Question Regarding Prime95, AVX, and CPU damage

  • Thread starter Thread starter Deleted member 2969713
  • Start date Start date
D

Deleted member 2969713

Guest
I just wanted to get some viewpoints on this subject since I haven't found a recent thread specifically about this topic and I've seen conflicting views on the matter.

Some say that running Prime95 risks damaging your CPU because it's an "unrealistic" work load utilizing AVX at a capacity that processors aren't designed for. Others say that if running Prime95 damages your CPU, then it's a problem with the hardware, not the software. Others still say that it's fine as long as you don't run it for too long, or it's fine if you aren't overclocking or messing with voltages.

So, can using Prime95 for stress testing damage your CPU, or not? If so, under what circumstances? Is it strictly the temperature of the CPU that is a concern, or is there more to it than that?

I ran Prime95 for a total of about eight and a half hours, doing small FFTs, large FFTs, and blend tests. During that time, the CPU never exceeded 75 degrees Celsuis according to CoreTemp. In a case like that, is there any possibility of my CPU having been caused damaged or having a noticeably reduced lifespan?
 
Thanks for your replies. Here is an example post from 2019 from a user (a moderator, in fact) making a statement about Prime95 similar to others I've seen:

When any version of Prime95 with AVX code runs on AVX capable processors, it can impose an unrealistic 130% workload, which can adversely affect stability and severely overload your processor.

So is this not true, then?
 
I just wanted to get some viewpoints on this subject since I haven't found a recent thread specifically about this topic and I've seen conflicting views on the matter.

Some say that running Prime95 risks damaging your CPU because it's an "unrealistic" work load utilizing AVX at a capacity that processors aren't designed for. Others say that if running Prime95 damages your CPU, then it's a problem with the hardware, not the software. Others still say that it's fine as long as you don't run it for too long, or it's fine if you aren't overclocking or messing with voltages.

So, can using Prime95 for stress testing damage your CPU, or not? If so, under what circumstances? Is it strictly the temperature of the CPU that is a concern, or is there more to it than that?

I ran Prime95 for a total of about eight and a half hours, doing small FFTs, large FFTs, and blend tests. During that time, the CPU never exceeded 75 degrees Celsuis according to CoreTemp. In a case like that, is there any possibility of my CPU having been caused damaged or having a noticeably reduced lifespan?
Running 95 24/7 will eventually kill the the processor and or motherboard. As long as you don't go crazy with CPU speed and/or GPU tweaking you should be okay..
 
Thanks for your replies. Here is an example post from 2019 from a user (a moderator, in fact) making a statement about Prime95 similar to others I've seen:



So is this not true, then?
None of the words he used mean damage.
Your CPU could freeze/hang which is the "adversely affect stability" part and the other thing just means very high temps that you won't see doing anything else.