Artic Silver 5, Ceramic or Noctua

[Countryman]

I am about to buy a Noctua NH-U12F. Can anybody advise as to whether I should use the heatsink paste supplied or buy some Artic Silver 5 or AS Ceramic. If the latter, should I apply to just the cpu, or the heatsink as well as per the Noctua instructions?

Many thanks.

 

[gmanjt]

I'd like to know also.

 

[BonzaiDuck]

First, read this:

Citarella at OverClockers, January '07

and this:

Citarella's DIY Paste Experiment

I'd recommend earlier articles at OverClockers on this topic, but the chart of thermal conductivities shown in Citarella's 2005 article says it all.

I will also be candid: I believe Citarella is entrepreneurially involved in the company known as "Innovation Cooling." Despite that, I also experimented with a concoction of micron diamond abrasive powder and JetArt CK4800 -- with much better results than Citarella's, but the JetArt uses diamond and was formulated to use it.

So despite J.C's "financial interest," I guarantee that IC Diamond is better than Arctic Silver 5 -- and for processors over-clocked to between 110W and 150W of thermal power dissipation, will show an improvement over AS5 of 2+ C degrees. At around 105W, off the top of my head, Innovation Cooling's own test shows an improvement of about 5 to 6F degrees (that's Fahrenheit -- 1C = 1.8F).

The silver compounds degrade in effectivness over time, while the diamond products do not. Silver particles could cause problems electrically (although the material in AS5 is mostly capacitative and minimally conductive), but the diamond has no electrical conductivity or capacitative properties -- you can spread it all over circuit traces and components and there will be no adverse effect at all.

Here's the best that's available:

Innovative Cooling IC Diamond Paste at Heatsink Factory

 

[VinDSL]

AS5 and AS Cermaic is GREASE not PASTE...

BonzaiAirDuct recommends Tinky Winky Icey Diamonds, or whatever it's called - the latest gay fad GREASE...

If you really want the best PASTE, get MX-1

 

[BonzaiDuck]

VD -- somebody oughta kick you off these forums. You're behaving like a puerile little *****le.

This is a site and forum for serious discussion of computer hardware, software and related issues.

The facts are clear about the products. The IC Diamond paste (or grease, or thermal interface material, or whatever you want to call it) -- trumps Arctic Silver 5 and it trumps JetArt CK4800 (the latter also using nano-diamond particles, but only at a 10% mixture).

Anything that improves thermal transfer and reduces temperatures is just better, period. And that goes for heatpipe coolers as well. If the cooler reduces temperatures and improves thermal transfer over your second-best Tuniq, it's just better, although one can weigh in evaluations about ease of installation, design-flaws in contact with the heatspreader, and other things we've discussed.

Now, I think I will go to the moderators to complain. About you're being a moe-ron.

 

[GalvanizedYankee]

Originally posted by: VinDSL
AS5 and AS Cermaic is GREASE not PASTE...

BonzaiAirDuct recommends Tinky Winky Icey Diamonds, or whatever it's called - the latest gay fad GREASE...

If you really want the best PASTE, get MX-1

VinDSL acting out like the bull of the woods is why I left the Cases & Cooling Forum. Meh.

Why bother why complain, it just gives this airhead bully attention.

 

[Fallen Kell]

I was going to also mention MX-1....

 

[BonzaiDuck]

Fallen Kell --

You might find a review for us that either compares to AS5 or IC Diamond. You probably won't find it compared to IC Diamond, because that product is so new to the market.

The ShinEtsu paste doesn't give an indication of its contents, and I haven't found anything about the MX-1.

My basic assumption, though, unless I see some white-papers on more complex materials or compounds (chemically speaking) that have peculiarly robust thermal conductivity, is that the chemical element with something like triple the conductivity of silver is Carbon in the crystalline (diamond) state.

Someone else had sent me a news item about an Asian professor who had found that carbon-black also had promise, but it had to be applied under extreme pressure, so it was unfeasible at this time.

Other mixtures use aluminum oxide and boron hydride in addition to silicon, and then there's that metallic stuff that behaves like mercury and offers some risk to your electronic components -- unlike diamond.

We need to get some comparisons in benchtests from reviews on the latest products. IC Diamond was only released in May. Citarella's article that provided comparison testing was published this January, and his earlier experiment with diamond powder was published in 2005. I'd be willing to bet that he had a financial interest in Innovative Cooling, but since the product proves itself as expected -- it doesn't really matter. Now -- in the case of our elected officials, their ties to certain companies, and their "warranty" about WMDs and terrorists, it looks a lot different than that. But no more. I'm just making a comparison -- not trying to stimulate a political discussion that belongs on a different forum.

 

[BonzaiDuck]

I wouldn't have said anything to anyone about mentioning MX-1 -- let me emphasize that.

I objected to someone personalizing the discussion with veiled insults -- originally insinuated -- and substantiated -- within the last day or so on another thread. There has been a pattern, and I've seen others become exasperated about our "fellow member."

Without pinning a tail on that . . . . source . . . . I noticed also someone pointing at a review as though it was rave proof of performance, but the review excluded some competitive products from the comparison, and it offered no facts. The facts may have been there at one time, but weren't available when I followed the link.

More and more in recent history, we find supposedly professional researchers forging results, just as we find people raiding corporate assets, stealing from the public, and doing other dishonest things. So it is always possible that people will forge or falsify facts to support their idea or their product. Since this occurs in professional academic journals, one should be even more suspicious or critical over gaps or unproven claims in commercially-oriented hardware reviews.

We live in a time when Greed and Short-sightedness trumps Integrity and Honesty.

 

[VinDSL]

Jeez! Sorry!

I apologized to you, BD, in another post... and I'll do it here too!

I was in a whimsical mood last night, and I guess I went a little overboard.

I'll try to temper my posts, lest someone take PERSONAL offense!

It's not my intention to drive anyone away from these forums - au contraire.

The funny things is - I'm always the most loved and most hated person on ANY forums that I'm on - all at the same time...

Anyway, sorry, dude!

 

[JEDIYoda]

There is alot of real fine pastes on the market!!
With that said there really is no need to go out and purchase a tube of past over what came with your heatsink unless the tube has the words Colgate or Crest on it!!!

But if your looking for the best then once again it`s trial and error and of course $$$.

Knowing Bonzai like I do he has been there and done that and would not steer anybody wrong!!
The thing I really like about Bonzai is all his opinions are based on actual usage of time and materials.
No unfounded opinions!! Just the facts!!
Peace!!!

 

[BonzaiDuck]

Actually, Vin, why don't you also apologize the EEYore? I mean, you quoted him as saying "blah blah blah . . . . "

Whether he, or I -- present large volumes of information or pass comments about the weather is nothing for you to demean with these comments. It's just not very mature. I can say that I [emphasis] have still some aspects of immaturity as I approach 60.

I can also say that as a much younger man, I was homophobic and made jokes about gays, but I accept them as people now. Nevertheless, casting innuendo at someone, as you did -- whether or not they are "of that persuasion" -- is not polite. It's downright rude, especially with the insinuation of the dictionary quote. I CAN say it was subtle, but it was neither polite nor civil. And that also goes for referring to a product as a winky twinky gay fad.

You can have fun here, surely. But we're trying to discuss issues about hardware and software -- cutting edge computer technology. If something bores you to death -- don't read it. If two people are having a polite exchange -- do not belittle them.

And think before you write. I also can continue to practice getting better at that, but I haven't made the sorts of remarks or innuendo you made, except in responding to them.

. . . . and I'll try to avoid abbreviating your handle to "VD" even if it means fewer keystrokes.

If you think MX-1 is better than IC Diamond, you should just find some reviews with benchtest results, and say something like this: "Well, BonzaiDuck, IC Diamond may be good, and diamond is superior for thermal conductivity, but look at these benchtest results. This MX-1 stuff must be pretty good."

Then, none of us here -- Galvanized Yankee, JediYoda, I or others -- wouldn't be wasting all this time discussing behavior . . .

 

[VinDSL]

Originally posted by: BonzaiDuck
I can say that I [emphasis] have still some aspects of immaturity as I approach 60.

Holy sh!t...

Double sorry, gramps!

 

[Eeyore69]

BDuck,

Been off for a while but did want to let you know a little more about Shin-Etsu. I finally took out my Q6600 processor and the two chipset heatsinks and lapped them. One thing I found out about aluminum (the two chipset heatsinks) is that aluminum doesn't lap nearly as nicely as copper; the aluminum has coarser lines even when using used 2000 grit paper.

But, as I mentioned before, I was going to try the Shin-Etsu x23 that I bought. That was a mistake. I started with the 2 chipset heatsinks but I found that the SE is impossible to apply. It is thick and dry like old wood putty. I couldn't get it spread very thinly (SE recommends .001"). When I tried to spread it very thin it wouldn't stick to the chip or the heatsink. The only way I could get any to stick was to press the SE down without trying to smear it across the chip or heatsink and then it was too thick. I tried heating it and the heatsinks with a hairdryer, heating the tube of SE in boiling water and just working it for a while with my finger in plastic wrap. That was of no help. I did finally get some on both chips (not very smoothly) and installed the heatsinks but decided to use AS on the cpu. When I reinstalled everything my cpu temps are now 2-3C cooler but I am worried about the chipsets. I tried to increase my OC from 3.2G and am only able to get up to 3.45G. I don't think it is the cpu since the temps are in the mid 40C range at 100% load on all four cores with it OC'd to 3.45G (1.3 vid). Tomorrow I am going to replace the SE on the chipsets with AS and see what happens. I will never use the SE paste again!

By the way, VD doesn't really bother me. I've just gotten into my 60's and come to realize that modern day kids are much different than when I grew up. Disrepectful and sometimes just downright mean to other people. And most of the time they are unwilling to listen to other opinions and always try to insist that their opinion is the only one that counts. When I run into such people, whether in person or online, then I just ignore them. I am getting older and I don't have the time left in life or the inclination to waste my time with them. I will stick to conversing with people that are more understanding and respectful of others.

Eeyore

 

[VinDSL]

Originally posted by: Eeyore69
By the way, VD doesn't really bother me. I've just gotten into my 60's...

I am getting older and I don't have the time left in life or the inclination to waste my time...

Mind if we test the theory?

Originally posted by: Eeyore69
But, as I mentioned before, I was going to try the Shin-Etsu x23 that I bought. That was a mistake. I started with the 2 chipset heatsinks but I found that the SE is impossible to apply. It is thick and dry like old wood putty. I couldn't get it spread very thinly (SE recommends .001"). When I tried to spread it very thin it wouldn't stick to the chip or the heatsink. The only way I could get any to stick was to press the SE down without trying to smear it across the chip or heatsink and then it was too thick. I tried heating it and the heatsinks with a hairdryer, heating the tube of SE in boiling water and just working it for a while with my finger in plastic wrap. That was of no help. I did finally get some on both chips (not very smoothly) and installed the heatsinks but decided to use AS on the cpu. When I reinstalled everything my cpu temps are now 2-3C cooler but I am worried about the chipsets. I tried to increase my OC from 3.2G and am only able to get up to 3.45G. I don't think it is the cpu since the temps are in the mid 40C range at 100% load on all four cores with it OC'd to 3.45G (1.3 vid). Tomorrow I am going to replace the SE on the chipsets with AS and see what happens. I will never use the SE paste again!

Please see my sig...

Shin-Etsu, et al, rocks!

Sorry if YOU can't figure out to apply it... but, oh, well!

 

[VinDSL]

Originally posted by: JEDIYoda
The thing I really like about Bonzai is all his opinions are based on actual usage of time and materials.

Oh, please...

I've been using/recommending Shin-Etsu paste for over a year, plus I pass them every day on the way to work!

I've warned ppl a zillion times what a PITA it is to apply - yet they're still surprised. No pain, no gain! No story, no glory!

I think I know a little bit more about Shin-Etsu MicroSi than you-know-who... Most ppl think it's made in Japan! :roll:

 

[DrMrLordX]

Originally posted by: VinDSL

Shin-Etsu, et al, rocks!

Sorry if YOU can't figure out to apply it... but, oh, well!

In all fairness to Eeyore, the X23 he's working with is not the same stuff as MX-1 (G751). X23 should be a wee bit better than MX-1, though I have no experience applying X23 to chipsets/chipset heatsinks; I did apply it to my CPU using the old AS5 "grain of rice" method where I allowed the downward pressure of the heatsink to spread the paste out over the IHS.

Seeing as how he struggled to apply the paste to chipset heatsinks, I'm thinking he may have encountered problems applying the paste to course surfaces. I've read that MX-1 is better on course surfaces like that, while X23 clearly dominates on reasonably smooth surfaces (machined properly) or lapped surfaces.

As to whether or not IC Diamond 7 is worth the money, well, it's cheap (as cheap as AS5) and better than AS5. So, why not? MX-1 is also better than AS5, and so is X23.

So far as I can tell, you're going to do well with any of those pastes over AS5, though I would choose between the IC Diamond 7 or the X23 (MX-1 only for course surfaces, in normal situations X23 beats MX-1/G751 by around 1.5 degrees C).

 

[MadScientist]

Vinny, Vinny. I'm surprised you didn't mention AC's new MX-2 paste. They claim it's better than AS5 and their own MX-1. Link. Google for reviews. Available from Jab Tech..

FYI. Thread at OC forum on IC Diamond 7 paste. Biggest complaint is application.
Joe C's reply:
The Customer feed back on application has been mixed so I would like to recommend a failsafe method.

The best way to apply this would be to apply a pea sized amount on the center of the IHS, Then mount your heatsink - not a full mount just enough to squish it to the edges of the IHS - Separate And let solvent dry for 10 min. Then reattach permanently. Excess will just squish out

This will spread the compound evenly, better than a finger spread or razor spread would and leave less waste.

The thick character or "tackiness" of the compound is the nature of the beast. If we thinned it anymore you would lose performance and it would be subject to pump out affecting long term stability.

This was our best compromise, we had one version that had double the performance of ICD7, But was so thick as to be pretty completely unusable except for maybe 1 or 2% of the totally hardcore market That might want that extra 1/2 -3/4 deg C it would have provided.

Volume wise you will find we have the same amount you will find in AS5, using the method described above you should be able to match coverage area of other TIM's on the market

 

[Eeyore69]

DrMrLordX,

Thanks for the info on Shin-Etsu (Mine is the X23-7783D variety). My heatsinks were all lapped to a fairly smooth surface. At first I tried using the finger/plastic wrap method of spreading and then using a credit card to smear the grease; it seemed like it should work since the initial consistentcy out of the syringe was similar to AS. But it seemed to dry out very, very quickly to where I could not use the normal AS application method. When that didn't work because of the thickness of the grease I tried the methods suggested by people on other forums; I heated the grease and heatsink, applied a "prime" coat, put small dots on the heatsink and then pressed down on the grease to get it to spread. This last method worked somewhat but it left the surface irregular and bumpy. I installed the heatsinks with the grease in that condition but I am wondering if there might be air pockets in the contact area, because of the initial bumpiness of the surface, which are compromising the heat transferring ability of the grease.

I am somewhat wary of trying again with SE. Since I have no way of monitoring the temps of the P35 and ICH9 on my motherboard I wanted the best possible heat transfer and I would rather be sure of decent heat transfer using AS than having poor contact and heat transfer with the SE. But I might try your method on my cpu since I can monitor the 4 core temps; but the current ambient temp is about 70F and the hottest core is 39C with all 4 running at 100% (1.300v). I wonder if it is really worth the effort to change to SE when my cpu temps are fairly decent; yesterday when the ambient was ~95F the highest temp I saw was 48C? Then if that improves the cpu temps I can try again with the chipset since I think the chips may be what are restricting my OC to 3.4G.

Thanks again for the good information (and not just worthless criticism),

Eeyore

 

[DrMrLordX]

Well Eeyore, as the old saying goes, if it ain't broke don't fix it. I wouldn't recommend messing with your current TIM application on your CPU seeing as how your quad isn't breaking 39C at load (which is a good temp).

It would be a bit hard to tell you what to do with your chipset heatsinks unless you're willing to go out and get a thermal diode or something else that would give you fairly accurate temp readings on the chipsets. The way I look at it, the old "grain of rice" method is probably what you'd want to use with either AS5 or X23, so in the end, you'll probably be okay with either one versus what ever TIM was applied by the mobo manufacturer. Using multiple small dots I think it supposed to help on multi-core CPUs (AS actually changed their application instructions, at least for Core 2 chips anyway), but I don't know what it would help on chipsets much. I would agree with your concern about possible air bubbles being in there the way things are now.

Glad to help in any case.

Originally posted by: MadScientist
I'm surprised you didn't mention AC's new MX-2 paste. They claim it's better than AS5 and their own MX-1. Link. Google for reviews. Available from Jab Tech..

I've heard a little about MX-2 (yes I know this was directed at Vin, not me). From what I've seen, MX-1 does about as well once it has had more time to cure. MX-2 cures faster. Otherwise I've seen reported temps around 1C lower with MX-2 which puts it in about the same ballpark as X23-7783D, maybe just behind it.

The best way to apply this would be to apply a pea sized amount on the center of the IHS, Then mount your heatsink - not a full mount just enough to squish it to the edges of the IHS - Separate And let solvent dry for 10 min. Then reattach permanently. Excess will just squish out

This interests me. Wouldn't removing the heatsink for 10 minutes and then re-applying it invite the possibility of air bubbles? I know that comes from Joe C, granted.

 

[BonzaiDuck]

The advice about the IC Diamond seems quite sound, and in fact, I've done it that way.

The diamond paste has a 90%-plus diamond-particle loading. It's abrasive stuff, but even removing the heatsink after "squishing" [not swishing -- see below] -- and letting it "dry" for 10 minutes will not remove all the oils or base. so little "icing peaks" will again be compressed and spread out.

The idea of placing "little dots" of the stuff on an initial layer is also a good idea. It is difficult to spread, but the cooling advantages make "ease of spreading" a rather dumb and silly performance basis for comparison.

No matter how you come up with an equivalent, you will pay more per application for the diamond TIM. The amount sold by IC is enough for maybe one GPU and one CPU. There IS a performance advantage of between 1C and 3C in temperatures -- someone here cited performance differentials of 3/4C degrees in different applications, but there is still an advantage. This performance margin would, of course, narrow as thermal power increases, just as it would for the heatsink of choice.

If you use JetArt CK4800 (10% diamond loading), you can mix the equivalent of IC Diamond by adding abrasive diamond dust rated at below 2-micron particulate-size, but 5 grams of the stuff is going to cost you close to $100. Even so, in comparison to IC Diamond's $5 price-tag, 5 grams and several $5 tubes of CK4800 are going to offer myriads of reapplications if ever needed. But for any single computer -- you won't need to "reapply." The stuff doesn't deteriorate. If you're worried that some of the base has evaporated, you could probably add a third-of-a-grain of rice-worth of CK4800 and it would minimally degrade the loading, but I've found that the paste is still gummy enough so that you can just replace the heatsink without doing anything to clean and replace the TIM.

Until I see comparison reviews -- there was a good comparison of several TIMs done in 2005 -- and the recent comparison reviews show that some exotic concoction for things like Shin-Etsu and MX-2 really work as well or better than a TIM based on particulate synthetic diamond, I'll stick to my recommendation of that paste, which one of our members called "the latest tinky-winky gay fad." [I should promise never, ever, never ever to bring up this matter again with such quotes.] Anyway, the particles are a dull, gray dust, and have no value set in silver-earring mounts. As for that -- I don't even wear a class ring or a "white-gold" dress-watch. But that's all about affinity for jewelry and other things, and this stuff looks more like freshly mixed concrete before it sets.

 

[JEDIYoda]

Originally posted by: VinDSL

Originally posted by: JEDIYoda
The thing I really like about Bonzai is all his opinions are based on actual usage of time and materials.

Oh, please...

I've been using/recommending Shin-Etsu paste for over a year, plus I pass them every day on the way to work!

I've warned ppl a zillion times what a PITA it is to apply - yet they're still surprised. No pain, no gain! No story, no glory!

I think I know a little bit more about Shin-Etsu MicroSi than you-know-who... Most ppl think it's made in Japan! :roll:

I didn`t mention your name because i was addressing what I have always liked abour BonzaiDuck!! Even when we first butted heads!
He has become a good friend on these forums!!

As I hope we can be!!

 

[JEDIYoda]

Also I learned early on as I bought better and better heatsinks and started collecting them that 99 times out of 100 they all shippd with real good quality thermal paste!!

In fact I cannot recall any of the many heatsinks I have ever shipping with crappy thermal paste!!

I have used every heatsink I have for at least several weeks before shelving the heatsink!!


Peace!!

 

[VinDSL]

Originally posted by: JEDIYoda
I didn`t mention your name because i was addressing what I have always liked abour BonzaiDuck!! Even when we first butted heads!
He has become a good friend on these forums!!

As I hope we can be!!

I understand, and can appreciate what you're saying, but...

The thing is, these <seasoned citizens> demand your respect, simply because they made it past their boozing n' drugging days!

The fact of the matter is, old guys aren't any smarter than you and me... they're just more cautious - and ppl tend to mistake caution for wisdom - when it's actually just cowardice, you know? They'd probably prefer to see me spilling my guts in Iraq rather than voicing an opinion on AnandTech, while they sit at home collecting pensions and playing around with ICey Diamonds...

Plus, these kind of guys tend to have 'glass jaws', e.g. they can dish it out, but they can't take it!

Anyway, I like BD & eyeSore just fine! They're the ones that have a problem with me, don't you think?

For gawd sake, we're just discussing thermal compound here...

BACK ON TOPIC

Originally posted by: DrMrLordX
I've heard a little about MX-2 (yes I know this was directed at Vin, not me). From what I've seen, MX-1 does about as well once it has had more time to cure. MX-2 cures faster. Otherwise I've seen reported temps around 1C lower with MX-2 which puts it in about the same ballpark as X23-7783D, maybe just behind it...

My understanding is... MX-2 is simply MX-1 with more solvent added to the mix (easier to spread - cures faster)!

Am I wrong about this?!?!?!?

 

[DrMrLordX]

You're probably right, which might explain why MX-2 comes out better in some reviews that do not allow for curing. I've heard MX-1 does just as well after 200 hours of operation.