Thermal Paste Shootout - Q209 @ [H]

FrgMstr

Just Plain Mean
Staff member
Joined
May 18, 1997
Messages
55,532
Thermal Paste Shootout - Q209 - We take 10 high profile thermal pastes used for mating heatsink surfaces to CPUs and other components and subject those to over 1200 hours of testing to find out which one reigns supreme. Drop or spread? And we tried American cheese too, just in case you find yourself in a pinch.


When determining what thermal paste affords you the best value, there are many factors that can influence your buying decision. How often do you change your thermal paste? Do you require the best performance? Are you air cooling or cooling with extreme methods?
If you change your paste often due to new components or paranoia then the...
 
Damn, I guess I'll have to eat my cheese :(

Nice review though, and nice to know that AS5 is still some of the best
 
THANK U KYLE!!! it's quite rare to see thermal paste shootouts. i wanted somn like this more than 10+ yrs ago... but perhaps it wasn't as common back then =P.

one little recommendation: if the drop vs. spread is such a point of contention why not DO BOTH so u can bet best of both worlds?

spread AROUND the drop-point (edges and corners) but leave the center blank until the drop is applied. apply drop, then squish for full coverage without affecting final temps. that's how i do it >P.

i'm assuming of course, thermal paste is much better than thermal adhesives or tape-formats.
 
Wow, all that testing took almost 2 months. There's some dedication! Good job [H], that was an interesting read. Also very interesting was the thinner stuff worked better with just a pea-sized drop. Glad I was doing it correctly with my AS5.
 
Damn, I guess I'll have to eat my cheese :(

Nice review though, and nice to know that AS5 is still some of the best

Not really as it degrades over time (i.e you are meant to re-apply it every year), a problem all the silver based pastes suffer from. For most people who don't change their cpu often buy a paste that lasts (e.g. artic cooling MX-2).
 
Nice one. Looks like I can keep my policy of using arctic ceramique (dirt cheap, non consuctive) without much worries about performance.

Now it's a pity you didn't inlude most [H]ard of all thermal pastes - Coollaboratory liquid pro - which consists of liquid metal (but i don't know if you can buy it in US)
http://www.tweakers4u.de/artikel/coollaboratory_liquid_pro_wlp,,12,2005,,458.htm

And well Zalman STG-1 also would be nice to be included.
 
Thanks for the review.

Next time could you guys include OCZ Freeze Extreme and Innovation Cooling "7 karat" TIM?

Thanks


/I cant believe you guys tested cheese. How was the curing time for that?
 
Great article.


LMAO at the cheese. I was quite surprised it actually allowed the cpu to stay at idle temps, though that may be due to the oilly grease/fat content. Was any thought given to testing margarine also?:D
 
Great article.


LMAO at the cheese. I was quite surprised it actually allowed the cpu to stay at idle temps, though that may be due to the oilly grease/fat content. Was any thought given to testing margarine also?:D

I would think that butter or magarine would be too soft and the melting temperature would be too low. Its also probably not a bad electrical conductor. The cheese, so long as it doesn't reach its melting point quickly would probably be ok.
 
Really nice article Marc and Kyle. I was hoping to see something like this one of these days. It really lets me know that I made the right choice with MX-2. I might pick up the Thermalright stuff for my next upgrade though as I'm going to be using it on two CPUs (my own and my brother's), a GPU (need to put the stock cooling assembly back on my 9800GTX) or two, and possibly a north-bridge (hoping better thermal paste will help keep the heat down on the 780i's NB). Beyond that I tend to upgrade my CPU or mobo about once a year anyway so it'll be nice not to have to re-buy stuff for a while. I'm not a huge OCer so I don't need to worry about getting the best of the best, not to mention for the price I would never consider Shin-Etsu stuff over the MX-2 as the difference is so minimal.

Thank you for sharing the charts with the community. I know they will be a big help to the users of the site I am with. Hopefully this week I'll get a small article or something written up to use the data in.
 
Another vote for Innovation Cooling 7 Carat Diamond paste. I've been using it for about a month now, don't have the money to compare it against the others though (just bought some 18x9 wheels and now closer to being poor ;) )

I do have an old tube of AS5, dunno if age does anything to it, maybe I could compare the 2 and post some results though since AS5 is in the top 5.
 
I was looking for something like this a while back when I was starting to get paranoid about my aftermarket heatsinks, I ended up just getting a tube of AC5 (reliable and already known) for $5.

Thanks for the review this will help me out in the future, lesson learned here for me was to not really give a crap about what you use as long as you can get it for cheap and clean it off easy (AS5 is not that easy...)
 
Dammit, I was so impressed by the cheese performance at idle and so disappointed when it Failed at the OC settings. Was this true "Cheese" or was it "Cheese Food" ? I am speculating that the oil content of fake cheese aka cheese food would have performed better. Oh well I guess with all the die shrinks making CPUs smaller my dream of a 4GHz grilled cheese sandwich maker is now realy just a dream. Back to work on my single bay mini-pizza oven.

Seriously, very nice tight workup. Excellent work.
 
How can you do all that and not have a "American Cheese = [H]ardOCP FAIL" logo!

Outside of that, very nice work.
 
Was this true "Cheese" or was it "Cheese Food" ? I am speculating that the oil content of fake cheese aka cheese food would have performed better.

This was the real deal Thin-Cut deli sliced American cheese...mmmmmmmm

It worked surprisingly well until it came time to remove it.

Oh Noes! :eek:
 
Cheese. Awesome. How about a shoot out with Colgate, Crest and Cheese Whiz? ;)

I used the paste that came with my U-120. It worked pretty good. I don't know if it is the same as their Chillfactor paste. I used the spread goo method. Keeps my CPU very happy though.
 
Good review, was hoping to see the Zalman thermal paste that you apply with a brush, kinda curious about the temps on it.
 
Thanks! I grabbed the Excel as well for easy reference of the difficult-to-remember S.E. nomenclature.
 
Very nice review,

Did you guys have any trouble spreading the Shin-Etsu pastes? On my last build I purchased some X23 and Thermalright. Attempted to apply the Shin-Etsu and after 30 minutes of trying just said "to hell with it" and used the Thermalright paste. The Shin-Etsu was an absolute booger bear to spread.
 
Need to add:
Tuniq TX-2
OCZ Freeze Extreme
Innovation Cooling 7 Carat Diamond

The first two, I can understand leaving out, but since so many have tried and seen IC Diamond 7 whoop everything else out there, I cant imagine why you wouldnt include the 'word of mouth' champ in such a comparison.
 
Perfect timing for this review to come out.
I was just about to build a system and i have about half of the compounds that you listed.
Thanks for the review.

One thing i noticed though is that while you included cheese, you did omit a test without any paste at all.
I would expect that even without any thermal paste, you should have probably been able to get a CPU working in idle, or even mild load on a non overclocked CPU working without failures.
Something to think about for the next review, or just in case you run out of cheese.
 
awesome review guys. Where do i pick up the cheese? I may be mistaken, but price versus performance it can't be beat!
 
Awesome. It certainly helps to show what each paste is capable of, and what it does over time. The key is the correct method of application.
 
Damn, no Tuniq TX-2 tested. Oh well, good review none the less.

We could have easily pushed into testing 4 months solid. Obviously testing across different machines would have introduced variables that would have fouled the metrics.

THANK U KYLE!!! it's quite rare to see thermal paste shootouts. i wanted somn like this more than 10+ yrs ago... but perhaps it wasn't as common back then =P.

one little recommendation: if the drop vs. spread is such a point of contention why not DO BOTH so u can bet best of both worlds?

spread AROUND the drop-point (edges and corners) but leave the center blank until the drop is applied. apply drop, then squish for full coverage without affecting final temps. that's how i do it >P.

i'm assuming of course, thermal paste is much better than thermal adhesives or tape-formats.

Best I can give you is a Paste Shootout from 2001.

We covered application on page 3, and doubling our testing time to 24 HUNDRED HOURS for such insignificant differences seemed, well, dumb.

FragTape vs. Thermal Paste from 1999. There's your 10 years ago. :)

Wow, all that testing took almost 2 months. There's some dedication! Good job [H], that was an interesting read. Also very interesting was the thinner stuff worked better with just a pea-sized drop. Glad I was doing it correctly with my AS5.

Thanks for the kind words. Doing thermal paste RIGHT is not an easy job. Marc also shoehorned multiple other projects in at the same time so my hat's off to him.

Did you guys really use cheese? Really? Pics!!!

Yes.

Thanks for the review.

Next time could you guys include OCZ Freeze Extreme and Innovation Cooling "7 karat" TIM?

Thanks


/I cant believe you guys tested cheese. How was the curing time for that?

I think something you can take away from this is that high end pastes are just not that much different. We will keep them in mind for next time though.

Great article.


LMAO at the cheese. I was quite surprised it actually allowed the cpu to stay at idle temps, though that may be due to the oilly grease/fat content. Was any thought given to testing margarine also?:D

Next time you get that build together and realize you forgot the TIM, reach for a slice of cheese to hold you over till UPS gets there!
 
This was the real deal Thin-Cut deli sliced American cheese...mmmmmmmm

It worked surprisingly well until it came time to remove it.

Oh Noes! :eek:

That had to be interesting to remove.

I do have a question though. One method I've seen for Artic Silver is taking the paste and going over the whole bottom and kind of rubbing it into the bottom of the heatsink, then wiping off the excess, before applying your normal spread on it. i was always paranoid of electrical conductivity and cleaned off the excess with light swabbing and rubbing alcohol. I can't claim that it has any effect on temps since I don't have the tools to get an accurate reading, but do you think that it could make a difference in performance?
 
very cool article. i had no idea the shin etsu x23 stuff worked so well. hp and apple ship several tubes of them whenever you order a main board replacement. i thought they were all oem crap and never considered using them in any of m builds. i've got tons of this stuff sitting around my work shop. i guess my current tube of as5 will be my last.
 
very cool article. i had no idea the shin etsu x23 stuff worked so well. hp and apple ship several tubes of them whenever you order a main board replacement. i thought they were all oem crap and never considered using them in any of m builds. i've got tons of this stuff sitting around my work shop. i guess my current tube of as5 will be my last.

Yeah the Shin-Etsu stuff is pretty good. Intel has shipped it with their retail boxes processors off and on over the years.
 
i can't wait for the sequel thermal CHEESE shootout! lol. swiss, cheddar, goat varieties, etc.:p

thanks kyle for the previous links. i must've missed those :eek: it's my ADD i swear it! :mad:

i think most of us [h] riggers change their CPU or HS enough that AS5 is more than good enough. besides the 'aging' of silver isn't that significant to negatively impact thermal performance, is it?
 
Would have been nice to see some Easy Cheese, but nice shootout overall.

0f4170b8.jpg
 
Last edited:
If I had to explain the differences it would come down to viscosity. Every paste the performed better when using the pea drop method has a thinner consistency that spreads easily. The pastes that performed better using the thin layer were all thicker and thus required extra help to spread evenly that the heat sink could not provide.
I might try the bead in middle next time with AS5. I've always used the spread method. How thick is the other stuff? I have AS5 and I thought it was pretty thick when compared to the old AS3. Are they like fresh caulk thick? Week old caulk? lol
 
would have been nice to see OCZ Freeze and some ic7 Diamond stuff also. But nice shootout overall.
 
I would also like to have seen the OCZ Freeze as that is what I am currently using. But as was stated, there really doesn't seem to be a lot of difference between any of them. The cheese should have been on a sandwich. At least then you would have something to eat when it was all over. :D
 
Back
Top