Overclock.net banner

1 - 2 of 2 Posts

21 Posts
Discussion Starter #1

I decided to use prime95 to find out what my CPU load temps would look like
(Athlon XP Barton core 3200+ not_Oced at the moment, Gigabyte G-power cooler pro).

So, i used the torture test-in_place large FFTs option which states that it causes the max heat, power cons., e.t.c.

My cooler was set at the maximum speed and i started the test.
My idle temps were 40oC and went up to 42oC, after 15 minutes no further increase.

I wasn't sure if Prime95 was really torturing my CPU with CPU temperature only at a +2oC, so i turned the cooler at the lowest speed setting.

CPU temperature increased at 46oC and then momentarily it would jump to 51 or 50, but then immediately back to 46oC (used MBM5 and Speedfan to measure the temps). This was not happening constantly but happened 2-3 times. (might be a sensor inaccuracy, or maybe a momentarily peak)

Anyway, i kept my cooler at lowest fan speed for 5 minutes and then raised it back to highest fan speed. The temps stabilized at 45oC, after 5 minutes and then i turned prime 95 off.

After turning prime95 torture test off, the CPU temps remained at 45oC (normally are at 38-40oC idle) and wouldn't change even if i turned my cooler to highest fan speed or lowest... what do you think might be the cause of this strange behaviour.

note that i always have speedfan on and the temps readings are not flactuating or have any strange behaviour.

Finaly, how many minutes should i leave prime95 torture test on to accurately test my CPU's load temps ? (note that i am not talking about stability, just to get the load temps).

Thanks in advance for any advice you might give me...

Premium Member
4,847 Posts
Prime95 is pretty reliable for testing, it nearly killed my core :-S

If your not convinced by it then you could SuperPi it for like a jillion years, it wont be a TRUE test but itll push it pretty far.

OR there is the funner option of just playing some games at high and use speed fan to log the temps
1 - 2 of 2 Posts