mersenneforum.org  

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

Reply
 
Thread Tools
Old 2017-02-24, 12:54   #1
bgbeuning
 
Dec 2014

3·5·17 Posts
Default Suspect results

One machine of mine has been getting "suspect" results recently.
The prime95 output looks like

Quote:
[Feb 24 07:35] Iteration: 67110000 / 77920153 [86.12%], ms/iter: 29.509, ETA: 3d 16:36
[Feb 24 07:35] Possible hardware errors have occurred during the test!
[Feb 24 07:35] 14 ROUNDOFF > 0.4 of which 1 were repeatable (not hardware errors).
[Feb 24 07:35] Confidence in final result is very poor.
[Feb 24 07:40] Iteration: 67120000 / 77920153 [86.13%], ms/iter: 31.756, ETA: 3d 23:16
[Feb 24 07:40] Possible hardware errors have occurred during the test!
[Feb 24 07:40] 14 ROUNDOFF > 0.4 of which 1 were repeatable (not hardware errors).
[Feb 24 07:40] Confidence in final result is very poor.
Since it says "not hardware errors" what kind of problem is it?
This is using version "Windows64, Prime95, v28.5, build 2"
on an "i7-4770K @ 3.5GHz" with no overclocking.

A few months ago I added a GTX 970 to the box and the suspect results may
have started about the same time. (It has GTX 760 & 970 running mfaktc.)
bgbeuning is offline   Reply With Quote
Old 2017-02-24, 16:00   #2
VBCurtis
 
VBCurtis's Avatar
 
"Curtis"
Feb 2005
Riverside, CA

2·2,339 Posts
Default

The message says 1 of the 14 were repeatable; a repeatable error is a software error rather than a hardware error. A "software error" is usually that the FFT is barely too small for the test being run. Often P95 will increase the FFT size when such a repeatable error is found, but in your case 13 of the 14 errors are non-repeatable, meaning they're hardware (e.g. bitflips).
VBCurtis is online now   Reply With Quote
Old 2017-07-19, 13:17   #3
bgbeuning
 
Dec 2014

FF16 Posts
Default

The 750W PSU died in this machine.
Maybe it was contributing to the bad results.
bgbeuning is offline   Reply With Quote
Old 2017-07-19, 16:06   #4
kladner
 
kladner's Avatar
 
"Kieren"
Jul 2011
In My Own Galaxy!

22×2,539 Posts
Default

I had a terrible time a while back when a PSU had gone crazily irregular on voltage. It took me a while to realize what was happening. Meanwhile, I was running SUM(Inputs) and Round off checking, and backtracking on a regular basis.
kladner is offline   Reply With Quote
Old 2017-07-19, 23:06   #5
lycorn
 
lycorn's Avatar
 
Sep 2002
Oeiras, Portugal

23×181 Posts
Default

Quote:
Originally Posted by bgbeuning View Post
The 750W PSU died in this machine.
Maybe it was contributing to the bad results.
As you said in your OP, errors likely started after you plugged in a second GPU, therefore increasing substantially the load on the PSU. I wouldn´t be surprised if that was the cause of the errors .A dodgy PSU is a very likely cause for trouble, and that tends to show up under heavy loads.
lycorn is offline   Reply With Quote
Old 2017-07-20, 00:10   #6
kladner
 
kladner's Avatar
 
"Kieren"
Jul 2011
In My Own Galaxy!

100111101011002 Posts
Default

My situation included an AMD FX-8350, a GTX 570 and a 580, both OC'd. However. that was an OCZ 1 KW supply, so it should have been adequate.
kladner is offline   Reply With Quote
Old 2017-07-20, 10:36   #7
henryzz
Just call me Henry
 
henryzz's Avatar
 
"David"
Sep 2007
Cambridge (GMT/BST)

2·41·71 Posts
Default

Quote:
Originally Posted by kladner View Post
My situation included an AMD FX-8350, a GTX 570 and a 580, both OC'd. However. that was an OCZ 1 KW supply, so it should have been adequate.
A dodge 1KW power supply can be fine at 250W but not at 500W. Often problems can be heat related.
henryzz is offline   Reply With Quote
Old 2017-07-20, 16:18   #8
kladner
 
kladner's Avatar
 
"Kieren"
Jul 2011
In My Own Galaxy!

22·2,539 Posts
Default

Quote:
Originally Posted by henryzz View Post
A dodge 1KW power supply can be fine at 250W but not at 500W. Often problems can be heat related.
True, and it was running over 800W. The problem, as I remember, was wild Vcore swings going from idle to various levels of load. I don't think it was running hot, but it has been a while. What I mainly remember is that switching from that Gold PSU to a Platinum 1.2KW dropped the draw on the line power by a significant amount, and eliminated the instability.
kladner 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 result stebbo PrimeNet 23 2017-06-03 11:14
Suspect Software Incompatibility Dionysus Software 3 2016-02-07 13:49
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 06:11.

Sun Mar 7 06:11:11 UTC 2021 up 94 days, 2:22, 0 users, load averages: 1.81, 1.55, 1.50

Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2021, 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.