Arctic Silver 5 is no longer the best thermal paste.

I mention this in every thread I see it come up, but AS5 is not the best thermal paste anymore, in terms of temperatures and safety. It's a capacitative paste which means it can damage your hardware if it makes an improper connection. I personally recommend the Tuniq TX-2, it's what I'm using in my current desktop build. As for temperatures, here are some reviews:

http://www.madshrimps.be/?action=getarticle&number=5&artpage=3071&articID=635
http://www.cluboc.net/reviews/cpu_accessories/thermal_paste/tuniq/tx2/2.asp
http://www.overclock3d.net/reviews.php?/cases_cooling/tuniq_tx-2_thermal_paste/5
http://lanoc.org/reviews/?p=247
http://www.3dxtreme.net/index.php?id=tuniqtx-2p1
http://www.modders-inc.com/modules.php?name=NDReviews&op=Story4&reid=186

Now, for why AS5 being capacitive is especially a poor choice in notebooks:

Both Tuniq TX-2 and Arctic Cooling MX-2 are superior choices for both temperatures and safety of your components. I know AS5 was the best recommendation for a long time but I've seen enough threads where people on here are recommended AS5 that I think this thread is needed to clear things up a bit with more recent data.

Summary for those who don't like reading long posts: Arctic Silver 5 has been surpassed in cooling, design, and hardware safety by both Arctic Cooling MX-2 and Tuniq TX-2, so stop recommending AS5 to notebook users.

 

Thanks for the info...will give the mx-2 a try at some stage.

 

im wondering if its because of the difference in viscosity, and the fact AS5 needs a fairly long curing time.

would be interesting to see a test with "fully" cured AS5

 

Read the reviews I linked, one of them let it cure for 3 weeks.

 

i did read them, i must have missed the one that did that

 

If I were to take anything away from this it would be other factors besides purely thermal performance.

The few good, quality, hi-performance TIMs all will work reasonably well for the average user when measured solely in thermal qualities. It now becomes other attributes that hold sway, with safety, being one of the foremost, IMHO. In-service and shelf longevity, along with ease-of-application, rank highly with me, as well.

I definitely could see recommending non-conductive/non-capacitive TIMs for their safety to those with little experience in working with electronics. I've seen some amazingly horrid applications of AS5 et al in the various fora around the net!

Thanks for posting.

 

http://www.overclock3d.net/reviews.php?/cases_cooling/tuniq_tx-2_thermal_paste/5

The Mediocre
• Unproven stability over time.

Hmmm, how long do I need to reapply the Tuniq TX all over again? Does the AS5 last forever?

 

AS5 wont last forever, but i havent noticed any performance degradation over a period of 5 years

im not sure about the others, as i havent tried them before

but yea, some thermal pastes tend to degrade fast.

 

At the time of review, it hadn't been out long enough to hear stability over time. I haven't heard complaints about it online, and mine's working just fine, so in that review mediocre is simply because of an unknown, not any tangible complaints.

 

I'll stick with AS5 why? "If it ain't broke dont fix it"

Since I know what I am doing no danger of causing a short and I know it performes great and last longer than I would keep a computer for.

If i want MAX performance there is something better than both of these:
http://www.coollaboratory.com/en/product_metalpad.shtml

The liquid metal pad is almost 100% perfect fusion of the cooler & object. It has a fool proof install as there is no "best method" and trying to spread paste around. You just cut the metal pad to the right size and install it. It will melt and form a perfect bond after you start up.

It was almost 3C better than AS5 when I used it on my C90S.

 

There are better thermal pastes out, however the the cost and availability AS5 is probably the most accessible solution. I've tried finding other brands and it's a lot harder to find, ie; MX-2 some computer stores have never even heard of it. That said, AS5 is still way better than the stock grey stuff that comes with OEM coolers which is why it's still recommended. I've also seen some reviews of the Zalman ZM-STG1 being comparable to both AS5 and MX-2.

 

In my regular machines I've taken them apart as long as 4 years later and my AS5 was still fine in consistency.

In my Distributed Computing pharm I've seen the AS5 dry up in as little as a year. But these were running full-load 24/7. The machine in question was running ClimatePrediction.net on a Q6600 quad core, all 4 cores fully loaded 24/7, overclocked, overvolted, and water cooled. Not typical use, I'd venture.

 

Newegg reviews:

http://www.newegg.com/Product/ProductReview.aspx?Item=N82E16835154003

Here: http://www.techpowerup.com/printreview.php?id=/Sunbeam/TX-2 - the conclusion is that it does not perform as well when broken in. Ehh what's the point, if you apply AS5 well you're fine. I've never heard of anyone actually causing shorts due to AS5, it's so easy to apply...if I didn't already have AS5 and needed new paste, I would definitely go with TX-2 though.

 

Techpowerup's test is far less thorough than others, besides conjecturing about long term performance.

 

Where do you get the liquid metalpad? I found a German seller on eBay (not feasible because I'm in the US). I'd like to try that stuff.

 

http://www.sidewindercomputers.com/colimeforamd.html
http://www.crazypc.com/products/liquid_metalpad-50125.html

Sounds okay, but I need to see more tests. Supposedly removal is a pain.

 

A while ago, I came across another TIM that greatly outperformed any others but it cost about 2x as much for the same amount.

EDIT: it was Shin-Etsu (not sure which specific TIM)

 

If that's the case, then why even bother with AS5? Why not just don't touch it at all?

Seriously, it's all hype for the fanatics to drop just a couple to few degrees, while on some machines you won't even see a difference. In fact, I've seen many cases where the temperature actually increased because he or she couldn't/didn't assemble the components back properly. I've owned a few laptops over the years and even my oldest (7+ years) one is still running with the original OEM stuff.

 

It really depends on how effective the notebook cooler was built, of course if a user doesn't know how to reassemble a notebook/PC back together properly they shouldn't be applying thermal paste in the first place.

Most cases temperatures don't drop by very much, but that doesn't mean that AS5 isn't more effective than stock. Subtle differences such as a slower temperature rise, or slower less noisey fan rotation, lower max load temperatures etc. is considered an improvement and worth the additional effort. And don't forget enthusiast overclocking benefits as well.

Longevity and cost is the only reason why OEMs still prefer the stock grey stuff. Most other thermal paste aftermarket requires frequent replacing (less than 2 years) required to be re-applied for maximum cooling benefits.

 

I'm not convinced.

The only review that showed any difference was the one that openly admitted to having a heat sink that wasn't being connected properly (and thus constantly). The others all showed a difference of only a few degrees.

Also, the whole 'it’s not capacitive' thing is irrelevant. You should never leave something on electronic components because they add risk. This will 100% attract all the dust in your case. Dust has been known to cause short circuits along with other damage. You really want to risk getting it on a high-power component and thus shorting it out, months after you apply it? Just clean the stuff with some alcohol and who cares if it is capacitive. Even conductive shouldn't really matter.

In the end, until TX-2 has reviews that say it last for 10 years if needed without any loss of cooling ability, I will stick with the AS5 you can find everywhere. (Is $1 difference really relevant to the cost of your computing?)

The only thing to hope for really is that AS comes out with a good product that beats both of these to regain their lead over the competition.

 

Agreed. The only application area for Thermal pastes are CPU, GPU and Northbridge cores. Most of these cores are surrounded by the green PCB material which is fibreglass plastic composite which are not capacitive. Even if a 100% capacitive thermal compound bleeds over, it will not short anything as long as it doesn't get onto the motherboard components. You will have to be very careless to do something like that.

 

I use Shin-Etsu x23

 

What about cost? Is AS5 viable as a cheaper alternative?

 

Costwise, most are about the same, AS5 just has the brand name out there so it's the 1st one people think of.

 

I use the Liquid Pro, not the metal pad, from the same company. It works wonder, and easy to clean and to remove. Just follow the instruction carefully, a drop goes a long way.

And yes, this thermal liquid "paste" has outperformed the AS5 that I used before on my comps.

cheers ...

 

http://www.hardwarelogic.com/news/137/ARTICLE/2752/2008-03-03.html
Another site that did a huge review.
Personally i think the "Cure time" in AS5 is a load of BS.

I haven't used AS5 since when i swapped by CPU From E6550 to an E8400, i had a HORRID baked out effect on my sink/CPU. I use OCZ FreeZe now, or IC Diamond 7 for all my cooling needs now.

 

This reviewer certainly didn't have such a good time with the liquid metal pad.

 

For the amount of lingo they use marketing the TIM's, you'd think they'd be charging more than $4.00 for 3.5 g. I don't even see why cost ever enters into the discussion.

 

I did maybe 20-25 desktop builds, rebuilds, and CPU upgrades over the last ~5 years using AS5 exclusively. I even attempted several times to follow the AS5 break-in thermal cycle pattern. I can't say I ever saw anything positive from this break-in effect. Even when my machines were under constant load conditions and I had digitally controlled air-con maintaining a relatively constant room temp during the summer, and watching for the effect, I never saw it.

I admit, these weren't scientifically controlled studies, but really, if anything we're talking a couple degrees C, if that. That's w/i the window of our ambient room temp variance.

IMO, practically speaking, a TIM's other qualities become more important than trying to squeeze the last .001% of thermal performance out of it.

 

In my personal opinion, is it really worth trying for a 1C drop by changing TIM brand.

However, saying this I did have a pleasant surprise the other day. I ordered a no-name water block as it was cheap and I needed to make a HTPC set up. It came with some, what I thought was, cheap looking silver TIM. It resembled the thickness of double cream when I got it out of the tube (it poured out basically). After cleaning up some of the mess by sucking it back up into the tube and connecting everything back together it actually runs very well, personally better than the AS5 I had in there originally.

Just goes to show that you may spend 10's of pounds on good TIM but applied correctly its usually just similar sort of results.

BTW, the best TIM was found to be tooth paste, well for about 10mins until it dried out. This is because of the micro fine particles that crush easily when you mount the heat sink.

 

Ceramique ought to work sufficiently for most people, and it's also non-conductive/capacitive.