Best Thermal Paste?

[nekrosoft13]

for long long time I been using Artic Silver, is there something better out now?

 

[felt]

http://hardforum.com/showthread.php?t=1832045&highlight=

some good suggestions there

 

[fuzchunk]

coollaboratory liquid pro. makes a mess though.

 

[Arct1c0n]

I finally retired my usage of Arctic Silver after everyone gave me shit for buying another tube of it for my new build, so now i'm using this highly rated stuff.

 

[Tsumi]

I use Arctic Ceramique 2. Mostly because I bought a 22g tube 5 years ago and haven't even gotten through half of it yet.

 

[Rav3n]

I use Arctic Ceramique also, and have for 15+ years. Never had a single issue with it. Cheap, easy and rarely dries out.

 

[t1337Dude]

Tom Hardware's thermal paste shoot-out suggests Gelid GC Extreme Thermal Paste is the best and nets a decent performance boost over AS5, particularly when applied to videocards. There are a few better but they can't be applied to everything and require more experience to apply.

 

[Arct1c0n]

Tom Hardware's thermal paste shoot-out suggests Gelid GC Extreme Thermal Paste is the best and nets a decent performance boost over AS5, particularly when applied to videocards. There are a few better but they can't be applied to everything and require more experience to apply.

Since I was redoing my HSF AGAIN last night, I decided to go ahead and pull apart my new STRIX 980. HOly God, they did a craptastic job at the assembly line. Took me 15 mins to clean all that crap up and do it properly with my Gelid DC Extreme.

Sadly, I havent' seen a squat in difference in temps

 

[t1337Dude]

Since I was redoing my HSF AGAIN last night, I decided to go ahead and pull apart my new STRIX 980. HOly God, they did a craptastic job at the assembly line. Took me 15 mins to clean all that crap up and do it properly with my Gelid DC Extreme.

Sadly, I havent' seen a squat in difference in temps

Agreed on the poor assembly line job. I'll admit that your job was vastly better looking than mine. The core on my GPU wasn't even fully covered, there were patches of dry-surface all over it. As for temps, it's really difficult to compare temperatures unless you recorded data from some benchmarks with the GPU fan set to a constant RPM before and after applying the paste. If you're using a fan curve, you likely won't notice a difference in temps, you'd only notice the videocard not getting as loud as it tries to achieve its target temp.

I forgot to take it off a fan-curve so it's difficult for me to compare too. Really, I've only made a couple of observations that stuck out. I notice that my card now idles at ~27c and below as opposed to ~34c where it used to (GPU fan on minimum). That's the only thing that particularly sticks out to me. Also, while benchmarking my card was quick to hit 85c when overclocked, and my GPU fan hovered at the peak of the fan curve. After doing the same amount of benchmarking after applying new paste, I still I haven't seen my card exceed 81c, or hear it get particularly noisy yet.

81c seems pretty hot but it's actually an improvement for my card. Especially if I don't have to hear the fan, which I rarely do anymore, even at my card's max OC. All of these changes could also be from a completely new case layout, as I switched from a case with not a ton of airflow to a case that I have decked out with 5 big Noctua fans.

 

[Arct1c0n]

My idle temp didn't change at all with the same fan curve I had, I was rather dissapointed after all that work

 

[DeadSkull]

This stuff. Been using it for 3 years now after trying Liquid Pro & Ultra and like 15 other TIMs

I finally retired my usage of Arctic Silver after everyone gave me shit for buying another tube of it for my new build, so now i'm using this highly rated stuff.

 

[Reality]

CoolLaboratory Liquid Ultra if you ask me, expensive though

It's conductive though, so be warned when using it on GPU dies

 

[durquavian]

I have been using IC Diamond paste and it has been extremely worth every penny. CPU temps on an 8350 dropped 10C at max load. Even on my GPUs saw a 10C drop. My 290 went from 36 idle to 28C (air conditioned case).

 

[Jorona]

Word of warning with IC Diamond, it does etch. So if you tend to clean and reseat your coolers often, it'll eventually wear grooves into the heat speaders/die/base of cooler over time.

I also recommend GC Extreme. After using AS5 for years til the tube ran out, I did see acouple degree improvement on everything I used it on.

 

[durquavian]

I already resurfaced CPU and Cooler, didn't my GPU cooler (warranty and all) but have never seen any sign of surface scratching. Others with far more experience using it have had no issues either. Seems some time ago there was some fiasco that eventually turned out to be false, but unfortunately still seems to linger.

 

[Dayaks]

I used AS5 too... going to switch to Gelid on my next build. Non-conductive is a big plus.

 

[durquavian]

Now you got me wondering so I looked for charts of thermal pastes. Funny thing is, IC Diamond isn't in any of the recent ones. I mean there are a ton of users that swear by it in many OCing sites but it seems no benching/comparisons done recently use it in their testing suite.

Just looking from the technical side, it is superior to Silver 400mw/k to 2000mw/k. IC Diamond lists a chart of the differences using results from users across a number of forums, but it would be nice to see a test using it like they do with the others (usually 30-40 of them at a time).

 

[FRZ]

I use pk-3. Works quite well for me.

 

[lord_emperor]

I use Arctic Silver 3 lol. I wonder where it belongs on a moden benchmark chart... somewhere between Silver 5 and Alumina? I guess my CPU will survive being 3.2ºC hotter without the newest and best thermal paste.

 

[durquavian]

Actually they should do a test with a resurfaced CPU/heatsink and a stock one, of course the heatsinks are the bigger issue. My 8350 needed it bad, My phenom not so much.

 

[Comixbooks]

LOL in that graphics they use toothpaste funny stuff

If they used Cool Whip that would make my day..

 

[Comixbooks]

Maximum PC did a review of them like two years ago

http://www.maximumpc.com/article/features/geek_tested_17_thermal_pastes_face

 

[Jorona]

I already resurfaced CPU and Cooler, didn't my GPU cooler (warranty and all) but have never seen any sign of surface scratching. Others with far more experience using it have had no issues either. Seems some time ago there was some fiasco that eventually turned out to be false, but unfortunately still seems to linger.

Nah, I've seen scratched dies on acouple laptops. Its really if you remove the stuff without using any cleaner, just trying to wipe it off. Nothing serious, and I've never seen it "ruin" anything, but it does happen.

 

[sun_bean]

Since I was redoing my HSF AGAIN last night, I decided to go ahead and pull apart my new STRIX 980. HOly God, they did a craptastic job at the assembly line. Took me 15 mins to clean all that crap up and do it properly with my Gelid DC Extreme.

Sadly, I havent' seen a squat in difference in temps

I always clean out manufacture crappy thermal paste and use Artic Silver 5

 

[turbolude]

lets use our brains here. 1-3deg difference doesnt mean anything on gains. Unless we are talking phase change or something extreme. I could be wrong with combined temp drops from cpu and gpu temps. it would only make a diffence if ur running air and 4 graphics cards.

 

[Bone_Enterprise]

I hear a slice of American Cheese works great.

 

[durquavian]

The only other issue I have is testing after a longer cure time. My IC DIamond drops temps by 2-4C the day I apply. The next day they are 6-10C lower (usually greater at Idle).

 

[TheGamerZ]

I highly recommend NOT using LiquidPro or Liquid Ultra, as it will bond and make a serious mess of copper headsinks, requiring sanding/grinding to get it off.

The only real reason to use it (IMO, of course) is for when you're delidding your CPU and use that to replace the stock crap (Looking at you Ivy Bridge/haswell!)