The Official Thermal Interface Material thread

[Ruru]

I'm wondering if I should buy 3-4 of the lesser known pastes and do a comparison review of them against my "go to" paste NT-H2 and maybe one other like TFX, all on one of my 5950x's. That would be a fun late fall/winter project when I'm not outside as much.

Why not there's always that arguing which paste is the best and how much they perform better than those cheaper ones.

 

[Shrek]

I assume by 'best' people simply mean lowest temperature; I always felt there was so much more to the art, for otherwise 40% diaper cream would be considered a rather fine thermal paste

 

[xzel87]

GD900 checking in here, used it on everything and it works great!

Here's some before and after HW Monitor and test setup. Basically idle temps vs temps running prime95 with AVX off for 10 minutes. Ambient temps when test done is 33.4C +/- 0.3C

Running a Xeon 1241v3 stock clocks (Undervolted 0.1v) with cooler master hyper212 turbo white edition

Idle temps, ambient is around 33.4C

10 minutes Prime95 run with AVX off (small fft)

I don't have the data on hand to show but I am getting better temps vs using Cooler Master Mastergel Pro. I only paid about equivalent USD4 with shipping to get a 30G tube shipped to me. I've had the tube for over a year now and there's still more than half left, the paste separates after awhile but a quick mix (just slowly pull and push the syringe plunger a few times and it's good as new).

While typing this I forgot to turn off the test and the temps stabilised at 69C max, I believe this was running for more than 20 minutes at this point. Also to note, the 3 other cores are cool as cucumber only that one core gets real hot.

 

[Shrek]

I have also moved over to GD900 (120g for $15 with postage), but wonder about its longevity; I got the jar version (30g a jar) so there would be no problem remixing.

Core 0 is also slightly warmer for me; is this generally so (the first task of many always being given to Core 0?)

 

[ThrashZone]

Hi,
Yep core 0 is the main core
Depending on board it may show you in bios which are priority cores with this symbol *

HWinfo also should show this too with perf #'s

 

[xzel87]

I have also moved over to GD900 (120g for $15 with postage), but wonder about its longevity; I got the jar version (30g a jar) so there would be no problem remixing.

Core 0 is also slightly warmer for me; is this generally so (the first task of many always being given to Core 0?)

I have repasted desktop systems almost exactly 12 months ago, but unfortunately these are all office machines so it didn't occur to me to take freshly applied temp data to compare to a later date. I may have a an ancient Q6600 system I can compare to as from memory in idle those 4 cores are in their early 50s after repaste (the stock cooler was meant for a measly dual core pentium, I used it for the q6600 anyways since office has AC).

I may be able to test my wife's laptop as I've repasted hers last year with GD900, but laptops are a hit and miss due to their flimsy mounting mechanism which doesn't have a lot of mounting pressure...so softer paste like the GD900 tends to get pumped out.


EDIT EDIT EDIT UPDATE UPDATE UPDATE

I've repasted my wife's laptop and I ran Prime95 with avx off at at least 30 minutes to get to max temps, cleaned the paste, applied fresh gd900 paste (same tube) and ran the same test parameters to get max stabilised temps. I recorded a difference of 1C cooler with new paste, as far as I'm concerned this is within error margin and the conclusion can be made that the paste has not deteriorated at all in the almost 1 year in my wife laptop. I also did a fresh windows reinstall and let all updates run before running the tests (install was done because i notice her windows was acting wonky and real strange, I haven't taken a look at her laptop and done software maintenance etc. in awhile).

BEFORE REPASTE (almost 1 year old GD900 paste)

AFTER REPASTE (Fresh GD900 paste, from same 30g tube)

 

[Yaxpak2021]

I'be been using MX-5. Seems to do okei compared to more expensive pastes like TG Kryonaout Extreme. I wasted 100 dollars on the TG stuff.

 

[Intervention]

I'be been using MX-5. Seems to do okei compared to more expensive pastes like TG Kryonaout Extreme. I wasted 100 dollars on the TG stuff.

LOL You mean like this?

 

[Yaxpak2021]

LOL You mean like this?

But why is this soooo expensive? Like seriously. I expected the same results as in the video. I saw another one too that compares both pastes. Almost the same crap.

 

[Intervention]

But why is this soooo expensive? Like seriously. I expected the same results as in the video. I saw another one too that compares both pastes. Almost the same crap.

Well, MX-5 I would say is for the casual "I want a good thermal paste that's going to perform and I won't have to worry about for years" kind of paste. While TG Kryonaut extreme is more for extreme overclockers, like LN2 OC. In a lot of cases, folks fall for this hype and spend crap tons of money on products which will not meet their expectations, for their intended purpose.

 

[Yaxpak2021]

Well, MX-5 I would say is for the casual "I want a good thermal paste that's going to perform and I won't have to worry about for years" kind of paste. While TG Kryonaut extreme is more for extreme overclockers, like LN2 OC. In a lot of cases, folks fall for this hype and spend crap tons of money on products which will not meet their expectations, for their intended purpose.

Yeah I've been there sadly

 

[freeagent]

Check out SYY-157, you can get on Amazon. It’s good stuff.

 

[Yaxpak2021]

Check out SYY-157, you can get on Amazon. It’s good stuff.

I have never heard of this one until now. Thanks so much I will do that

 

[Intervention]

Check out SYY-157, you can get on Amazon. It’s good stuff.

I'll have to check that out too. Thanks man.

 

[freeagent]

It’s cheap too, 8 grams for 15 bucks I think.. you would have to check your region

 

[Yaxpak2021]

It’s cheap too, 8 grams for 15 bucks I think.. you would have to check your region

15W/mk???? what? thats insane!

 

[Intervention]

It’s cheap too, 8 grams for 15 bucks I think.. you would have to check your region

Well I live in Norway. I think Amazon UK might have it. Now I want try this LOL

 

[80-watt Hamster]

15W/mk???? what? thats insane!

Take all thermal conductivity ratings with a suitable number of grains of salt. It's kind of like UTQG for tires; there are testing standards, but the resulting values are only really comparable to products from the same manufacturer due to variances in testing methodology.

 

[freeagent]

Its about the same as Thermalright TFX, but easier to apply.

These guys are trying to get their name out there, they make a good product.

 

[Shrek]

15W/mk???? what? thats insane!

I still don't understand this fasciation for thermal conductivity

Power 100W
Area 3cm x 3cm
Thickness 0.05 mm

• Thermal conductivity 10 W/K m -> 0.6°C temperature drop
• Thermal conductivity 20 W/K m -> 0.3°C temperature drop

Is the second paste twice as good? Since the CPU reports temperature to the nearest degree one is probably not even going to notice the difference.

 

[ThrashZone]

Hi,
Yep not enough between chip and cooler to matter all that much
Think it was mostly which ones melt the least under heat.

 

[freeagent]

Indeed. That SYY stuff is fairly thick, there will no pumping anything out

Same with TFX, no pump out because its thick.. I do have a couple of CPUs that make real heat, not that pretend stuff

Earlier I said I haven't seen pump out, but I think I did on my GPU with TF8.

 

[Shrek]

I could be wrong (am often wrong), but don't get how thick will resist pump out.

The forces associated with thermal expansion are enormous, so the heatsink will expand over the CPU and thick or not, some thermal paste will be carried out and wiped away when things cool back down; but it will take a lot of thermal cycles to notice.

 

[freeagent]

I am probably just talking shit to be honest. But so far I have not seen any pumpage on my CPUs.

 

[Shrek]

For me its a big deal if there really is no pump-out, so don't let me discourage you.

I'm also wondering if pump-in is possible, namely a high surface tension that would have the thermal grease pull back in.