Good Thermal Paste?

With proper application, which can change your result by at least several degrees, you really can't go wrong with any of the better thermal pastes. And by better, I mean anything arctic silver ceramique and above.

Conductive/capacitive semantics? Bridging transistors? You can't really talk about it if you have no clue.

 

So... I'm wrong? You can slather on AS5 onto points where electric current runs and it'll be fine?

I was wrong about the capcitive/conductive thing. Fine. The point I'm trying to make is a messy application of, say, MX-3 won't do naughty things to your components like a messy application of AS5.

 

I feel like you would have to try really hard to mess anything up. I played with it and my multimeter around 5 years ago and I couldn't get anything out of it. Also there really aren't any horror stories about using it and it has been out there for what, 6 years? The only reason it isn't as popular as it used to be is because there are slightly more thermally conductive pastes out there. In all that time, including a few years as the best and most popular paste out there, "incidents" were and continue to be nearly unheard of. As long as you aren't a complete nitwit and squeeze out the tube into the socket you should be fine.

 

I just blew coke out my nose. Thanks.

 

Alright I get it and get the humor. I do not want to read all 6 pages. Is all the stuff I read about ICD and a few others not really of issue? AS5 or above all really work. Has been a few years since I dug into this. But back then much this brand vs that brand was hype. Are we still there?

 

I think the stock paste will be ok for a while, but it might bleed.

So I would get some other paste, whatever you like best.

 

ICD is the best, its not too expensive if you look around (well not in NA anyways, maybe harder to get in other areas)

Its just as good or better than all the other top pasts with little or no risk factor.

Its also pretty fool proof to do a good install since it uses compression install methods. AS5 for instance while a great TIM requires a thin layer to be spread about. More often than not novices would do a bad install of the paste and get undesirable temperatures.

I do believe that this is in great reason why ICD is doing so well because the number of users that perform a bad install are significantly less. If you did AS5 correctly chances are there will be almost no real difference in temps

ICD also has another inherent advantage in being so thick so it bridges gaps and fills holes so if your cpu or heatsink has imperfections, it easily makes up for it, it can even be used as a thermal pad replacement (done it myself speaking out of experience)

 

Just FYI for the readers here. When a poster says;

Brand X is the best.

What that really means is;

Brand X is my favorite.

Please don't confuse opinion with fact or intertwine drastic, emotion-fueled conclusions with findings based on solid, scientific based testing. Do your research and you'll end up with a good product and a good installation.

With a little effort you can't go wrong as long as you choose one of the commonly accepted quality thermal interface compounds and install it carefully.

 

I saw my temps drop when i applied AS5 over the stock thermal paste. May be the stock thermal paste was getting a bit old i don't know. However, compared to the AS5 ICD7 gave me better results.

I was a bit sceptical at first but it worked for me. If you haven't tried some of the recommended brands then may be try them.

AS5 took my lowest temp down to about 25C, this doesn't indicate poor application at all. ICD7 though undercut that and i saw my lowest temps drop to under 23C.

Edited to add:
When i first applied AS5 at idle my lowest temp was 24C but this was basically doing nothing and after a cold boot. After some months i started averaging about 27C. I swapped to ICD7 and my like for like temps dropped.
I've been using it for about 6 months plus.

These are my most recent results after my computer had been on for quite some time. I can get it to drop lower but well that's kind of pointless really.

http://forum.notebookreview.com/att...d1274190018-sony-sz-owners-lounge-thermal.jpg

 

No I mean ICD is the best

Sure its an opinion, but its an educated opinion.

Reasons:

A.) Its made from diamond materials (thus the name) and as that is the most dense common material it has a much higher thermal conductivity compared to metals like used in most other compounds.

B.) Its not very expensive, some of the other good compounds carry very high price tags. Despite the high level of performance they do not try to gouge the prices with this product.

C.) Its not conductive, again within the relm of high performance TIM the greater majority of them are conductive to electricity. This makes them more dangerous to use and limits the flexibility of the product.

Use AS5 as an example, yes you can carefully install it on a cpu die or gpu die, but what about as a thermal pad replacement ontop of a RAM chip on your GPU. Even if it had the characteristics to hold its shape it would short out the GPU and kill it.

If one product can do something another cant, it makes it better.

D.) Ease of installation. Many TIM's are messy, sticky, runny, hard to work with, and if not installed perfectly do not give the proper benefits. ICD uses compression install put a glob in the center of your application area and install the heat sink. Physics will do the rest leaving you with a pain free perfect install with no air gaps, and no hassles.

E.) Does not deteriorate over time (for incredibly long amounts of time). Lab tests have been done in extreme situations to test the durability of many top end TIM's and it looks like ICD was the one that held out the best. The others being more liquid based would dry up and leave air pockets and gaps resulting in loss of efficiency.

F.) None destructive to materials. Yet again another characteristic associated with some high performance TIM's is how it is corrosive to certain materials. As far as I know ICD has no materials that produce this effect.

Sum it up. Sure I do have an opinion that ICD is the best. However I can reasonably conclude that in a overall comparison of competing products its also factually the best as well.

You can of course create your own sub-set of "best" and say ask "whats the best TIM for the $$$" or what is the "best" tim for ease of removal.

Within specific categories you may conclude a different product is "best" and have your own different opinion. But when one does not specify a specific category and just uses the term "best" it would convey that one means "overall" and in that regard ICD is the best based on all the factual data I have accumulated via studding TIM's for over the last decade.

 

That's a bit of a paradox you have there, since all liquids and semi-solids have some Material Thixotropy, so when you say you take that out of the equation, you instantly remove all the conductive greases we're talking about. If you are careful about the application, and assuming we have a new synthetic substance that can hold its viscosity while remaning a liquid/grease, then you should not have any problems, again, assuming you're careful about the application of such. otherwise, I don't kow of any grease that can be used as a thermal gap pad replacement, so it doesn't matter whether it's conductive or not.

being that diamond is the hardest material on earth, wouldn't squishing it between a copper heatsink and silicon die be extremely abrasive?

I wouldn't say that's true, most high performance tims have universally not been conductive since manufactures stopped using heavy metals like copper and silver 5-10 years ago and began using carbon and diamond powder greases.

Again, I cant see how diamonds wont cause micro pitting/scratching when you squeeze it between two surfaces, and until we know the average micron size, I personally won't be using diamond powder products. Although I haven't looked around for such problems, I just can't see how it wouldn't. The stuff is basically polishing compound that has high thermal conductive properties.

 

If you ever need to remove ICD 7, check to see whether that material has caused scratches on the DIE surface. With it being such a hard dense material many have warned that it could cause a lot of scratches. As I said before, I did notice this when I tested it on an old unused chip but I don't know whether this happens in general use.

Here are some comments from others who noticed similar results. If you google you can see a few more.

I have applied thermal paste on CPU's and GPU's many times with out any mess, problems or any material going anywhere it shouldn't be but it does not mean that I would use a paste that is conductive when I have other options that work just as well, if not better that are non conductive.

 

IC Diamond 7 is not the best. A couple of Shin-Etsu pastes are at least as good if not better.

 

I would say that the above statement is subjective at best. With the myriad of hardware loadouts available, not to mention the types applications that the hardware will be placed under, it's difficult to say if anything is really "the best". You would need to test all 3,292 types of TIMs in order to ascertain what is "the best" for a given hardware configuration.

 

Which is why I don't call anything the "best."

 

lol, as soon as somebody says something is, "The Best!" I can't help but discredit anything they say in my mind from that point forward.

 

IC Diamond has put some very light scratches on my X9000. I still prefer it over AS5. I am not saying that AS5 is a bad Thermal paste(I still use it on family/friends computers) but ICD has given me lower temps.

 

No, you're just saying that for your applications, Shin-Etsu is a better TIM than ICD7.

 

What is your problem? I said IC Diamond 7 is not the best and that a couple of shin-etsu pastes are at least as good if it isn't better. What is so hard to understand?

 

Shin Etsu X23-7783D is probably better than IC Diamond.

 

Antec Formula 5 is good. Dropped my temps by 8C, on processor.

 

Your results are subjectively based on the system and applications you based your TIM test results on. ICD7 may not be best for you with your testbed, but it may be better than Shin-Etsu for someone else on theirs. What is so hard to understand?

 

I can attest to the fact that ICD7 scratches the CPU and GPU dies, it did so in my laptop. It seems to scratch the dies and make fine lines and smudges on both the heatsink and the die. I just went back to the Shin X23, I've used it for ages and its been the best so far.