more about hardware errors

graeme · 2003-07-04, 09:19

Hi everyone -
I'm 30% of the way through checking an exponent and I've started getting possible hardware failures (SUMINPUTS!=SUM[OUTPUTS] ), about 1 per day.
So there's probably a hardware error, but I can't find it and it only shows up in prime95.

It seems I shouldn't use my poor computer for exponent checking, but should I let it continue with the current one, or is it not worth it.

Presumably I can still look for factors?

What do people advise,

Graeme

lycorn · 2003-07-04, 13:58

The type of error you are reporting is normally a serious one, in the sense that the final result has a fairly high probability of being wrong.
It will most likely be caused by faulty memory and/or excessive heating/overclocking of the system.
If you´re running the system at stock speed, my advice is not to use it in LL testing, at least until you find and correct the root problem. Trial factor is, AFAIK, less stressful to the machine, so you may want to switch to that type of assignment.
If you browse the hardware forum, you will find some threads addressing the subject of errors that may as well be helpful.

outlnder · 2003-07-04, 17:09

Update all your drivers. Motherboard, especially chipset, video and sound are the most common culprits.

2003-07-04, 19:03

Are you using a PC100?
That is usually the problem, where the memory is thrashed with large input and output files. Overclocking is rarely the problem in my experience.

graeme · 2003-07-08, 09:14

hi All,

three more errors over the weekend. This run seems to be wasted time anyway, so I unreserved the exponent, selected mersene numbers to factor and off we go, finding factors for M23066921.
Ah well,
GRaeme

2003-07-04, 09:19	#1
graeme Jul 2003 41 Posts	more about hardware errors Hi everyone - I'm 30% of the way through checking an exponent and I've started getting possible hardware failures (SUMINPUTS!=SUM[OUTPUTS] ), about 1 per day. So there's probably a hardware error, but I can't find it and it only shows up in prime95. It seems I shouldn't use my poor computer for exponent checking, but should I let it continue with the current one, or is it not worth it. Presumably I can still look for factors? What do people advise, Graeme

2003-07-08, 09:14	#5
graeme Jul 2003 29₁₆ Posts	bit the bullet hi All, three more errors over the weekend. This run seems to be wasted time anyway, so I unreserved the exponent, selected mersene numbers to factor and off we go, finding factors for M23066921. Ah well, GRaeme

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
Reproducable hardware errors?	cmokruhl	Software	2	2002-09-17 19:04

2003-07-04, 13:58	#2
lycorn "GIMFS" Sep 2002 Oeiras, Portugal 3×23² Posts	The type of error you are reporting is normally a serious one, in the sense that the final result has a fairly high probability of being wrong. It will most likely be caused by faulty memory and/or excessive heating/overclocking of the system. If you´re running the system at stock speed, my advice is not to use it in LL testing, at least until you find and correct the root problem. Trial factor is, AFAIK, less stressful to the machine, so you may want to switch to that type of assignment. If you browse the hardware forum, you will find some threads addressing the subject of errors that may as well be helpful.

2003-07-04, 17:09	#3
outlnder Aug 2002 2·3·53 Posts	Update all your drivers. Motherboard, especially chipset, video and sound are the most common culprits.

2003-07-04, 19:03	#4
TTn 3751₈ Posts	Are you using a PC100? That is usually the problem, where the memory is thrashed with large input and output files. Overclocking is rarely the problem in my experience.