Best thermal paste TODAY?

[hypeMarked]

I have a question. Do you spread the compound all over the CPU's IHS? Most of the application guides shown the old AMD CPU which has a very small IHS surface.

 

[JEDIYoda]

Originally posted by: hypeMarked
I have a question. Do you spread the compound all over the CPU's IHS? Most of the application guides shown the old AMD CPU which has a very small IHS surface.

you follow the manufacturers instructions!

 

[VinDSL]

w00t!!!

SHIN-ETSU G-751 THERMAL COMPOUND 1.0 GRAM - $2.99 w/ *FREE* SHIPPING

 

[Fadardo]

Originally posted by: Aflac
That one guy who made his own diamond thermal paste got something like an 8C improvement. The thread's somewhere here...

He's probably just on the DeBeers payrole

 

[Idontcare]

Originally posted by: Fadardo

Originally posted by: Aflac
That one guy who made his own diamond thermal paste got something like an 8C improvement. The thread's somewhere here...

He's probably just on the DeBeers payrole

Actually if you do read thru that thread, and it is quite lengthy, you will see that just about everyone ends up with roughly the same temps between AS5 and IC7.

There are just a very few outliers who see 6-8°C improvement with IC7. I'd bet heavily those guys did not have AS5 installled well even despite their self-assessment that they did know what they were doing.

The overwhelming body of evidence supports this conclusion as well. One guy on that thread even compiled everyone's temps and ran t-test stastistics and showed there was truly no difference between AS5 and IC7. (its all there in the thread)

For example see this post in that thread (updated from thread page 6): http://www.ocforums.com/showpo...?p=5382181&postcount=3

 

[overbyte]

Originally posted by: Idontcare

Originally posted by: Fadardo

Originally posted by: Aflac
That one guy who made his own diamond thermal paste got something like an 8C improvement. The thread's somewhere here...

He's probably just on the DeBeers payrole

Actually if you do read thru that thread, and it is quite lengthy, you will see that just about everyone ends up with roughly the same temps between AS5 and IC7.

There are just a very few outliers who see 6-8°C improvement with IC7. I'd bet heavily those guys did not have AS5 installled well even despite their self-assessment that they did know what they were doing.

The overwhelming body of evidence supports this conclusion as well. One guy on that thread even compiled everyone's temps and ran t-test stastistics and showed there was truly no difference between AS5 and IC7. (its all there in the thread)

For example see this post in that thread (updated from thread page 6): http://www.ocforums.com/showpo...?p=5382181&postcount=3

That's not what he said at all

"" There is more to it, of course. If we're looking at the differences between the two groups, we can calculate a confidence interval (excel doesn't do this, so I plugged it somewhere else) - in this case, the 95% confidence interval for the mean drop in temps would be 1.38 to 5.29, so there is a 95% chance that the average drop in temps from IC7 over other thermal compounds is within this range.""

95% confidence the temp improvement users are are seeing is between 1.38C to 5.29C for an average improvement of 3.3 C of those that reported before and after results

pretty good IMHO


As you are dealing with a sample, averages filter out the noise - For example if you have a few bad mounts with AS5 you also have to assume you have some bad mounts of ICD7 in the mix and if the sample is large enough your confidence level increases as these outliers get smoothed out into the mix.

Numbers in total are more accurate in this format overall than the individual anecdotal posts which are subjective and not indicative of the total picture.

I'll take the 95% view over the 5% on this one

 

[Capt Caveman]

MadShrimp's Spreading it Thin TIM Roundup 2007

1. Tuniq TX2
2. Arctic Silver AS5
3. Arctic Cooling MX-2
4. Arctic Silver Ceramique
5. Nanotherm PCM+
6. Intel stock thermal interface material

 

[Martimus]

Originally posted by: Capt Caveman
MadShrimp's Spreading it Thin TIM Roundup 2007

1. Tuniq TX2
2. Arctic Silver AS5
3. Arctic Cooling MX-2
4. Arctic Silver Ceramique
5. Nanotherm PCM+
6. Intel stock thermal interface material

That was a very interesting article. Thanks for linking it.

 

[VinDSL]

Yay, Tuniq!

Truthfully, my Tower 120s came with Tuniq TX1 - and it looks, feels, applies, et cetera, exactly like MX-1, aka Shin-Etsu G-751...

I'm not a betting man, but I'd be willing to wager that Tuniq TX2 is nothing more than repackaged MX-2! (most likely Shin-Etsu X23-7783D)

AS5 ranking so highly, in the results, makes the whole article suspect in my mind!

Whatever...

 

[Idontcare]

Awesome link to the Madshrimps review. Thanks!!!

That is crazy how much of a difference he got between the dollop vs. pre-spread techniques. Personally am not surprised that pre-spreading increases liklihood of entrapped air pockets...how could it not?

Oddly enough TX-2 is much cheaper than AS5 at Newegg.

TX-2 for $3.99 (3.5g tube)

AS5 for $5.99 (3.5g tube)

Shipping is the same for both.

 

[DrMrLordX]

Actually, TX-2, MX-2, and X23-7783D are all different pastes. I have some X23 and, as I recall, it looked different than either the TX-2 or the MX-2.

From what I understand, MX-2 is just MX-1 with more solvent. It performs about the same but it cures much faster. TX-2 is its own beast so far as I can tell. If it is repackaged X23 then it'd be a surprise to me.

 

[VinDSL]

Originally posted by: DrMrLordX
TX-2 is its own beast so far as I can tell. If it is repackaged X23 then it'd be a surprise to me.

Yep! I'm just conjecturing...

I've never used MX-2 or X23 - I'm so happy with MX-1.

Next time I run out of grease/paste, I think I'll try some Shin-Etsu X23-7783... unless Tuniq TX-2 turns out to be 'Sleeper of the Year'...