Which thermal paste should I use?

Dont get me wrong, MX-2/3 is marginally better from everything I have seen, but it should not be OMG LOOK AT ME better. Unless you get that excited over 1-2C (hey, some might!) ;)

1-2c can make or break a good overclock and add to stability.
 
1-2c can make or break a good overclock and add to stability.

Haha. Not in the real world it cant.

In the real world, most high-end thermal pastes are the same. If your OC-ing to the fringe where thermal paste makes 1-2C degree difference, your pushing your chip too far. Cause in the real world, ambient temps differ. By more than a few degrees.

Whats the difference between a stable OC at 70* C and 65*C? Nothing.
 
Haha. Not in the real world it cant.

In the real world, most high-end thermal pastes are the same. If your OC-ing to the fringe where thermal paste makes 1-2C degree difference, your pushing your chip too far. Cause in the real world, ambient temps differ. By more than a few degrees.

Whats the difference between a stable OC at 70* C and 65*C? Nothing.

Exactly, the room where my PC is fluctuates by about 6-8C because of the sun and seasons. If you are running so close to the edge that 1-2C "makes or breaks" your OC you should look at better cooling overall IMO.

Good paste is a plus, but it is one part of the equation. I am not saying AS5 is the best, it clearly is not, but it is up there and I know how it works/responds after years of using it.
 
So how do those two stack up against Shin Etsu X23-7783D? Is it like 2-3C difference or even higher?

Right now I just got the cpu lapped, planning on lapping Megahalems really soon so wondering if I should get some more extreme TIM to take advantage of both cpu/heatsink lapped.

My cpu's are lapped too - and I do prefer LiquidPro on surfaces like that. I always get very good temperatures with LiquidPro.

If you mount your cooler 4 times in a row - after cleaning and reapply tim - I can almost guarantee that you will get 4 different temperatures within a good degree. That is no matter how careful and correct you do the mount.

skinee labs has done - what I think is - some of the better tests of tims out there @ http://vapor.skinneelabs.com/TIM/IXSupplement/IXSupp.html

If you ask 10 experienced overclockers on which tim is best - you probably will get ten different answers :p
 
Last edited:
Just to be different, I'd like to recommend Thermal Fusion 400, it's the same performance as the MX-2 but available in more places.
 
My cpu's are lapped too - and I do prefer LiquidPro on surfaces like that. I always get very good temperatures with LiquidPro.

My CPU and TRUE are both lapped, and I also like using LiquidPro, I like the very thin film I am able to apply.
 
So how do those two stack up against Shin Etsu X23-7783D? Is it like 2-3C difference or even higher?

Right now I just got the cpu lapped, planning on lapping Megahalems really soon so wondering if I should get some more extreme TIM to take advantage of both cpu/heatsink lapped.

I've got the Shin Etsu on my W/C setup... It's easy to work with, basically performs more or less like AS5 but without the cure time. All you need is a pea sized amount for something like an i7 heatspreader.
 
I've got the Shin Etsu on my W/C setup... It's easy to work with, basically performs more or less like AS5 but without the cure time. All you need is a pea sized amount for something like an i7 heatspreader.

I don't find ShinEtsu (at least the three types I got) easy to work with :p not even after I heated it in hot water before applying it.

My experience with ShinEtsu vs. AS5 is that ShinEtsu is several degrees better - but they are both like sticky clay :)
I still have a tube of AS3 - and I found that AS3 easily outperforms AS5, I guess AS3 is really poisonous tho - and that was one of the reason why it dissappeared from the market.

ShinEtsu is very good for more coarse surfaces like non lapped cpu's / coolers.
 
Based on Benchmarkreview, temp difference between the best(Shin-Etsu 751) and free Thermalright Chill Factor that I got with HR05 is about 1c. I don't think I want to shell out money for 1c but how often would I have to reapply Chill Factor? I mean it not Aluminum Oxide, Polysynthetic Silver or Aluminum Oxide like the best one reviewed but it's Silicon Oxide Grease or it wouldn't really matter?
 
This is a subject of much debate but just know that the difference is about 1 degree.
 
My experience with ShinEtsu vs. AS5 is that ShinEtsu is several degrees better

ShinEtsu is very good for more coarse surfaces like non lapped cpu's / coolers.

Darn, and I have shinEtsu and a lapped CPU.

How much better is the Liquid Pro over the x23-7783D ?

Because of Vapor's tests, and a low low price, I just went ahead with the 7783D.
 
This is a subject of much debate but just know that the difference is about 1 degree.

Yes - nothing can be more true Earley ; as I said "If you ask 10 experienced overclockers on which tim is best - you probably will get ten different answers"

Based on Benchmarkreview, temp difference between the best(Shin-Etsu 751) and free Thermalright Chill Factor that I got with HR05 is about 1c. I don't think I want to shell out money for 1c but how often would I have to reapply Chill Factor? I mean it not Aluminum Oxide, Polysynthetic Silver or Aluminum Oxide like the best one reviewed but it's Silicon Oxide Grease or it wouldn't really matter?

That one degree can be just enough to stabilize you at xxxx MHz.... You never know before you tried - do you ???? ;)


I must say, that's quite a collection of TIM!

I just ordered some Indigo Xtreme to give it a try.

Thanks for your detailed observations! (Thumbs-Up!)

Thank you :)

Belive me or not - last month I threw away twice that amount with tim.... + hundreds of old fans, a LOT of aircoolers, waterblocks, pumps, cabinets +++++... you name it :)
I was cleaning up my old workshop since I will soon move to a better one :D

Sooner or later one have to let go on old things - I wasn't really good to that can you say - I collected all kind of shit. And after some +20 years in the business I had a L O T of shit :)

Darn, and I have shinEtsu and a lapped CPU.

How much better is the Liquid Pro over the x23-7783D ?

Because of Vapor's tests, and a low low price, I just went ahead with the 7783D.


That tim isn't bad - but my experience are

Fine surfaces = good for thin tim
Coarse surfaces = good for thicker tim

And as always - quality and performance costs, hunting for that last ~100MHz can be very expensive :)
 
At work we use ShinEtsu for stress testing and all servers are sold with ShinEtsu as the thermal paste. It is meant for long term, high-end use without reapplication ever.

When we change out CPUs during testing, we scrape off the paste from the CPU and reapply it to the next CPU without using new paste unless absolute necessary. These systems then go in very hot rooms and run very stressful tests for days at a time.

ShinEtsu is made to be there for a very long time.

Just my two cents however.
 
At work we use ShinEtsu for stress testing and all servers are sold with ShinEtsu as the thermal paste. It is meant for long term, high-end use without reapplication ever.

When we change out CPUs during testing, we scrape off the paste from the CPU and reapply it to the next CPU without using new paste unless absolute necessary. These systems then go in very hot rooms and run very stressful tests for days at a time.

ShinEtsu is made to be there for a very long time.

Just my two cents however.

btdt :) I have also used copper paste (yeah the type meant used on bolts for exhaust manifolds ! ) on a server on a oil drilling rig since we didn't have the proper stuff.
Thats 2 years ago and that server is still running - with copper paste :D
 
Shin Etsu is great when its been put in boiling water so it can actually be spread. MX2 is great if you are lazy. It spreads very easily.
 
I have also used copper paste (yeah the type meant used on bolts for exhaust manifolds ! ) on a server on a oil drilling rig since we didn't have the proper stuff.
Any port in a storm! :D


last month I threw away twice that amount with tim.... + hundreds of old fans,....
You shoulda posted your location...I'm sure everything you threw woulda been caught by someone just like a bride's bouquet at a wedding! :D
 
Any port in a storm! :D
You shoulda posted your location...I'm sure everything you threw woulda been caught by someone just like a bride's bouquet at a wedding! :D

I live in Norway :) so freight would have been extremely expensive.
I have been giving away - a lot locally
 
Last edited:
I really like Zalman Super Thermal Grease, its extremely easy to apply and the temps I get with it are identical to Arctic Silver 5 without the and burn in time.
 
Lapped Megahalems (2000 grit) + lapped Q9650 (2000grit) + 1 x Delta afb1212she (1 down)

and PLUS Liquid Pro = sweet ass temps. Was testing out these same settings yesterday (with Shin-Etsu X23-7783D) and I was getting upwards of 58-58-etc etc with fans at max.

Ambient 69-70F.
Vcore: 1.36125V in bios
Vtt : 1.28V

sweet_ass_temps.png
 
Last edited:
Lapped Megahalems (2000 grit) + lapped Q9650 (2000grit) + 1 x Delta afb1212she (1 down)

and PLUS Liquid Pro = sweet ass temps. Was testing out these same settings yesterday and I was getting upwards of 58-58-etc etc with fans at max.

Ambient 69-70F.
Vcore: 1.36125V in bios
Vtt : 1.28V

sweet_ass_temps.png

Do you know what kind of temps you would get with the same setup and AS5? I really haven't heard of many people using the liquid pro stuff. My mega shadow is in the mail :)
 
Do you know what kind of temps you would get with the same setup and AS5? I really haven't heard of many people using the liquid pro stuff.

I was getting in the upper 50s with Shin-Etsu X23 when I was testing yesterday and according to review sites AS5 is 1-2C worse then Shin-Etsu so either way big drop. I guess I should mention the TIM i had on before in my post.
 
I was getting in the upper 50s with Shin-Etsu X23 when I was testing yesterday and according to review sites AS5 is 1-2C worse then Shin-Etsu so either way big drop. I guess I should mention the TIM i had on before in my post.

Hmm. I think ill stick with the AS5 for now, but that is very interesting. Since its liquid on application I bet it allows a very thin layer, probably very good if your stuff is lapped nicely. I will consider trying it in the future after seeing those temps.
 
Hmm. I think ill stick with the AS5 for now, but that is very interesting. Since its liquid on application I bet it allows a very thin layer, probably very good if your stuff is lapped nicely. I will consider trying it in the future after seeing those temps.

Thats pretty much the reason I got this. Once I got both the chip and heatsink lapped SE (Shin-Etsu) became a PITA to install. I tried three different times, pea in center, spread, thin spread and got about 3-4C in between the applications at worst.

Versus a perfect SE or AS5 application there probably shouldnt be more then 2-3C difference but unlike AS5 and SE this stuff is very easily to apply.


I used a very small amount of Liquid Pro, around 2mm by 2mm pea drop and still after spreading there were little droplet clusters.
 
Indigo Extreme is top of the heap by all modern tests. It also has a more "interesting" installation.
Other liquid metals are next (and are harder to install)
IC7 Diamond is probably top of the line for any regular "thermal paste", but is harder than other "thermal pastes" to install/spread correctly
TIM Consultants 0098 is next on the list, and the best "traditional' style thermal paste. It was developed by former shin-etsu engineers. This is what I'm currently using.
Shin Etsu comes next and is the new "standard" for high performance thermal paste.
Everything else is below..
 
Haha. Not in the real world it cant.

In the real world, most high-end thermal pastes are the same. If your OC-ing to the fringe where thermal paste makes 1-2C degree difference, your pushing your chip too far. Cause in the real world, ambient temps differ. By more than a few degrees.

Whats the difference between a stable OC at 70* C and 65*C? Nothing.

Maybe. My i7 doesn't really seem to care how hot it runs at 4GHz but keeping my Q6600 at 3.6GHz was always a balancing act between temps volts and if I sneezed.

Let me put it this way, if using a $10 tube of thermal paste and spending 10min to redo the application to run cooler why not? Besides, this is [H], pushing our stuff "too far" is what we do.

Also, my ambient temps don't change that much, my computer room be it winter or summer is 95% of the time 22c, almost year round.
 
Back
Top