3770K temps with Prime95

bgt

Senior member
Oct 6, 2007
573
3
81
Why is there such a high temp. when running prime95? The CPU value stays low but core value is high.
3770kPrime95a.jpg
 

fixbsod

Senior member
Jan 25, 2012
415
0
0
Core looks wrong. Low 60s are good temps assuming that P95 has been running for at least 30+ mins. However that program may be wrong as it doesn't appear to be reading the temps right as you pointed out. Try HWinfo64 or HWmonitor
 

Vectronic

Senior member
Jan 9, 2013
489
0
0
As suggested, HWMonitor, HWiNFO, RealTemp, AIDA64, CoreTemp

Personally I prefer RealTemp for just monitoring CPU... fast, accurate, options, stays out of the way. This version of it has a few more options.

64C seems a bit low if you are air-cooled (or running Blend, and not far into it).
 

bgt

Senior member
Oct 6, 2007
573
3
81
Switched turbo off but temps are still high. Temp difference between CPU and core is crazy. HWinfo64 does not work. It freezes the pc.
3770kPrime9a.jpg
 
Last edited:

Idontcare

Elite Member
Oct 10, 1999
21,110
64
91
Speedfan was outdated 3yrs ago, to be trying to use it now with a modern processor is just D:

Use RealTemp or CoreTemp so you at least know the correct temperature per your processor's DTS. Then benchmark that against other programs if you prefer to use a different program, but make sure the other programs are correct.
 

bgt

Senior member
Oct 6, 2007
573
3
81
With realtemp and coretemp the same high temps. It uses 100W less than the the 8350 but the temps are worse. Power use with prime95= 112W w turbo off.

3770kPrimea.jpg

3770kPrime5a.jpg
 
Last edited:

Vectronic

Senior member
Jan 9, 2013
489
0
0
Just for curiosity, double check the TJMax in RealTemp under Settings (should be 4 grayed out boxes at the top all saying "105")

If those are 105, then
(nevermind, Math says 105, lol)... your temps seem to be "accurate".... I don't know if you meant "same high" as in equal, or "same high" as in HOLY S**T TEMPS!!!... but, mid 60's is easily safe (though you haven't run P95 for very long, try the Small FFT setting, or LinX).

So yeah, use RealTemp, or use SpeedFan and ignore it's "CPU: X" reading.
 
Last edited:

Vectronic

Senior member
Jan 9, 2013
489
0
0
To get the temps lower than the 60's you will need to delid.
Not necessarily... he's at 3.5GHz with 1.112. Even with Turbo enabled, he might be able to drop his voltage, and if he's planning on sticking with Turbo disabled, he might be able to drop it as low as 0.9 or 0.95 volts... which would reduce the temps a lot.

Regarding temps, I've had mine (3570K, but basically the same) up to 105 3 or 5 times now, and have done LinX test @ 4.5GHz (what' i'm currently "testing") in the 102/103 peak temp ranges for significant durations... *so far* no ill effects...lol

With mine, my day to day 100% usage, the peak temp will be about 30C lower than whatever LinX/P95 maxes out at... if I'm doing 3D rendering (4 threads, 6GB+ memory usage), 4.5GHz is only about 68C.
 
Last edited:

Rvenger

Elite Member <br> Super Moderator <br> Video Cards
Apr 6, 2004
6,283
5
81
Those temps are normal. Mine hits 78c with IBT @ 4.2ghz.
 

ShintaiDK

Lifer
Apr 22, 2012
20,378
146
106
With realtemp and coretemp the same high temps. It uses 100W less than the the 8350 but the temps are worse. Power use with prime95= 112W w turbo off.

3 small things to include as well:

1. AMD uses a different method and location to measure the temperature. Hence you cant compare the 2.

2. Tjmax is 105C, meaning under 105C is fine. 60C is very nice.

3. Temperature and wattage cant be mixed like you do. A 1W needle size could be 100C, while a 2000W radiator might just be 50-60C. You would also notice that your IB cooler is alot colder than an FX cooler would be.
 

bgt

Senior member
Oct 6, 2007
573
3
81
My FX cooler feels cooler than the IB cooler.......no issues there. The IB has a very dynamic thermal response....hence the small die probably. It seems its no issue to have the IB reaching much higher temps then the FX. Ah well, no problem then:whiste:On with testing this little devil.