Are these temps normal for Thermalright SA?

[tizm]

I've been stress testing my i7 2600k CPU (with HT on) using prime95 and was just wondering if these were ok temps:

Haf X with all fans running max speed
Room temp ~70-75 F
Prime95 running 12+ hours ~ highest so far on just core#1 is 67 while the rest varies from 55-60
Edit: Overclocked to 4.5Ghz and 1.305 fixed vcore in BIOS but voltage goes up to 1.33 on full load

If anyone else has an i5 k series or i7 with noctua/thermalright SA, I'd like to compare to see if it's worth reapplying thermal paste / reseating my SA.

 

[stahlhart]

What are your idle temps?

 

[BonzaiDuck]

I've been stress testing my i7 2600k CPU (with HT on) using prime95 and was just wondering if these were ok temps:

Haf X with all fans running max speed
Room temp ~70-75 F
Prime95 running 12+ hours ~ highest so far on just core#1 is 67 while the rest varies from 55-60
Edit: Overclocked to 4.5Ghz and 1.305 fixed vcore in BIOS but voltage goes up to 1.33 on full load

If anyone else has an i5 k series or i7 with noctua/thermalright SA, I'd like to compare to see if it's worth reapplying thermal paste / reseating my SA.

That doesn't seem too far out of whack. It actually seems about right with that VCORE and speed. Which Prime test are you running? Is HT enabled?

We've explored related factors on another thread: what's reported for voltage and temperature can well be BIOS-dependent.

I have a Noctua NH-D14. i7-2600K @ 4.60Ghz -- about the same voltage at full load under PRIME95 Large FFTs. And at that room ambient the average of the core temperatures seems to be in the neighborhood of 60C.

I've had several ThermalRight coolers. They tend to ship with a slight convexity on the heatsink base. My two Noctuas -- a U12P(?) and an NH-D14 -- also had this convexity. I've posted much here about the cumulative improvements from "three different grains of rice:" Flattening the surfaces; removing the nickel plate; and the choice of a TIM. If you simply installed the TR cooler "as is," I could see that our results are roughly equivalent.

What thermal paste/interface are you using?

PRIME95, even the toaster-oven Large FFT test, gives "pussy" results in temperatures with the i7-2600K, even with hyperthreading enabled. You'll get a different set of numbers using LinX or IntelBurnTest, with libraries updated to load up the processor's iGPU. If you have the iGPU disabled, well -- like I said -- the temperatures are fairly lame . . .

 

[tizm]

What are your idle temps?

Around 25-35. It's weird because one of the cores is almost 10 degrees hotter than the 2 coolest ones - in idle and on load. Based on some of the screenies I've seen, all 4 cores don't fluctuate THAT much...

That doesn't seem too far out of whack. It actually seems about right with that VCORE and speed. Which Prime test are you running? Is HT enabled?

We've explored related factors on another thread: what's reported for voltage and temperature can well be BIOS-dependent.

I have a Noctua NH-D14. i7-2600K @ 4.60Ghz -- about the same voltage at full load under PRIME95 Large FFTs. And at that room ambient the average of the core temperatures seems to be in the neighborhood of 60C.

I've had several ThermalRight coolers. They tend to ship with a slight convexity on the heatsink base. My two Noctuas -- a U12P(?) and an NH-D14 -- also had this convexity. I've posted much here about the cumulative improvements from "three different grains of rice:" Flattening the surfaces; removing the nickel plate; and the choice of a TIM. If you simply installed the TR cooler "as is," I could see that our results are roughly equivalent.

What thermal paste/interface are you using?

PRIME95, even the toaster-oven Large FFT test, gives "pussy" results in temperatures with the i7-2600K, even with hyperthreading enabled. You'll get a different set of numbers using LinX or IntelBurnTest, with libraries updated to load up the processor's iGPU. If you have the iGPU disabled, well -- like I said -- the temperatures are fairly lame . . .

Yup HT is enabled and I was doing a 24 hour blend test. It was weird because when I came home, the temps were like 55, 64, 61, 53. But sometime during the day when I was at work, one of the cores actually hit 68. Again though, I'm a bit worried about the fluctuation.

I'm using MX-4 thermal paste and tried doing the pea sized application. I think I might have put a bit too much (compared it to what other people considered "pea-sized").

And with LinX / IBT, I hear it really toasts your CPU so I'm not sure if I'd be comfortable running those. Even with a Prime95, I hear in reality your CPU will most likely never even reach those temps.

 

[stahlhart]

Around 25-35. It's weird because one of the cores is almost 10 degrees hotter than the 2 coolest ones - in idle and on load. Based on some of the screenies I've seen, all 4 cores don't fluctuate THAT much...

My idle temps are in the low 30s with core at 1.2V, in a consistent b-b-c-a state (a=coolest, c=warmest), about 7C range low to high. If you're at your temps at this core voltage then you've clearly done a better installation job. :thumbsup: I've tried it I think about three times, and end up with the same temps pretty much. Occasionally the lower ones will drop into the high upper 20s, but that's about as low as I can get at this core voltage.

Intel Burn Test at maximum gets temp averages into the mid 60s, again at stock multiplier and 1.2V core.

I have not messed with overclocking much yet; waiting until I get some PC1600 RAM for that.