My thermal paste collection...how would you rank them from best to worst?

[Vario]

Really? This has happened twice. Granted, both times are from the same tube. Maybe I have a bad sample?


Ok, so I'm not alone. Perhaps a bad batch?

If it can happen to Arctic, it can happen to Noctua.

Happened to me with a number of mounts, I ended up giving away all the tubes I had of it with other pc hardware I sold on eBay and forums.

I tried a few different brands went back to AS5, the old reliable. I have tubes that are 15 years old that still work fine.

 

[dirtyferret]

That's weird, I've used a lot of NT-H1 and never seen that, even on installs lasting several years.

really? This has happened twice. Granted, both times are from the same tube. Maybe I have a bad sample?

Several reviewers use it as their got to and I probably use it more often that other pastes but I also have several tubes of it. Over the years of use I would simply say it does the job without any issue but I often tell people just use the paste that comes with the cooler. If you need to ask what type of thermal paste/compound/etc. you should get, you are not doing the type of OC that is going to need every last bit of temp for stability.

As I stated earlier on, there are only fractional differences. Thermal paste industry is partly a scam industry

not so much a "scam" because it does work and you obviously should use it but similar to many other parts of the PC industry (and others), there is so much marketing BS in it for positioning and differentiation because the products are in fact so similar.

 

[Shrek]

GD900 for me i dont waste my money virjazzles .

Another vote for GD900

 

[lexluthermiester]

The MX-5 is garbage, in my opinion. It's as thick and sticky as chewing gum. Performance-wise it's OK, but applying it is a real pain in the backside that I wouldn't even wish on my enemies.

My experiences have been all positive, I bought boxes of the stuff for my shop and no one has had the experiences you and others have described. However, I have no desire to argue about the subject, so let's agree to disagree.

 

[Bones]

There's been a number of them.

Yes, in reference to the question as asked earlier a prime example was this: Thermal Grizzly Kryonaut scratching and denting. | TechPowerUp Forums

In truth nearly every maker has had issues before.

In terms of cost vs what you get MX-4 and NT-H1 are about the best values for TIM I'm aware of, each performs well and for desktop use are perfectly fine without paying a premium for what you want it for.
I mean why pay 2-3x the cost per gram for something that's not going to deliver the same amount of performance improvement over the other if you don't have to?

Incorrect. It's electrically capacitive. There's a HUGE difference. It's only a problem if some of it get on and bridges contact points. But even then, it's easily cleaned and resolved.

My personal take:
Doesn't really matter if it's conductive or capacitive, if it can cause a problem that way I woudn't want to use it and frankly I don't in the case of original AS5 TIM.

If it's easily cleaned or whatever is an non-point, you can't just wipe away the aftermath of an issue with a papertowel or whatever else it once it occurs.
That would mean real hardware damage occured and that can only be resolved by repair/replacement of the affected component and nothing else, sometimes the cost of a papertowel or cloth pales in comparision to the cost of fixing things related to it.

My point is if you suspect there may be a problem, it's best not to take the chance of Mr. Murphy paying you a visit because he will if you openly invite him to..... He's bad enough about barging in on his own as is.

 

[ThrashZone]

Hi,
Yeah my mx-5 was/ is pretty wet and as wet as mx-4 2019 is so someone saying mx-5 is thick as bubble gum just shows why it's eol, very inconsistent mixture.

 

[Bones]

Hi,
Yeah my mx-5 was/ is pretty wet and as wet as mx-4 2019 is so thick as bubble gum just shows why it's eol, very inconstant mixture.

I've had a couple of tubes of that before and each was good.
They probrably sent you a tube that was old from sitting on a shelf somewhere. My experience with it has been all good, it's too bad it's EOL now because that stuff let me do some crazy stuff before and you guys know what I do sometimes.

 

[ThrashZone]

I've had a couple of tubes of that before and each was good.
They probrably sent you a tube that was old from sitting on a shelf somewhere. My experience with it has been all good, it's too bad it's EOL now because that stuff let me do some crazy stuff before and you guys know what I do sometimes.

Hi,
I wasn't saying mx-4 or 5 was thick as bubble gum some else did
I was saying both were wet.

 

[Bones]

Hi,
I wasn't saying mx-4 or 5 was thick as bubble gum some else did
I was saying both were wet.

Gotcha on that.
I prefer TIM's that are a bit "Wet" myself, just not too wet (Runny) - That simply will not do.
What is too wet for one may not be for another is all I can say about that.

 

[bonehead123]

I have used all of those the OP mentioned and others too, and under normal circumstances, with sufficient care/patience during application, I have rarely ever seen more than 2-4° difference between them in average, everyday builds, and yes, that includes most typical gammin rigs too

However, this doesn't count the so-called "exotic/high-end" stuff that I have treid here & there, which is usually severely overpriced, over-hyped and adds very little to no REAL benefits over the everyday stuff IMHO....

Also, I would NOT recommend using any paste that is moar than 2-3 years old, just to be safe with those shiny new $500+ CPU's or $1K GPU's

 

[cvaldes]

Look at this recent thermal paste roundup from Tom's Hardware:

Best Thermal Paste for CPUs 2025: Over 50 Compounds Tested and Ranked

The right thermal paste will improve performance.

www.tomshardware.com

paying special attention to the legacy list of 85 pastes.

The most noteworthy observation is that the conventional pastes (not liquid metal, not pads) are almost all in a very narrow 3 °C performance window. They all perform similarly if they are correctly and consistently applied and the tester knows the proper way of securing a heatsink to a chip's IHS something that the Tom's Hardware testers presumably can do.

I've used maybe 6-7 thermal pastes in the past few years and they all thermally perform similarly. This includes Noctua NT-H1, NT-H2, Arctic MX-4, MX-5, a small tube of Alphacool branded paste that came with a waterblock, and one or two others that I have since forgotten (probably because they behaved the same as the others). The main difference is paste consistency, handling, and ease of application. Oh, and price.

There are several TPU forum participants who run PC repair businesses. Their opinions and experiences are different than those of Joe Consumer or the occasional hobbyist because the former apply thermal paste frequently, just like a pastry chef frosts cakes more frequently than the average home cook.

All thermal pastes will eventually dry out; they aren't expected to stay liquid forever. They're shipped in a paste format to facilitate application. The main point of a thermal paste is to provide a thin layer of thermally conductive substance (ideally a couple of molecules thick) to fill out any air pockets that might be created between the microscopic pits and bumps in the two metal contact surfaces. The paste is just a fluid to transport the thermal compound molecules just like painting media is there to transport pigment. Remember that thermal paste components never conduct heat as well as the two metal surfaces they are bonding so using more thermal paste isn't better.

The main takeaway is to find a reasonably priced thermal paste that you find easy to apply and can do so consistently without 5-10 attempts to get it right because you may not do it again for months, even years. I don't need a one-gallon bucket of this stuff, I'm not running my own PC repair business in the same way that I don't need 50-pound bags of all-purpose flour.

@A Computer Guy
Identify the one you find easiest to apply and use that. Or use them all. From a thermal performance standpoint, it's not going to make a significant difference. Your skill at applying a thin and consistent layer of paste and properly tightening your heatsink is far more important than the brand printed on the paste's tube/packet.

 

[Shrek]

All thermal pastes will eventually dry out; they aren't expected to stay liquid forever.

Silicone oil is good to 500°F, so one might not expect much drying out at say 200°F

 

[Kapone33]

NTH1 was pretty good Arctic dries in months and the XT50 is much less viscous.

 

[Shrek]

Drying may actually not be an issue, a solid is less susceptible to pump out.

 

[caroline!]

AS-5 - good stuff at a reasonable price
Corsair - big syringe, I like that
Noctua - seen better tbh, like MX-4
TT - meh, I had a few mini syringes that came with coolers and it's oily af, looks bad
TG - ridiculous prices for normal quality
EK - at the bottom because I despise the brand in general tbh, should be casted into oblivion forever with NZXT and Razer.

You asked how I'd rank them and that's how. As for the paste I've always used AS-5 and MX-4 and had zero issues with them. For low end/midrange builds with locked chips and stock coolers I use white mayo and just works. As several users have said, this new "gaming" paste market is a scam.

 

[silentbogo]

Ok, so I'm not alone. Perhaps a bad batch?

Or it could be related to those "freebie" syringes? Not as good as retail? I assume yours came from a Noctua-branded heatsink as well.
NT-H2 seems to behave much better, though.
But bad batch is not out of the question. My L9i was from '15-'16 or so(bought it in '16 but it was slightly used). Also had some crappy experience with later MX4, where I could've just written it off on a "knock-off", if I didn't actually know where it came from. It was all runny, and if your heatsink alignment was at least a little bit off, all the "silver" nano-particles would clump-up to the lower-pressure side, while the higher-pressure side would be covered by a completely transparent silicone snot. Worked fine on CPUs with heatspreader, but was a total nightmare on laptops and GPUs. That's when I switched to MX2 - it's a little more viscous.

 

[A Computer Guy]

from top to bottom in the picture

• Thermal Grizzly Something (came with my water block)
• XTM50 (about 6 months old)
• Noctua NT-H1 (about 1 yrs old)
• EK Ectotherm (about 2 yrs old)
• TG-7 (about 2 yrs old)
• Artic Silver (about 3 yrs old)
• Artic Sliver (about 5 yrs old)

still waiting for delivery

• Thermal Grizzly Kryonaut

​

I haven't used many pastes but in my small collection my top 3 are

1. Noctua NT-H1
2. EK Ectotherm
3. Artic Silver

Artic Silver was my standard go to for many years. EK Ectotherm and Noctua NT-H1 seemed to hold up pretty well after my 1-year teardown and inspection for the past two years. The NT-H1 seemed a bit drier after 1 year so I might worry about it a bit for longer term applications like 2-3 years.

For my rebuild I was gong to try Thermal Grizzly Kryonaut next for my final build application. Perhaps just the one shown in the picture that came with my CPU block. Looks like that's good enough for 2 maybe 3 applications. Thank you everyone for your feedback.

 

[ThrashZone]

I haven't used many pastes but in my small collection my top 3 are

1. Noctua NT-H1
2. EK Ectotherm
3. Artic Silver

Artic Silver was my standard go to for many years. EK Ectotherm and Noctua NT-H1 seemed to hold up pretty well after my 1-year teardown and inspection for the past two years. The NT-H1 seemed a bit drier after 1 year so I might worry about it a bit for longer term applications like 2-3 years.

For my rebuild I was gong to try Thermal Grizzly Kryonaut next for my final build application. Perhaps just the one shown in the picture that came with my CPU block. Looks like that's good enough for 2 maybe 3 applications. Thank you everyone for your feedback.

Hi,
Yeah nt-h1 is #1 in my book
2nd would have to be TG hydronaut though not ek.

 

[lexluthermiester]

Doesn't really matter if it's conductive or capacitive, if it can cause a problem that way I woudn't want to use it and frankly I don't in the case of original AS5 TIM.

My point is if you suspect there may be a problem, it's best not to take the chance of Mr. Murphy paying you a visit because he will if you openly invite him to..... He's bad enough about barging in on his own as is.

Here's the reality, Arctic Silver 5 is very easy to work with. As such, one would have to be VERY careless or incompetent to actually get enough AS5 in places it doesn't belong in enough amounts to be a problem. Anyone who has to worry about such should NEVER be allowed to work on a PC.

 

[A Computer Guy]

Here's the reality, Arctic Silver 5 is very easy to work with. As such, one would have to be VERY careless or incompetent to actually get enough AS5 in places it doesn't belong in enough amounts to be a problem. Anyone who has to worry about such should NEVER be allowed to work on a PC.

Is there a way to change the font of the reader. My first take on this was a bit off because S and 5 look very similar and my eyes are getting old..

 

[Bones]

Here's the reality, Arctic Silver 5 is very easy to work with. As such, one would have to be VERY careless or incompetent to actually get enough AS5 in places it doesn't belong in enough amounts to be a problem. Anyone who has to worry about such should NEVER be allowed to work on a PC.

While I agree with this, you missed an issue about unfamiliarity with the TIM to know how much to use so it doesn't come out and spread out too far (Pump out), going where it doesn't need to.
I've used AS5 before and it can spread a pretty good ways depending on the exact batch you get in the tube.

That's why I say if you're not sure, then don't take the chance.

 

[lexluthermiester]

Or it could be related to those "freebie" syringes? Not as good as retail? I assume yours came from a Noctua-branded heatsink as well.

Nope. I bought mine specifically for testing last year. It was a retail package as shown in the thread in question. Look up the MX-5 thread and it was last year around June.

Is there a way to change the font of the reader. My first take on this was a bit off because S and 5 look very similar and my eyes are getting old..

To clarify, we were talking about Arctic Silver Five ( 5 ) earlier.

 

[Bones]

Is there a way to change the font of the reader. My first take on this was a bit off because S and 5 look very similar and my eyes are getting old..

I feel your pain because my eyes are older too and it does at times look like we're taking about us spreading ASS on a CPU.....
So AS-5 should be a better way to spell it.

 

[lexluthermiester]

While I agree with this, you missed an issue about unfamiliarity with the TIM to know how much to use so it doesn't come out and spread out too far (Pump out), going where it doesn't need to.

1. Pump-out as commonly theorized does not exist for common computing applications.
2. Again, anyone who using so much AS5 that is squeezes out the sides in enough volume to "drip" on other areas of the PC should NOT be allowed to work on a PC.
3. Even when AS5 gets on areas it isn't designed for, it rarely causes problems as it is capacitive, not conductive. It's capacitive effect is greatly offset by it's inherent resistance factor.
4. Arctic Silver 5 is perfectly safe for use by anyone who knows how to spread jam or butter on bread and not get it anywhere but on the bread.
5. Arctic Silver 5 is perfectly safe for use by anyone who knows that you only use what you need.

That's why I say if you're not sure, then don't take the chance.

While that's a fair thing to say, one would really need to muck things up to get it wrong and again, that kind of person should never work on a PC.

 

[Ripcord]

No one has mentioned syy157 really good performance but cheaper than thermal grizzly kryonaut .