Earlier today I finished retesting of NT-H2. Paste pattern I saw after taking plate off indicates a) I didn’t use too much/little of paste, and b) Noctua’s pattern recommendation leads to using more paste than necessary, applying single 5-6mm central dot and spreading it with spatula seems to end up just right. Still that didn’t lead to any improvement in NT-H2’s temperatures over previous attempt, leaving them slightly better over stock paste and Kryonaut.

If the new pastes work just as good as the old paste, what is the problem?

Prolimatech PK-3 and Gelid GC-Extreme arrived today.
 
PK-3 was almost as good as Noctua NT-H2 but overall I was more pleased with Noctua.
 
GC-Extreme was huge flop. I got 3.5g tube from Performance-PCs.com that arrived in packaging used before they changed it some months ago (blister plastic one).
 
Attempt to spread it with spatula was impossible. It felt completely un-malleable. Nothing would stay on IHS, all of it would get picked up by spatula.
 
Attempt to apply it by heating up syringe and using single pea drop didn’t improve anything. Temperatures were 3C higher than with Prolimatech PK-3 and climbing. When I removed cold plate leftover paste was again hard and as dry as a desert.
 
This, in combination with Luumi’s YouTube video about bad experience with GC Extreme and reports by others of their bad experiences (including temperatures going up not that long after the application), leads me to believe Gelid has quality / quality control/consistency issues and that I can’t trust my system to it long term wise.

HeavyHemi

ZoranC

HeavyHemi
I believe you're misunderstanding the entire purpose of TIM. The LESS TIM between the die or IHS and your cooler the better.  More TIM means LESS thermal transfer. Ideally you'd have a perfect machine match between surfaces. Of course that is not possible. So, TIM act to fill those micro voids in the surface to increased contact area and thermal transfer.

My understanding of TIM is exactly what you said. Please notice I've more than once said I'm not expecting paste to form edge to edge layer completely covering surface, but not practically nothing in between either. CLC360's plate has very thin concentric groves in it (they can be seen under 7x magnification) so I would expect that to be filled. At least stock paste does.

It can't be if you're making the same fundamental error of believing there should be  a LAYER of a particular thickness of TIM between the surfaces. In fact ideally you should have 'practically nothing' between the surfaces. In fact you should have coverage from edge to edge of the die. Again, for some reason you believe the TIM layer is what enables efficient contact. It is not. It is the TIM maximizing contact area by filling microvoids. The LESS TIM you have between mating surfaces the better. This fundamental precept will not change.

It seems you are hearing my words way you want to hear them, not the way I am trying to communicate them. I am unable to communicate it any better than I already did.
 
Regardless of that my bottom line result remains unchanged: Stock paste seems to hold up rather good when compared to others. Kryonaut doesn't seem all that and a bag of chips, and others report the same. NT-H2 is doing better than Kryonaut. PK-3 is almost as good as NT-H2. MX-5 is doing noticeably worse than Kryonaut. And GC-Extreme was absolute disaster, same like number of others reported.

ZoranC

HeavyHemi

ZoranC

HeavyHemi
I believe you're misunderstanding the entire purpose of TIM. The LESS TIM between the die or IHS and your cooler the better.  More TIM means LESS thermal transfer. Ideally you'd have a perfect machine match between surfaces. Of course that is not possible. So, TIM act to fill those micro voids in the surface to increased contact area and thermal transfer.

My understanding of TIM is exactly what you said. Please notice I've more than once said I'm not expecting paste to form edge to edge layer completely covering surface, but not practically nothing in between either. CLC360's plate has very thin concentric groves in it (they can be seen under 7x magnification) so I would expect that to be filled. At least stock paste does.

It can't be if you're making the same fundamental error of believing there should be  a LAYER of a particular thickness of TIM between the surfaces. In fact ideally you should have 'practically nothing' between the surfaces. In fact you should have coverage from edge to edge of the die. Again, for some reason you believe the TIM layer is what enables efficient contact. It is not. It is the TIM maximizing contact area by filling microvoids. The LESS TIM you have between mating surfaces the better. This fundamental precept will not change.

It seems you are hearing my words way you want to hear them, not the way I am trying to communicate them. I am unable to communicate it any better than I already did.
 
Regardless of that my bottom line result remains unchanged: Stock paste seems to hold up rather good when compared to others. Kryonaut doesn't seem all that and a bag of chips, and others report the same. NT-H2 is doing better than Kryonaut. PK-3 is almost as good as NT-H2. MX-5 is doing noticeably worse than Kryonaut. And GC-Extreme was absolute disaster, same like number of others reported.

Hmm... well, can't say I've ever had as much problem as you have with using TIM and I've use quite a few different types. Hard to believe many of them are more viscous than IC Diamond.
Anyhow, you're not hearing what others are posting to you, but I guess you'll just have to learn the hard way. I'm not fond of wasting my time.

HeavyHemi
… can't say I've ever had as much problem as you have with using TIM and I've use quite a few different types.

 
Am I really having nothing but problems and it is all just me or you are somehow misinterpreting what I posted? Let’s try to quickly recap and see:
 
NT-H2 I had nothing but good words about. Performed exactly like number of reviews and owners indicated it would.
 
PK-3 I didn’t have bad words about either, I said it performed almost as good as NT-H2 but that I preferred Noctua, which is exactly in line with what number of reviews and owner posts said about it.
 
Kryonaut, regardless of did I correctly diagnosed / use correct term for what I saw or not my bottom line experience with it reflects exactly what number of others reported too.
 
MX-5 (whose behavior, I’m guessing, should be at least same of an MX-4, if not better) behaved for me no differently than what others reported about MX-4.
 
And GC-Extreme too performed same for me as number of others reported it did for them.
 
So is it really just me? IMHO when one’s experiences reflect exactly what others have too reported then that indicates it is definitely not just him. Actually it indicates his experience is in line what others too can expect.
 

HeavyHemi
Anyhow, you're not hearing what others are posting to you but I guess you'll just have to learn the hard way. I'm not fond of wasting my time.

 
I too am not fond at all of my time and money being wasted, nor of learning the hard way. But when one finds praising reviews/”reviews” often don’t give complete picture then his time and money are being wasted, he ended up forced to learn the hard way.
 
My experience with GC Extreme is exactly what Luumi reported in one of his reviews. He later dismissed it as bad batch. But his initial report was at end of September of 2020, dismissal was at beginning of November of 2020 (after manufacturer sent him new batch), I got my tube from official reseller beginning of March of 2021, and my experience reflects exactly what Luumi and number of others reported (can be found when one starts to search).
 
Are you saying Luumi didn’t know what he was doing when he did initial test? Are you saying all of us are having that experience because they are not hearing what you posted? Or it is other way around, you are not hearing what is being said and in turn wasting time?
 
… and you know, with all due respect to Luumi, when you get fresh batch from manufacturer that sent it to you because they became aware of problem you reported, and now everything seems OK on first day you applied it that doesn’t mean you can declare all is fine now cause it doesn’t guarantee rest of us that don’t receive personal attention from manufacturer will have same experience. Nor it guarantees Luumi will still have good experience year down the road on daily use system.

Don't some pastes have a curing time? As in after application and heat cool cycles, they will work a bit better.  Some don't have any cure time. 

notfordman
Don't some pastes have a curing time? As in after application and heat cool cycles, they will work a bit better.  Some don't have any cure time. 

Eh, that's a marketing term.
 
That is addressed in the video I embedded on the previous page.
 
The idea behind "cure time" is that thermal paste heats up, thins out, squeezes out over time, and the two surfaces get better contact and better heat transfer as a result.  I think that they just chose "cure time" as a marketing term to sound fancy rather than trying to explain heat, pressure, squeeze-out, viscosity, and thermal material thickness in paragraph form.
 
It's silicon-based with metal powders mixed inside.  It doesn't "cure" in a traditional sense.  Thicker thermal pastes can improve in thermal performance over time though, yes.  But after enough time, they all start to then lose thermal performance due to pump-out effect.  "Cure time" isn't infinite; it's on a curve.  Different viscosities will take different amounts of time to squeeze out and have optimal contact.  The thinner high-performance stuff will have a low "cure time" (maybe not even advertised because it is so insignificant) and the thicker long-lasting stuff will have a longer "cure time".  The thicker stuff will take longer to perform at its peak, will not ever perform as good as the thin high-performance stuff, but will perform at that mediocre level (within a couple degrees, so really what does it matter, and why even call it "mediocre"?), while the thinner high-performance stuff performs better for a shorter duration of time before pump-out effect takes hold.
 
Most of the traditional pastes are pretty much the same.  ZoranC has confirmed that with his endless testing and endless disappointment.
 
Liquid metal is where it is at.

ty_ger07

notfordman
Don't some pastes have a curing time? As in after application and heat cool cycles, they will work a bit better.  Some don't have any cure time. 

Eh, that's a marketing term.
 
That is addressed in the video I embedded on the previous page.
 
The idea behind "cure time" is that thermal paste heats up, thins out, squeezes out over time, and the two surfaces get better contact and better heat transfer as a result.  I think that they just chose "cure time" as a marketing term to sound fancy rather than trying to explain heat, pressure, squeeze-out, viscosity, and thermal material thickness in paragraph form.
 
It's silicon-based with metal powders mixed inside.  It doesn't "cure" in a traditional sense.  Thicker thermal pastes can improve in thermal performance over time though, yes.  But after enough time, they all start to then lose thermal performance due to pump-out effect.  "Cure time" isn't infinite; it's on a curve.  Different viscosities will take different amounts of time to squeeze out and have optimal contact.  The thinner high-performance stuff will have a low "cure time" (maybe not even advertised because it is so insignificant) and the thicker long-lasting stuff will have a longer "cure time".  The thicker stuff will take longer to perform at its peak, will not ever perform as good as the thin high-performance stuff, but will perform at that mediocre level (within a couple degrees, so really what does it matter, and why even call it "mediocre"?), while the thinner high-performance stuff performs better for a shorter duration of time before pump-out effect takes hold.
 
Most of the traditional pastes are pretty much the same.  ZoranC has confirmed that with his endless testing and endless disappointment.
 
Liquid metal is where it is at.

Thank you for the clarity ty_ger07, you are full of...Information!