Sign in with
Sign up | Sign in
Your question

Prime95 temps spike at 16 minutes

Last response: in Overclocking
Share
May 24, 2012 3:18:26 AM

So after testing out the overclocking potential of my new i5 3570k, I was able to get the chip up to 4.9GHz at 1.368v before temps started becoming a problem, hovering around 95 C on the hottest core using the latest version of Prime95. Temperatures stayed around that range until the test reached 16 minutes, then they spiked up to 105 C.

I backed the chip down to 4.8ghz at 1.296v, where temperatures maxed out at a more reasonable 83 C. However, once the test got to 16 minutes again, temps spiked up by 15 C!

Is this normal behavior for Prime95, and why is it happening at exactly the same point in the test? Do I just need to account for these spikes as my max temperature for a stable overclock?

a c 176 à CPUs
a b K Overclocking
May 24, 2012 5:38:59 AM

BVKnight said:
So after testing out the overclocking potential of my new i5 3570k, I was able to get the chip up to 4.9GHz at 1.368v before temps started becoming a problem, hovering around 95 C on the hottest core using the latest version of Prime95. Temperatures stayed around that range until the test reached 16 minutes, then they spiked up to 105 C.

I backed the chip down to 4.8ghz at 1.296v, where temperatures maxed out at a more reasonable 83 C. However, once the test got to 16 minutes again, temps spiked up by 15 C!

Is this normal behavior for Prime95, and why is it happening at exactly the same point in the test? Do I just need to account for these spikes as my max temperature for a stable overclock?

it may be where in prime95 most of the folding starts. you can try intel burnin and see if the max temp is reach sooner or latter then prime95. most time with burn in diags if there written right you want a slow ramp up to max temp then hold it for a set time to watch out for thermal runaway. at the max point you dont want to be drawing more power then the cpu or mb can handel...then you want to slowly ramp down and then stop the test. i used to work with environment chambers. ramp from -5c to 50c in 3-4 min. we used to split cpu and other parts from thermal shock if they had a defect.
!