Q6600 G0 Temps - Too big of a spread?

[lilbabimac]

I'm wondering if the temps on my Q6600 is normal. The spread from core 0 and 3 seems to be pretty big. This is with a TRUE-120:

Ambient temp: 25 C

Idle Temps:

Core 0: 43 C
Core 1: 40 C
Core 2: 36 C
Core 3: 35 C

Load Temps:

Core 0: 55 C
Core 1: 51 C
Core 2: 47 C
Core 3: 46C

These are what the temps have been for about a year, so maybe I should try reseating the heatsink?

BTW, these temps are at stock with no OC.

 

[error8]

First of all use RealTemp and set Tjunction at 90C manually for all cores. If those you've listed are red ain those conditions, then they are normal. Do you have the impression that 55 C is too hot? Mine goes up to 79 C and I don't have a problem with that.

 

[lilbabimac]

I'm using Core Temp to get these temps and loading with prime95. I have no problem with 55 C, but I am wondering if the difference between Core 0 and Core 3 is normal or too much.

 

[error8]

I have a difference of 5 C between core 0,1 and cores 2,3. It's a normal situation, don't worry about it.

 

[Idontcare]

The cause of the temperature delta is either above the IHS (i.e. bad TIM job, reset HSF, warped IHS, warped HSF) or under the IHS (bad solder job, warped IHS, electrically different IC's with significant power consumption deltas, etc).

If the source is above (outside) the IHS then reseating your IHS and redoing TIM is the first step...if care to try and eliminate it. As error8 said, 5C ain't no thing. I ran with 8C delta without issue, in my case the source was under the IHS so nothing I could do about it.

 

[BonzaiDuck]

Originally posted by: Idontcare
The cause of the temperature delta is either above the IHS (i.e. bad TIM job, reset HSF, warped IHS, warped HSF) or under the IHS (bad solder job, warped IHS, electrically different IC's with significant power consumption deltas, etc).

If the source is above (outside) the IHS then reseating your IHS and redoing TIM is the first step...if care to try and eliminate it. As error8 said, 5C ain't no thing. I ran with 8C delta without issue, in my case the source was under the IHS so nothing I could do about it.

. . . And that's why lapping the IHS may trim the delta, but it won't eliminate it. At LEAST, you should be able to trust the core sensors on Conroe/Kenstfield CPUs . . .

 

[lilbabimac]

Yeah, I figured it might be uneven surface on heatsink or CPU, but I don't really want to lap either one of them. I guess I'll try reapplying TIM and reseating the heatsink and see if that does anything.

 

[California Roll]

My core 0 and 1 are the same. Core 2 is 6C lower. Core 3 is 4C lower. Your temps are fine. This happens a lot with quads.

 

[Schmide]

Almost identical to my q6600, except mine is a few degrees cooler and some of the cores are mixed up (down stairs)

 

[lilbabimac]

Cool, thanks for the response guys - now I know it isn't just me. It would be nice for all the cores to idle around low 30's though.

 

[Qbah]

Those are very good temps. Don't worry! My Q9450 idles at 40/40/41/46 and loads at 48/48/49/55. If I open my balcony the hottest core goes to 40 at idle And my GPU drops to 36C (from 40) So you're really fine

 

[hclarkjr]

Originally posted by: lilbabimac
I'm wondering if the temps on my Q6600 is normal. The spread from core 0 and 3 seems to be pretty big. This is with a TRUE-120:

Ambient temp: 25 C

Idle Temps:

Core 0: 43 C
Core 1: 40 C
Core 2: 36 C
Core 3: 35 C

Load Temps:

Core 0: 55 C
Core 1: 51 C
Core 2: 47 C
Core 3: 46C

These are what the temps have been for about a year, so maybe I should try reseating the heatsink?

BTW, these temps are at stock with no OC.

you paid all that money for a true-120 and here i am with a arctic cooler freezer 7 pro and getting these temps -

CPU #1 / Core #1 30 °C (86 °F)
CPU #1 / Core #2 32 °C (90 °F)
CPU #1 / Core #3 25 °C (77 °F)
CPU #1 / Core #4 27 °C (81 °F)
room temp of 68F

 

[Schmide]

Originally posted by: hclarkjr

Originally posted by: lilbabimac
I'm wondering if the temps on my Q6600 is normal. The spread from core 0 and 3 seems to be pretty big. This is with a TRUE-120:

Ambient temp: 25 C

Idle Temps:

Core 0: 43 C
Core 1: 40 C
Core 2: 36 C
Core 3: 35 C

Load Temps:

Core 0: 55 C
Core 1: 51 C
Core 2: 47 C
Core 3: 46C

These are what the temps have been for about a year, so maybe I should try reseating the heatsink?

BTW, these temps are at stock with no OC.

you paid all that money for a true-120 and here i am with a arctic cooler freezer 7 pro and getting these temps -

CPU #1 / Core #1 30 °C (86 °F)
CPU #1 / Core #2 32 °C (90 °F)
CPU #1 / Core #3 25 °C (77 °F)
CPU #1 / Core #4 27 °C (81 °F)
room temp of 68F

Idel temps mean nothing. What are you load temps occt?

Chips differ idle temps more than cooling BTW.

 

[hclarkjr]

load temps are around 122F

 

[Schmide]

Originally posted by: hclarkjr
load temps are around 122F

In occt?

adding

The arctic cooler freezer 7 pro is a great cooler; however, your spread in readings is much greater 28-50c vs 38-50c thus the TRUE-120 is doing a better job cooling the chip.

 

[hclarkjr]

sorry, no that was with prime95 after running it for 3 hours the temps stayed right there the whole time

 

[lilbabimac]

Bringing back an old post, but after debating for a while, I decided to lap my Q6600 and TRUE-120 today in hopes of dropping and evening out my temps:

Lapped Temps:

Idle Temps:

Core 0: 39 C
Core 1: 37 C
Core 2: 33 C
Core 3: 31 C

Load Temps:

Core 0: 49 C
Core 1: 47 C
Core 2: 43 C
Core 3: 42C

VS. Unlapped Temps:

Idle Temps:

Core 0: 43 C
Core 1: 40 C
Core 2: 36 C
Core 3: 35 C

Load Temps:

Core 0: 55 C
Core 1: 51 C
Core 2: 47 C
Core 3: 46C

So there is an obvious drop in temperature on all cores, but I was really hoping to even out the core temps to within a 1-3C difference. I really don't understand why even after lapping it there is uneven temps between core 0 and core 3, but I guess there's really nothing else I can do now. Maybe I'm putting on too much TIM or not seating it right? I know the TRUE-120 is on as tight as it can be, so looseness isn't the problem. Any ideas?

Edit: Going back to the old posts, I guess the problem can be under the chip like Idontcare was saying and that's what's causing the spread in temps between cores.