Should I reapply my thermal paste?

witches_jelly

Distinguished
Apr 13, 2010
88
0
18,630
So about a week ago I built a new computer, so far so good aside from the graphics card dying after only 2 days, RMA'd it and got a new one no biggie. Any who I recently installed hardware monitor to see what kind of temps I'm getting from my i5-2500k with my Noctua NH-D14. At the moment I'm getting temperatures across the cores ranging from 111°F(max)[107°F atm] on core 1 to 104°F(max)[93°F atm] on core 4, cores 2 an 3 are a around the middle temp wise. I'm wondering if this difference in temperature across all 4 cores is something to worry about and if it can effect the processors longevity or performance and if I should reapply thermal paste to try and get more even temps. When I applied the thermal paste I used the method recommend in the NH-D14's instructions, putting a 4-5mm dot of it on the center of the CPU. Would it be better to use the vertical line method if I need to reapply?
 

arson94

Distinguished
Apr 18, 2008
867
0
19,010
The vertical line method is for heatpipe coolers where the heatpipes directly touch the CPU's heatspreader like the OCZ Vendetta II. The ridges prevent evenly distributing thermal paste from one dot in the center via pressure alone. Your NH-D14 has a heatspreader covering the heatpipes giving the contact surface a flat surface just like the CPU's heatspreader. So, the dot method is the method you should have used just like you did.

Your temps are fine. Temps differing between cores is also fine. They will almost always differ unless all 4 cores are experiencing the same amount of stress for the same amount of time. Core 1 will probably remain the hottest as it's always in use no matter what. When the workload is high enough or multi-threaded it'll be distributed amongst the other cores. But, until then, other cores remain idle and therefore cooler.

Now, your temps are definitely within Intel's max operating temp, but I'm not sure what the norm is idle for your CPU with your heatsink. Somebody else will need to answer that for you. But I can tell you that you applied the thermal paste using the right method and cores with different temps are nothing of concern.


EDIT: slcmike posted while I was typing too much lol
 

beenthere

Distinguished
Probably because most people use degrees Celcius when reporting temps, the OP's temps might look high but are just fine because they are in Farenheit not Celcius. His temp. is like 44C.

Variations in core temps is normal and a few degrees C difference is of no concern.
 

witches_jelly

Distinguished
Apr 13, 2010
88
0
18,630
Thanks, I wasn't aware the 1st core was always in use, I'd assumed that usage was spread across all cores at all times. Those temp numbers by the way are from just casual internet use, so Idle I guess. I've haven't checked them under load and am not really sure the best way to do that. as for it in Celsius the temps range from 32 to 41 C.
 

arson94

Distinguished
Apr 18, 2008
867
0
19,010


Idle temps would be around 30C, give or take a couple of degrees and depending on the temperature of your room. So you seem to be fairing well in that range. Internet use though would be a little more than idle and dependant on the context of the site you're currently viewing. Your motherboard probably came with monitoring/overclocking software that can also monitor your CPU usage and stats like current speed (in Mhz) and temperature. With Intel's Speedstep enabled in BIOS, which it usually is enabled by default, you'll see that your CPU's speed decreases in times of low usage. This decrease in speed would happen in times of idleness. Your temps of course would drop to their lowest during these times. To check your temps under load would be accomplished by stress testing your computer. To do this, the defacto program of choice is Prime95. http://www.mersenne.org/freesoft/ Your temps will soar to their highest after running the program for a bit. Prime95 is popular to stress test overclocking configurations for stability as well as maximum temperatures reached under whatever configuration.

That is all for future reference though and in case you're ever interested in it. For the purpose of this thread though, you're fine and seem to have no reason for further concern. You should put that new video card of your's and get to gaming!

 
@witches_jelly,

It's always a good practice to pull the heat sink and inspect the footprint, so you know exactly what's happening when it's reseated, the D14 is very easy to remove and reseat, I would want to know I had a good thermal spread, just enough and not too much of the NT-H1 compound.