Good Thermal Paste?

I got some free thermal paste (the package called it grease ) with my processor and I wanted to know if it was good or not.

It's packaging is half chinese and half english, but it clearly says that it's thermal conductivity is greater than 7.5W/m-k and it also says it has a thermal resistance of .06 Celsius/W. Is this decent enough paste or should I just go out and buy some for 6 bucks?

 

I recomend IC Diamond. It is one of the best on the market.

 

ASII ----- 10char

 

Where could I purchase this? I'd rather not get anything off the internet. Could you buy this at Wal-mart or something?

 

I think you would have to get it online.

 

Yep, unless you know some custom/repair shop nearby that might carry thermal compound.

 

AS 5 10char

 

AS5 is not the best anymore.

ICD is the best and arctic cooling mx2/3 are better.

The stock paste is going to be pretty bad compared to those.

 

I used Arctic Cooling MX-3 on my D901C's Q9650. Works great.

 

I used Antec Formula 5 on Asus G50VT-X5 and it dropped my CPU temp by 8C. Its under $10. I bought mine from eBay, but you can get it anywhere.

 

Most of the good thermal pastes don't differ much in performance, I just tend to get the one that is fairly cheap and easy to obtain and apply. It might not be the best but for a 1-3 deg C difference, I'm not going to go through the hassle.

 

OCZ Freeze is great and cheap.

 

I've used Arctic Silver Ceramique and Matrix and they've done very well for me. Break in time is looong though.

I also have IC Diamond (bought on ebay) but I've heard its not the easiest to install, so I am waiting to use it on something that I'll not be opening very often.

 

Do any of you guys have like a before and after? Temp with standard thermal paste, then with the new type. With mine it would be 37C idle with standard thermal paste. With Antec Formula 5, 29C. Room temp 25.5C.

 

If you want the absolute best performing thermal paste go for Shin Etsu G751. Here are some comparisons:
Techware Labs - Reviews - Thermal Paste Comparison
ITShootOut.com :: Thermal Compound Shoot out - Summary and Conclusion - Page 4

the problem is that it's a little hard to find (eBay is a good place to start)
and it's harder to apply.

 

You guys suggestions are great and all, but I'm also wondering if my random branded thermal paste will work. I got it with my CPU that I bought off ebay. I found the site that sells it, but I can't find it on the site. Apparently it's only 25% silver.

I also found another site that sells it.
Thermal Grease - Detailed info for Thermal Grease,Thermal Grease,Thermal Grease,HS-A1 on Alibaba.com

Edit: I found it.
http://www.pccooler.cn/cn/proview.php?pro_typeid=247&pro_typeid2=260&pro_id=1582

 

wrong. pre burn in maybe, after as5s burn in period there's very few (only 1 infact according to this review) pastes that can match it

80-way Thermal Interface Material Performance Test | Thermal Interface Material,Thermal Paste,Heatsink Compound,80-Way Thermal Interface Material Best Thermal Paste Heatsink Compound Cooling Performance Comparison Benchmark Tests

 

Well those don't match my results.

How long does the burn in take?

 

i'd steer away from those no name brand pastes. good ones are so cheap that there's really no reason not to use them.

 

as5s burn in period is up to 200hours. I agree for instant drops there are better solutions. however over time as5 will increase it's performance.

 

The problem with all these reviews is that the temperature difference all lie under standard error. In one test, X might score 1 deg better than Y but in another, it might be reversed. As I stated above, go through the review and pretty much any of the non-generic pastes will be suitable. Then go to your nearest computer shop or online and check out which one is easiest to apply and is fairly cheap.

 

yup they are all more or less the same.

 

while i normally agree with you guys, according to this chart : ITShootOut.com :: Thermal Compound Shoot out - Summary and Conclusion - Page 4 there's a clear winner.

 

Arctic Silver 3??? AS2???? That is one old, outdated comparison. Was hoping for something more current. I'm a big proponent of Shin Etsu though. Before that I was an AS5 user, then switched to ICD and now Shin Etsu is giving me best results.

 

That article is from 2003

 

yes, it's quite old. i can't find newer articles, but from my experience it is better than mx3 and as5.

 

here's one i found now : (it has as5)
http://enthusiast.hardocp.com/article/2009/04/07/thermal_paste_shootout_q209

 

I'll probably get shot down for suggesting it, but I've used automotive copper grease for over a decade now. It always works well, is easy to clean up, and dirt cheap. On the desktop, with a decent cooler on an E5300, cores are 35C and 38C. Ambient temperature in the room is a little over 21C.

Those figures are under 100% load... system has been up for over 2 weeks, still under 100% load, and never gone above 40C.

 

I wouldn't really use automotive copper silicone. It's designed to keep oil and water out and be resistant to heat, not to conduct it. Almost any liquid or grease will conduct some amount of heat, thermal interface pastes just do it better generally speaking. Unless you're talking about something like anti-seize? In which case yes, generally that can be used due to it's high metal content.

 

The review you posted uses 120hour curing time, This is not within the guidelines of the instructions for the application of AS5, the review I posted does use a 200 hour curing time for AS5. It does showcase however, that as I previously said, before the curing time has taken place there are better solutions for a quick drop in temps.

 

In the hardocp test It did show AS5 tied as the best at the end of 120 hours however. That speaks very highly for AS5 and if it's going to improve over the next 80 hours even better.

In the 80 way thermal interface material test AS5 again scored at the top with a perfect score of A+ (Along with a few others.)

80-way Thermal Interface Material Performance Test | Thermal Interface Material,Thermal Paste,Heatsink Compound,80-Way Thermal Interface Material Best Thermal Paste Heatsink Compound Cooling Performance Comparison Benchmark Tests

AS5 is certainly still a very valid choice when picking a product and one can expect stellar results.

Personally I think people might be placing a little too much importance on TIM and spending a little too much effort on it. There are many good products out there to use and none of it is going to fix severe cooling problems do to a poorly planned installation, cooling/airflow set up or within the tight confines of a laptop. It will help of course, but none of it is going to grant you miracles.

 

I had AS5 on my old t5550 for well over 200 hours and still lost a few degrees after installing IC Diamond. Also the fact that IC Diamond is nonconductive and easier to clean up is a plus for me.

 

Artic Cooling mx-3 it's the best and very easy to use, won't short circuit your system too if you have too much

 

as5 Is non conductive as well(as are 95% of pastes these days), and I would find easy to clean up a subjective response.

 

Also application plays a big role in temps. You can put it in there what you think is the same way for five times and get temperatures that vary by a couple of degrees. Also the way you put it on plays a big role too. I've found the best to be a very very thin and even layer over the die before installation. That way there is a lot less in the way compared to putting a blob on there and smashing it down.

 

Arctic Silver 5 is non-conductive? Did they do a new formula recently?

 

Found out I can get AS5 at my local Radioshack.

 

I didn't say copper silicone, I said copper grease. Yes, the same stuff as anti-seize compound. I used it when in a bind one day, and found it compared favourably to other pastes of the time. I'm not offering benchmarks, or group tests, just the advice from someone who spent a good degree of time working in the industry, and that it's not a bad performer and hugely cheap.

Having said that, and whilst I've used it for many years now, there may be better alternatives out there. But back at the turn of the century, it performed very favourably in my experience compared to the "flavour of the month" uber-hyped branded stuff.

 

Last I heard Arctic Silver 5 was thermally conductive and ICD 7 was great at scratching DIE surfaces.

 

I believe arctic silver 5 has always been non-conductive, it is however slightly capacitive

 

I have tried AS5 and AS Ceramique in the past and I recently ordered a few bottles of the Zalman ZM-STG1.

The reason why is really how it is applied. Easier to apply a nice, even, thin layer without wasting a bunch of the stuff. As for results in the 80 way tests I also question the actual experiment.

1. Same paste when tested repeated will obviously give you a range of results.
2. They used more than one heatsink. Heatsinks are not going to be identical in surface variances or composition.
3. Human error. The application of 80 kinds of thermal paste cannot be easy to replicate in the same manner.
4. Was the room temperature the same throughout the tests?

I honestly think the most important thing is to apply the paste properly no matter what paste you use. They should all do the job quite adequately.

Edit: Added reason #3 & #4.

 

I saw this years ago...

Thermal transfer compound comparison

 

I've tried AS5 and ICD7 and ICD7 lowered my temps about 4-5C lower than what i got with AS5.
You can get ICD7 from ebay as well.

 

yeah, I can definitely see that being used ina bind, the stuff is ~80% of what copper pastes were 5-10years ago anyway.

JK has it right, AS5 is completely non-conductive, however they recommend(which is just common sense anyway) that you don't glob it on transistors or pins and what not since it is slightly capacitive.

 

Conductive, capacitive, meh. I guess I should learn the symantics.

In any case, that's still the point. If you apply AS5 badly--like you said, bridging transistors--it'll be bad news.

 

in a worst case scenario yes. However I wouldn't be comfortable with any paste making contact with transistors.

capacitive is the ability to hold a charge, conductive is the ability to move a charge, it's not semantics.

 

Most CPUs and GPUs don't have sensitive circuitry right up against the socket (since you need room to apply the heatsink). As long as you do a little bit of planning beforehand, there is no reason why you should short out anything.

 

Agreed but with so many good pastes available that pose zero risk in that sense, I see no reason to risk it at all.

 

Anybody who can't install thermal compound without slopping it all over the place has far greater issues than choosing the right interface material. These particular people shouldn't take panels off their computer to begin with. To me that's a moot issue. Personally, I'll use any quality product and have allegiance to none.

 

I agree, if you have to calculate risk of not doing a clean job, you shouldn't be doing it at all.