNet 7 2017-07-20 16:18
What is a "suspect" result sixblueboxes PrimeNet 82 2015-09-20 01:07
Two very suspect results tha Data 6 2015-05-22 16:46
suspect LL assigned again to me rudi_m PrimeNet 10 2009-02-12 09:56

All times are UTC. The time now is 07:09.

Thu Oct 22 07:09:21 UTC 2020 up 42 days, 4:20, 0 users, load averages: 1.42, 1.40, 1.30

Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2020, 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.