Thermal Paste Recommendations?

[Rroff]

Working better than MX-6 for me also in current build, go figure.

In my experience in terms of performance MX-4 and MX-6 are identical except in full load temperatures on high TDP CPUs like the 14900K where MX-6 can be 2-3C lower, for most stuff I'd just use which ever of the 2 you can get cheapest, though the 9950s might warrant MX-6. For say a 7800X3D you probably won't see a difference between them.

There is some difference in how easy they are to apply but neither are difficult to use.

 

[Pulseammo]

MX-6 is what I use now, I like that it's a bit thicker but it's still easy to spread. Temps are decent too. Well priced as well usually, so value for money wise I'm very happy with it.

 

[martos123]

Using MX-6 in my current setup and happy with it. Easy to apply and also to clean too.

 

[Gray2233]

As an aside to all of this, if you buy a Thermalright HSF (and frankly, if you need a new HSF you probably should) they actually come with a syringe of their own brand TFX (it ranges through 4-8) paste, there's a few versions and I forget the exact you end up with. This includes their more entry level coolers such as the Assassin King 120 which sets you back all of £15-20.

Regardless, it's about on par with Arctic MX4 and 6 and there's enough for several applications. It's a stark difference to most coolers which come with an adhesive pad on the block, or a sachet or whatever nonsense.

 

[ssjninjamonkey]

Most are within margins of error of each other from all I have seen recently been looking myself, I decided to go with grizzly duronaut but all charts pretty much showed 1-2c ~ if that.

 

[Troop]

MX4 for the last 10years, easy to use and clean off when needed, and cheap, been on the same tube for 10 years.

 

[Mcnumpty2323]

Just bought 8g of mx6
Cost less than 1g or 2g of some of
The so called best/most expensive thermal paste too
Unless you absolutely need to eke out
The very last degree you can
Seems a waste of money to spend too much
On premium pastes

There's also ease of use to consider
And pump out/longevity of the thermal paste too
And whether the same paste works as well on gpus
As on cpus
If you also need to use it for that

 

[ATIorNvidia]

IMO just use whatever comes with the cooler.

 

[Brackish]

MX-4 gets my vote, it has been my go to for a while.

 

[deviation]

I have never noticed any difference when repasting the same CPU with a different paste, when I've run out of the previous brand and found some from a different brand.

 

[Mr Oxide]

Noctua NT-H2 and has been solid for me.

 

[StephenJ]

Noctua NT-H2 and has been solid for me.

Would have to agree with this Noctua NT-H2 for me

 

[jigger]

If you have some Kryonaught left then I’d use that, it’s free!

AS5 is the best all round paste though.

 

[Slogan]

Tend to use MX-4 most of the time although recently used some Honeywell PTM7950 for a laptop application. Wouldn't generally be wringing the neck of my CPUs most of the time but never had any issues.

 

[Nevakonaza]

MX-6 OR Thermal grizzly Duronaut

Both excellent pastes.

 

[Ice Tea]

At some point an elderly person will suggest AS5 even though it hasn’t won any benchmarks for years.

 

[Coogs]

Are you de-lidding?

 

[doyll]

Been using Thermaright TIM for as long as they've been selling it, no problems at all. That said, some is much easier to apply than others. I found it's much easier to apply if I heat tube in cup of hot water for a few minutes before applying.

Quality of seat is way more important than what TIM is used. A poor seat / print with best W/mk rated TIM won't transfer heat as well as a poor seat / print with much lower W/mk rated TIM.

 

[jigger]

At some point an elderly person will suggest AS5 even though it hasn’t won any benchmarks for years.

There is more to a TIM than showing the lowest temp after being freshly applied.

 

[Rroff]

At some point an elderly person will suggest AS5 even though it hasn’t won any benchmarks for years.

Most benchmarks are a poor test in reality especially of longer term results. Many benchmarkers of thermal paste persist in using the same application method for all pastes in a test for "consistency" despite the application notes for any given paste might prescribe a different method for best results.