i5-2500k core temp variance

S

speedlever

Limp Gawd
Joined
Sep 29, 2006
Messages
447
Is it common to see core temps differ by 10*C or more under load? My Core 0 (coolest) seems to run 10 to 11*C cooler than my Core 2 (hottest) under load. I'm wondering if I did a bad job on the thermal compound application or if that's just the normal scope of things.
 
Interesting. I put a tad more TP on the CPU than I intended, so I thought maybe I overdid it which might be causing the large variance.
 
My core 2 does the opposite actually. The gap is widest (~10C) at idle, and narrows under load to 3-5C. After 6 months of use, it seems harmless though. I think the worst thing it does is throw off the "overall" temperature that programs like CoreTemp report.
 
I find Real Temp 3.67 and Core Temp 1.0 to be very close, but HW Monitor temps seem way lower than RT and CT.

Here's a sample of them all after running IBT on very high after enabling the default OC on the P8Z68V Pro mobo.

Apologies for the large image. The 800x600 version was hard to read.

nDswC.jpg
 
I have noticed similar on a 2500k and a 2600k. One will spike a bit higher than any of the others.
 
30-40-41-46 here


So yeah :) Wish they would all run at 30c! I blame the actual device reading the temps. In reality they are probably much closer together (and 40C'ish)
 
speedlever said:
I find Real Temp 3.67 and Core Temp 1.0 to be very close, but HW Monitor temps seem way lower than RT and CT.
Click to expand...

Check if the HWMonitor.ini file is using a tjmax of 98, if not change it. Seems like it's using ~83.

Bad TIM application can cause variance but it's not uncommon to see variance with good TIM application on SNB which can leave you wondering if it's the TIM or not.
 
Actually, the value was 85. I changed the ini file to 98, but haven't noticed a difference. shrug.

Edit: the file is actually hwmonitorw.ini.
 
speedlever said:
Actually, the value was 85. I changed the ini file to 98, but haven't noticed a difference. shrug.

Edit: the file is actually hwmonitorw.ini.
Click to expand...

It's been a long time since I used that program so thanks for the correction. You need to change the ini file and save it before you run the program, at least that's how it use to work. Using 85 instead of 98 will make you temps read 13C lower than they should read.
 
Argh. Thought I saved it and then restarted the program. Tried it again. No joy. Finally, the 3rd time was a charm and 98 stuck. Yes! That does make a difference. Thanks for the tip.

LqpAq.png
 
These sensors are only accurate to +/- 5C so a 10C spread in reported core temperatures when fully loaded running Prime95 Small FFTs is normal. Many other testing applications do not load the cores equally like P95 Small FFTs does so the differences could be greater.
 
Yeah, my new 2600k does pretty much the same at the OP.

RealTempi7.jpg


I'm running a Corsair H70 in P/P with 1800 RPM Gentle Typhoons.
 
I just built an i5-2400 with the CM Hyper 212+/dual fan on a P8Z68V mobo. The core temps on it are remarkably similar without anywhere near the spread I see on my i5-2500k. I guess it's the luck of the draw... or maybe it might have to do with me using a bit less TIM on this build. I'm tempted to take mine apart and clean off the TIM and rework it. Or not. ;)
 
Hi guys I am new to this forum so hi to everyone first of all....:)

My question is about hyper 212 +. I was waiting for a better performance but I am searching the forums for days and realized that my temp readings seems normal. Not perfect just normal. Anyway I want to share my temps(C°) by hardware info 64:

İdle:
core0:36
core1:40
core2:36
core3:37
load, prime 95 30 minutes:
core0:51
core1:60
core2:57
core3:57

My system is:
THERMALTAKE Armor Case
INTEL corei5-2400
ASUS P8P67 DELUXE
4GB Hi Level ram(1600mhz)
ASUS Geforce 680 GTX
OCZ SSD Agility 3 240GB
SAMSUNG 500 gb harddrive
CREATIVE XiFi sound card
LG DVD writer
Cooler Master Hyper 212x heatsink

My room temperature is 25 C degrees.

What do you think?
 
You must log in or register to reply here.
Back
Top