Thermal Paste questions

[AgentCipher]

Hey guys,

I'm looking to build a new i5 core pc, and I'm looking to do some minimal overclocking. Since I've never done any OCing before, I always just ran with the default cooling that came in the cpu box, and I know next to nothing about this stuff.

All I can recall is that, back in the day, Arctic Silver was king due to the silver in its paste; however, this led to fried electronics due to its electric conductivity as well if you smeared it just a little. Oxide compounds didn't have this problem, but also didn't have the thermal conductivity of the colloidal metal.

So, what're your thoughts? What should I look for?

Thanks!

 

[Nocturnal]

Really? I still use AS5 on my computers. I haven't really checked lately what is the newest and greatest. In for an update.

 

[MadScientist]

Here's BMR's latest 80 TIM comparison: http://benchmarkreviews.com/index.php?option=com_content&task=view&id=150&Itemid=62&limit=1&limitstart=12

Looks like AS5 is still hanging in there. I ran out of AS5 and gave TIM Consultants T-C Grease 0098 I try due to the no-cure time. It scored in the top five in their previous 33 TIM comparison.

 

[dfuze]

I'm still using AS5 from a tube I bought a few years back and it's never let me down.

 

[Zap]

Here's BMR's latest 80 TIM comparison: http://benchmarkreviews.com/index.php?option=com_content&task=view&id=150&Itemid=62&limit=1&limitstart=12

Hey, thanks for the interesting link. I guess the "A" list is the top 20 of the 80 tested. I think it is important to note that the temperature spread was 0.7ºC, which is probably within the margin of error. The difference between the best and the worst out of the entire group of 80 is 4.55ºC. That's getting to be significant. However, from their "C" group, the worst is only 1.4ºC off from the best of the "A" group. Thus, the top 60 compounds are separated by only a hair.

BMR did talk about application methods on different types of bases and lines versus beads. I think how it is applied will have as great an impact on temperatures as what compound is chosen. I'm pretty sure that a crappy application of Arctic Silver 5 (I had to "fix" someone's computer earlier this year where he used "the single use syringe" ) will get worse temperatures than an excellent application of an average "C" paste.

One thing that BMR didn't go into was specific applications for thicker paste. I know that they mention the viscosity and that thicker pastes were better suited for rougher surfaces, but I liked how HardOCP's small TIM roundup showed the actual differences in performance for how TIM is applied, and that more viscous materials required different methodology for best results. HardOCP's roundup link here. Interesting to note that HardOCP found a 3ºC difference between some TIMs that BMR found less than 1ºC difference. Different test procedures? Different application methodologies?

My take on this is pretty simple. Use any thermal compound and apply it more or less "properly."

I do have thoughts on GPUs, specifically the NVIDIA GTX series (260/275/etc.). For the most part, and especially when re-using the stock cooler, use a high viscosity TIM and spread it over the whole core.

I'm pretty happy that my favorite TIM got a B+ with only 0.75ºC below the top TIM. My choice is Arctic Silver Alumina. My reason for choosing this TIM is that it is cheap (you get around 5x more for the money versus Arctic Silver 5), it does not require a cure time, it is thin and thus spreads really easily (don't know why they put "moderate" for viscosity, maybe splitting hairs?) and it is very easy to clean up.

Let's address AgentCipher's concern. Depending on what CPU AgentCipher has, I'd say the best performance gain would be from an aftermarket fansink versus just using "better" TIM on the stock unit.

 

[sticks435]

Hmmm, maybe I should switch from the AS3 to the Alumina I have. I thought I read somewhere that Alumina was adhesive though, so if you used it, you wouldn't be able to get it back off.

 

[billyb0b]

AS has been around so long and one of the 1st on the scene so it's popularity is high. There are better TIM's out there. The Shin Etsu stuff always gets high marks.

Me personally, I'm using IC7 on my i7 build and some tuniq tx2 on a linux server in the basement. Both are frosty.

 

[sticks435]

Yea, I'm just using it because it's what I have laying around. Not going to spend 5-10 bucks on a tube when I switch processors about every 2-3 years. I don't do extreme OCing, so no point.

 

[Fallen Kell]

I use TX-3 and have for a while. It needs to be applied differently then AS5 since it is "thicker" (more viscous), but when done correctly, it will give better performance. It will also not short out some components if you screw up and spill it or have excess seep out from between the CPU and heatsink onto the socket or surrounding motherboard PCB.

 

[Modelworks]

$1.99 silicon grease.
Been around for 40 years , is hard to apply wrong and works great.

 

[Zap]

I thought I read somewhere that Alumina was adhesive though, so if you used it, you wouldn't be able to get it back off.

It is available both ways.

Alumina

Alumina adhesive

BTW, Arctic Silver is also available as an adhesive.

 

[JEDIYoda]

It could be said and righhtly so that the days of there being one best of mthe best are long gone. Especially when its almost laughable the very small differences in temps between what some would call the top 10.....

Personally I have always liked the Zalman ZM-STG2 Super Thermal Grease, since it first came out.

Just follow Zaps advice and follow the manufacturers instruction when apply the thermal compound~!!

 

[Mike Gayner]

I thought we put all the thermal compound bullshit behind us years ago. Just use whatever comes with your heatsink.

http://www.dansdata.com/goop.htm

 

[daw123]

How long does TIM last if it stays in the syringe?

The reason I ask is that I have two syringes full of AS5 and some cheap Akasa stuff. Both syringes have been 'opened'.

 

[zagood]

AS5 can separate if left in one place for a few years, if you want to be safe you can empty the whole syringe and mix it. Unfortunately there's no foolproof way to tell...if you push a little out and it's watery, yup, it separated, but if it's thick, you can't be sure.

The Akasa will probably be fine.

How long does TIM last if it stays in the syringe?

The reason I ask is that I have two syringes full of AS5 and some cheap Akasa stuff. Both syringes have been 'opened'.

 

[daw123]

AS5 can separate if left in one place for a few years, if you want to be safe you can empty the whole syringe and mix it. Unfortunately there's no foolproof way to tell...if you push a little out and it's watery, yup, it separated, but if it's thick, you can't be sure.

The Akasa will probably be fine.

Thanks zagood

 

[Mike Gayner]

I tried to apply some old silver based grease that has been in a used syringe for a few years. It was hard like putty, and unusable.

 

[Knavish]

I'd like to see a comparison of the stock thermal pad that comes on an Intel or AMD cooler with the temps of the exact same cooler with the pad cleaned off & replaced with an "A+" rated compound, "B" compound, etc...

 

[ZimZum]

Cool laboratory liquid pro

http://www.hwreviewlabs.com/index.p...-up&catid=32:round-ups&Itemid=47&limitstart=3

 

[Zap]

I'd like to see a comparison of the stock thermal pad that comes on an Intel or AMD cooler with the temps of the exact same cooler with the pad cleaned off & replaced with an "A+" rated compound, "B" compound, etc...

Using only anecdotal evidence, IIRC back when "thermal pads" (rubber, waxy or gummy stuff) were being used, removing it and putting on decent thermal compound would indeed work wonders on lowering temperatures.

These days, however, most heatsinks come with some kind of thermal compound pre-applied, not a thermal pad. I don't think the temperature difference would be that much these days. Heck, don't know what the CPU manufacturer uses, but high end GPUs often use Shin Etsu MicroSI of some kind, so it is already decent stuff.

 

[JEDIYoda]

Using only anecdotal evidence, IIRC back when "thermal pads" (rubber, waxy or gummy stuff) were being used, removing it and putting on decent thermal compound would indeed work wonders on lowering temperatures.

These days, however, most heatsinks come with some kind of thermal compound pre-applied, not a thermal pad. I don't think the temperature difference would be that much these days. Heck, don't know what the CPU manufacturer uses, but high end GPUs often use Shin Etsu MicroSI of some kind, so it is already decent stuff.

If i may also add...these days almost all manufacturers use a descent thermal compound or ship with what you otrder a descent compound. Gone are they days of toothpaste being used as a thermal compound!

 

[lenjack]

Arctic Silver is NOT electrically conductive.

 

[zagood]

Arctic Silver is NOT electrically conductive.

Correct. It is, however, capacitive, so not completely safe if you slop it all over.

 

[JEDIYoda]

Cool laboratory liquid pro

http://www.hwreviewlabs.com/index.p...-up&catid=32:round-ups&Itemid=47&limitstart=3

There are issues with "liquid" thermal compounds. Most of the issues are that they conduct electricity, thus even a little drop on your mothert board and you void the warranty of your mobo.
Also ytou cannot clean 100% of the "liquid" thermal compound off!
Then most all "Liquid" thermal compounds are highly corrosive to alluminum...so again be careful...

There are some cons with Cool Laboratory Liquid PRO, first of all, it's a bit hard to handle and behaves similar to mercury (quicksilver). Even if it perhaps not as toxic as mercury, it is still not recommended to digest the paste. The paste also conducts electricity, which makes any spill on the motherboard a fatal mistake. Finally you can not use it with aluminum heatsinks as these will be damaged by the paste.