Which Thermal paste

Meridiusuk

Meridiusuk

Meridiusuk

Associate
Joined
9 May 2009
Posts
1,072
Hi all

I am going to get a 3900x and using a bequiet dark pro 4 Heatsink and would like to know what is a good paste.

I did some reading and it seems to between the Noctua NT-H1 or MX-4 but wanted the best temps i can get.

I have read that there is now a HT-H2 paste but not to sure on this one

would be great if you could let me know what your thoughts on this

thanks
 
Meridiusuk said:
thanks but not after reading this i am not going anywhere near it as they seem to have poor quality control, i can not trust somthing like that

https://forums.overclockers.co.uk/threads/kryonaut-abrasive.18827970/
Click to expand...

Never had any such issues with mine. Have had several tubes over the years and all has been very consistent.

I’ve tried mx4, mx5, gelid extreme and it beats them all by a few degrees from my testing.

If you are worried then gelid extreme is good, used to use that before I moved over to kryonaut.

Best bet is looking at a testing comparison and deciding from there if you want maximum performance.
 
Meridiusuk said:
mm been reading up and the NT-H1 suffers from somthing called pump-out effect, have no clue what that is yet.
Click to expand...

From what I've been able to ascertain, since I still have an unopened tube of that, and the H2, it appears that according to some tests, the viscosity of the H1 (doesn't say about H2), is rather high, and so when applied to certain types of situations (directly to the die of chip I guess from what I'm reading; largely on laptops), it starts to spread rather easily (from heat and pressure I believe from what I'm reading), necessitating a reapplication not long after. But from the sounds of things that's not what you're doing anyway, so the H1 should be safe.

Personally, I have the H2 as well, and have not used the H1 yet (H1 came with U12A and I bought the H2 for extra, but found it had plenty available after installation and use on GPU as well). But I largely got it because whilst I have very dextrous and flexible fingers and hands, ultimately I'm a huge dude, and so rather than tempt fate whilst I repaste, I got the non-conductive or capacitive H2 over something like say Arctic Silver, or Thermal Grizzly. So if I got any of it anywhere I shouldn't, it wouldn't kill anything. Performance wise, the H2 I'd expect is very accetpable for many except for the most ardent users who want the very best performance.
 
All of the major name brands TIMs transfer heat about the same. Take liquid metal TIM off the list and the difference is about 3.6c between next 26 TIMs. When testing TIM between 2 copper plates the difference is 3.36c for top 36 TIMs not liquid metal. Gelid GC Extreme and Kyronaut are best, but if you have to worry about less than 3.5c difference you have much bigger problems.

Key is not so much what TIM you use, but how good a print you get. The difference between a good TIM applicaton (print) and a poor one is way more than the 3.5c difference in preformance between TIMs.
 
Meddling-Monk said:
the viscosity of the H1 (doesn't say about H2), is rather high, and so when applied to certain types of situations (directly to the die of chip I guess from what I'm reading; largely on laptops), it starts to spread rather easily (from heat and pressure I believe from what I'm reading), necessitating a reapplication not long after. But from the sounds of things that's not what you're doing anyway, so the H1 should be safe.
Click to expand...
High viscosity means the substance has a high resistance to flow, therefore won't flow easily.

Low viscosity means the substance has a low resistance to flow, therefore flows very easily.
 
an0nym0us said:
High viscosity means the substance has a high resistance to flow, therefore won't flow easily.

Low viscosity means the substance has a low resistance to flow, therefore flows very easily.
Click to expand...

Thanks, I knew I it sounded funky when I wrote it but was in a rush at the time and didn't have time to double check. But yeah, that. :)
 
I've always preferred mid to low viscosity TIM because we don't want a layer of TIM between IHS and cooler base, we want metal to metal contact with TIM filling voids. We want as much metal to metal as we can get because heat transfer is like 50 times better than though TIM and hundreds of time better than air. The plated copper IHS and cooler base have a heat transfer rate of 400 W/m K, if cooler base is aluminum it's 205-227 W/m K depending on it's temperature. TIM is between about 5-38 W/m K and air (no TIM) is 0.024 W/m K.
 
JasonM said:
ID Diamond, it's incredibly reliable with performance at least on par with MX-4.
Click to expand...

Yeah, I my tube of IC - , wait... ID Diamond? Never heard of that one :p ;). I had a tube of "IC" Diamond 24 though, but used it up and hence why on Noctua's NH1 and 2 now. Agree it was quite impressive and non-conductive too. The only major downside to it was ease of scratching surfaces and also discolouring of surfaces too if you weren't careful. My H100 and H100i plates were slightly discoloured and minutely marked by them even with the extra care. But great performance otherwise.
 
Meddling-Monk said:
Agree it was quite impressive and non-conductive too. The only major downside to it was ease of scratching surfaces and also discolouring of surfaces too if you weren't careful. My H100 and H100i plates were slightly discoloured and minutely marked by them even with the extra care. But great performance otherwise.
Click to expand...

I don't care about the scratches, but to be honest I've never really noticed anything myself.

I have IC diamond that's been applied 7 years ago on an i7 3770k, i'm at 6 years on 2 x i7 4770K's. The stuff is super reliable and this is the reason I use it.
 
You must log in or register to reply here.
Back
Top Bottom