mersenneforum.org  

Go Back   mersenneforum.org > New To GIMPS? Start Here! > Information & Answers

Reply
 
Thread Tools
Old 2010-02-26, 23:12   #1
BlackOmega
 

400310 Posts
Default Prime 95 kills CPUs?

Hey everyone, I've got a question about P95 and CPUs'. I'm an overclocker and I use p95 for stability testing. I really like the program; I feel that it really puts a load on all of the components to truly check for stability at the smallest possible level.

Anyway, on another overclocker forum someone posted that P95 kills cpus; is this true?

Here's the story, a guy overclocked his i7 975 extreme to 4200 MHz. Him and his buddy made a bet that P95 would kill his CPU in 6 days time. So they let it run, 5.5 days later the CPU is dead. Now I know that an overclocked CPU is prone to electromigration (which is what I think killed it) especially @ 45nm.

But has anyone ever had a non-overclocked CPU die on them from running prime?

Also, correct me if I'm wrong, but don't some people run P95 24/7?

Any input would be greatly appreciated.
  Reply With Quote
Old 2010-02-26, 23:51   #2
retina
Undefined
 
retina's Avatar
 
"The unspeakable one"
Jun 2006
My evil lair

22·5·307 Posts
Default

It is my understanding that all electronic components are prone to electromigration regardless of overclocking. However it is a very long term thing, I doubt that 5.5 days would be enough to manifest an electromigration problem. I think you need to look for another answer as to why the system failed.
retina is online now   Reply With Quote
Old 2010-02-27, 00:47   #3
lycorn
 
lycorn's Avatar
 
Sep 2002
Oeiras, Portugal

2·36 Posts
Default

The PC I´m using to write this message (P4 Northwood 3200) has been running Prime95 for over 5 years now, 24/7, and overclocked for most of the time. It is alive and well. I have also recently retired from service an old Athlon XP 1533, that had been running P95 for over 7 years. So definitely no, Prime95 doesn´t kill CPUs.
If somebody runs an OCed CPU at full load 24/7, some precautions have to be taken as far as cooling goes. So I´m with retina in that the reason for the i7 to fry lies somewhere else, it is not a P95 issue.
lycorn is offline   Reply With Quote
Old 2010-02-27, 00:56   #4
BlackOmega
 

31·227 Posts
Thumbs up

Thanks for the replies!

I was sure that P95 is not the cause of the failure. This just re-affirms my suspicions that something else caused the failure, most likely temps.
  Reply With Quote
Old 2010-03-01, 02:14   #5
joblack
 
joblack's Avatar
 
Oct 2008
n00bville

72510 Posts
Default

Quote:
Originally Posted by BlackOmega View Post
Hey everyone, I've got a question about P95 and CPUs'. I'm an overclocker and I use p95 for stability testing. I really like the program; I feel that it really puts a load on all of the components to truly check for stability at the smallest possible level.

Anyway, on another overclocker forum someone posted that P95 kills cpus; is this true?

Here's the story, a guy overclocked his i7 975 extreme to 4200 MHz. Him and his buddy made a bet that P95 would kill his CPU in 6 days time. So they let it run, 5.5 days later the CPU is dead. Now I know that an overclocked CPU is prone to electromigration (which is what I think killed it) especially @ 45nm.

But has anyone ever had a non-overclocked CPU die on them from running prime?

Also, correct me if I'm wrong, but don't some people run P95 24/7?

Any input would be greatly appreciated.
If you have killed your CPU because of an extreme cpu core voltage you might have known it before. I never had such a case. The CPU downclocks from 90 C on so you have to have had an extreme voltage and run the cpu in downclock mode (which is senseless anyway).

Prime95 isn't killing cpus but showing if you haven't overclocked too much.
joblack is offline   Reply With Quote
Old 2010-03-01, 04:31   #6
BlackOmega
 

7,369 Posts
Default

So the truth finally came out. The guy was running his i7 @ 1.77v. Even though his temps were fairly low (60C), exceeding Intels recommended voltage by .37v is just foolish.
  Reply With Quote
Old 2010-03-01, 07:32   #7
cheesehead
 
cheesehead's Avatar
 
"Richard B. Woods"
Aug 2002
Wisconsin USA

22×3×641 Posts
Default

Quote:
Originally Posted by BlackOmega View Post
Anyway, on another overclocker forum someone posted that P95 kills cpus; is this true?
Everyone above has given sound answers. Here's another response you can give anyone making that P95-destroys-CPUs claim:

Nothing that Prime95 does is outside manufacturers' specifications in any way. It contains only legitimate instructions, executed in a legitimate manner, and never violates any manufacturer's warranty in any way.

If some hardware part of a system fails during Prime95 execution, it will be because of some defect in either that or some other part, or in the system environment. Such defects may be a result of inadequate cooling, faulty materials or construction, incorrect or faulty connections, or operation outside the manufacturer's specifications (voltage, temperature, ventilation, etc.). Possible reasons for inadequate cooling include blocked air ducts, dust accumulation, improper cooling design, cooling fan failure, and overclocking beyond manufacturer's specifications. There are similar possible reasons for other types of failure, but Prime95 execution by itself is not one of them.

Prime95 may very well operate your system closer to its limits than almost any other software you use, but any properly designed, manufactured, assembled, adjusted and situated system should be able to run Prime95 without harm within the manufacturer's specified limits. A system that fails during Prime95 execution would probably fail during execution of any other software that places the same legitimate load on the system as Prime95 does.
cheesehead is offline   Reply With Quote
Reply

Thread Tools


Similar Threads
Thread Thread Starter Forum Replies Last Post
Number of CPUs to use houding Software 3 2015-02-26 19:56
cpus running at 10% wildrabbitt Information & Answers 9 2015-01-20 13:31
M929 progress: Sure, drinking kills brain cells - but only the slow ones! fivemack Factoring 125 2013-03-06 20:21
kernel grsecurity kills mprime gue Software 0 2003-12-06 17:22
Windows XP logoff sometimes kills Prime95 markhl Software 8 2003-07-07 16:00

All times are UTC. The time now is 22:19.

Sun May 9 22:19:08 UTC 2021 up 31 days, 17 hrs, 0 users, load averages: 2.85, 2.96, 3.03

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.