I suppose that it depends on your application. If you're running games then these days the CPU really isn't stretching itself and on top of that the TDPs for such CPUs just aren't in the realm they used to be. That being said, there are some use cases that do really tax a CPU and throw off some heat and you can well imagine that if your mobo Tau settings are appropriate for sustained boost that your peak sustained performance could hinge on a couple of degrees. This being said, it's gamers that tend to be targetted by the brands and the subject of "gotta have the best" advertising. They're probably more impulse buy prone and science blinded (as well as RGB shiny shiny bling bling blinded - I mean, GOLD LIGHT UP RAM?!) than your real high end workstation / server user who buys with their head. I'd be interested to hear from such people as to whether they use high quality paste or just use the basic stuff and apply it right...
I hate RGB, but I do water cool and overclock. I do use high end pastes and TIMs, with the exception of liquid metal. I used to use it, but no longer. That's another story.
I don't do extreme OCing as in liquid nitrogen. I do enjoy dialing in the clock speeds to see how far I can go. Better pastes are worth it to me. They do generally make a few degrees difference.
It's much like motor oil. There are basic brands that are fine and will protect your engine. They will also be good enough to keep your warranty in effect. However, there are other oils that do a better job lubricating, keeping the engine clean and prolonging the life of seals. It's the same with better TIM pastes and greases.
Certainly not worth the money. Lower TDPs also lowers the potential of more expensive cooling products. If you are an enthusiast for higher numbers then go for it.
This depends as well. With lower TDPs, we also have a plethora of smaller form factors and with that often comes smaller and less performant cooling solutions. Higher quality TIM may still be important in this case. There is also the philosophy that even with a monstrous cooler you don't want your cooler held back by using a peanut butter class TIM.
I personally see a point at which the increase in cost of the TIM doesn't match the performance difference and recommend that my clients stick with a good quality TIM (not particularly hard to find) at a more reasonable price (this is lower than many think). That said, overall cost of the cooling solution plays a role here. As an extreme case, if you are going to spend $500 or more into a liquid cooling system, why would you risk using a lesser quality TIM to save a few dollars.
Mh. Not so sure about that. The Ryzen 5 3600 produces quite a bit of heat which is rather difficult to remove from there. Seems like the shape of the heatspreader doesn't help, which means you need a good paste to make contact. And my Noctua D15 is running COLD. Like, room temperature, while the CPU underneath is at 90°C. Because it will just keep the warmth to itself. We are talking about a CPU pushing at most about 100 W TDP into a cooler capable of much more, and yet it gets too hot.
Given how sensitive the Ryzen 3000 series is to warmth... as in they need to be cold in order to be fast... yeah, that is an issue. Anything that will help me transfer the heat to the cooler is very much appreciated.
Well.. considering I've been using Kryonaught for years, the benchmark they used says this new formulation is 1C cooler. To me, that's not worth the effort of switching thermal pastes. Maybe for the next guy?
"up to 2°c lower" = "up to" is the key phrase here.
Presumably, Noctua's coolers will still bundle the cheaper NT-H1? Surprised they're keeping this older version around; Noctua's never been one to shy from premium pricing.
Curious how this performs against the Thermal Grizzly Kryonaut.
I tried Kryonaut and NT-H1 and strongly prefer the latter for its ease of application. The consistency of the Noctua paste was so much easier to work with than the Kryonaut which was almost chewing-gum like.
I suspect a perfect application of Kryonaut would perform slightly better but seems much harder to achieve, so I strongly prefer the Noctua.
In the video they added, it says it's 1C cooler. In my mind, not worth upgrading to NT-H2. Also all the other pastes I've seen that cool better than Kryonaut tend to be conductive, liable to short circuit electronics..
Looks OK. Reasonable product for Noctua's loyal customer base, especially if they don't want to dabble with liquid metal products.
I still personally like using the liquid metal tube I got from Thermal Grizzly like 2 years ago in most places, even GPU reapplications. I never use aluminum heatsinks (ex: intel box cooler), and it performs well in copper or nickel plated copper surfaces, and a single tube lasts for quite a few applications since you use so little and spread it thin. No issues with corrosion here, just the usual light surface marring/staining which doesn't affect performance. I also protect the surrounding area with a good electrical tape, too. Liquid metal doesn't seem to harden or lose thermal properties over time, and still has its "wet" texture after dismounting the heatsink, and the better thermal transfer means better performance/lower fan rpm, depending on how you tune your settings after the fact.
You’d better off using none of those. I have the Arctic MX-4 (replaced the 2-years old Arctic Silver 5) in one of my rig and it’s been 7 years since. No change of idle temp or load temp. This old i5-2500k OCed to 4.5Ghz has been serving me well.
I like MX-4 as a good middle ground sort of thermal compound. It does what sonny73n is saying by lasting many years. It also has the right viscosity balance that makes application easy, but limits movement after application. It isn't electrically conductive so a bit of compound applied someplace where it shouldn't isn't going to cause a problem and it cleans up easily. The price per quantity is very good too and balances out its lower than liquid metal TIM performance. However, given that the person in question already has Noctua paste on-hand, it makes sense to just use what is available and not worry about it.
> Noctua hasn't highlighted which materials have been used.
Shouldn't there be a Material Safety Data Sheet, or some EU-equivalent?
Did they say anything about the suitability of their wipes for 3rd party compounds?
I've typically used rubbing alcohol and paper towels, for coarse removal. Then, I use a 2-stage cleaning/prep system from Arctic Silver (probably just mineral spirits and rubbing alcohol) + lint-free cotton cloth.
5-years is the minimum I would consider. I have 3 PCs I'm still using beyond that point (though I'm getting ready to replace one). One big reason I've stuck with Arctic Cooling's MX-4 is its longevity.
We’ve updated our terms. By continuing to use the site and/or by logging into your account, you agree to the Site’s updated Terms of Use and Privacy Policy.
Facebook
Twitter
RSS
34 Comments
Back to Article
Marlin1975 - Tuesday, January 29, 2019 - link
How much does thermal paste play into temps today? I do not see them talked about as much versus much older chips/designs.philehidiot - Tuesday, January 29, 2019 - link
I suppose that it depends on your application. If you're running games then these days the CPU really isn't stretching itself and on top of that the TDPs for such CPUs just aren't in the realm they used to be. That being said, there are some use cases that do really tax a CPU and throw off some heat and you can well imagine that if your mobo Tau settings are appropriate for sustained boost that your peak sustained performance could hinge on a couple of degrees. This being said, it's gamers that tend to be targetted by the brands and the subject of "gotta have the best" advertising. They're probably more impulse buy prone and science blinded (as well as RGB shiny shiny bling bling blinded - I mean, GOLD LIGHT UP RAM?!) than your real high end workstation / server user who buys with their head. I'd be interested to hear from such people as to whether they use high quality paste or just use the basic stuff and apply it right...Questor - Wednesday, January 30, 2019 - link
I hate RGB, but I do water cool and overclock. I do use high end pastes and TIMs, with the exception of liquid metal. I used to use it, but no longer. That's another story.I don't do extreme OCing as in liquid nitrogen. I do enjoy dialing in the clock speeds to see how far I can go. Better pastes are worth it to me. They do generally make a few degrees difference.
It's much like motor oil. There are basic brands that are fine and will protect your engine. They will also be good enough to keep your warranty in effect. However, there are other oils that do a better job lubricating, keeping the engine clean and prolonging the life of seals. It's the same with better TIM pastes and greases.
dromoxen - Thursday, January 31, 2019 - link
and the Penguins , toopiroroadkill - Wednesday, January 30, 2019 - link
Yeah, it's still as important as it once was, why wouldn't it be?zodiacfml - Wednesday, January 30, 2019 - link
Certainly not worth the money. Lower TDPs also lowers the potential of more expensive cooling products. If you are an enthusiast for higher numbers then go for it.BurntMyBacon - Friday, February 1, 2019 - link
This depends as well. With lower TDPs, we also have a plethora of smaller form factors and with that often comes smaller and less performant cooling solutions. Higher quality TIM may still be important in this case. There is also the philosophy that even with a monstrous cooler you don't want your cooler held back by using a peanut butter class TIM.I personally see a point at which the increase in cost of the TIM doesn't match the performance difference and recommend that my clients stick with a good quality TIM (not particularly hard to find) at a more reasonable price (this is lower than many think). That said, overall cost of the cooling solution plays a role here. As an extreme case, if you are going to spend $500 or more into a liquid cooling system, why would you risk using a lesser quality TIM to save a few dollars.
kadajawi - Friday, August 30, 2019 - link
Mh. Not so sure about that. The Ryzen 5 3600 produces quite a bit of heat which is rather difficult to remove from there. Seems like the shape of the heatspreader doesn't help, which means you need a good paste to make contact. And my Noctua D15 is running COLD. Like, room temperature, while the CPU underneath is at 90°C. Because it will just keep the warmth to itself. We are talking about a CPU pushing at most about 100 W TDP into a cooler capable of much more, and yet it gets too hot.Given how sensitive the Ryzen 3000 series is to warmth... as in they need to be cold in order to be fast... yeah, that is an issue. Anything that will help me transfer the heat to the cooler is very much appreciated.
FastCarsLike - Sunday, April 26, 2020 - link
Well.. considering I've been using Kryonaught for years, the benchmark they used says this new formulation is 1C cooler. To me, that's not worth the effort of switching thermal pastes. Maybe for the next guy?jameserlay - Tuesday, January 29, 2019 - link
"The new Noctua NT-H1 thermal compound builds upon the previous success of the NT-H1 material..."What?
quorm - Tuesday, January 29, 2019 - link
It's important to add an additional layer of NT-H1 so your processor doesn't get cold during the winter.dakishimesan - Tuesday, January 29, 2019 - link
lolSamus - Wednesday, January 30, 2019 - link
lmaogavbon - Tuesday, January 29, 2019 - link
Thanks for pointing that typo out - Editedikjadoon - Tuesday, January 29, 2019 - link
"up to 2°c lower" = "up to" is the key phrase here.Presumably, Noctua's coolers will still bundle the cheaper NT-H1? Surprised they're keeping this older version around; Noctua's never been one to shy from premium pricing.
Curious how this performs against the Thermal Grizzly Kryonaut.
ydeer - Tuesday, January 29, 2019 - link
I tried Kryonaut and NT-H1 and strongly prefer the latter for its ease of application.The consistency of the Noctua paste was so much easier to work with than the Kryonaut which was almost chewing-gum like.
I suspect a perfect application of Kryonaut would perform slightly better but seems much harder to achieve, so I strongly prefer the Noctua.
ikjadoon - Wednesday, January 30, 2019 - link
Oh, that's a great point. Huh, Tom's Hardware said NT-H1 was more viscous and harder to apply than TG Kryonaut? But perhaps they took shortcuts.https://www.tomshardware.com/reviews/thermal-paste...
I know TG now includes a flat applicator, but perhaps it's still too thick.
FastCarsLike - Sunday, April 26, 2020 - link
In the video they added, it says it's 1C cooler. In my mind, not worth upgrading to NT-H2. Also all the other pastes I've seen that cool better than Kryonaut tend to be conductive, liable to short circuit electronics..JoeyJoJo123 - Tuesday, January 29, 2019 - link
Looks OK. Reasonable product for Noctua's loyal customer base, especially if they don't want to dabble with liquid metal products.I still personally like using the liquid metal tube I got from Thermal Grizzly like 2 years ago in most places, even GPU reapplications. I never use aluminum heatsinks (ex: intel box cooler), and it performs well in copper or nickel plated copper surfaces, and a single tube lasts for quite a few applications since you use so little and spread it thin. No issues with corrosion here, just the usual light surface marring/staining which doesn't affect performance. I also protect the surrounding area with a good electrical tape, too. Liquid metal doesn't seem to harden or lose thermal properties over time, and still has its "wet" texture after dismounting the heatsink, and the better thermal transfer means better performance/lower fan rpm, depending on how you tune your settings after the fact.
modeonoff - Tuesday, January 29, 2019 - link
I just received an i9-9900K to be used with a Noctua NH-D15S. Is it better to use this new paste or just use the one that came with the NH-D15S?sonny73n - Wednesday, January 30, 2019 - link
You’d better off using none of those. I have the Arctic MX-4 (replaced the 2-years old Arctic Silver 5) in one of my rig and it’s been 7 years since. No change of idle temp or load temp. This old i5-2500k OCed to 4.5Ghz has been serving me well.Diji1 - Wednesday, January 30, 2019 - link
No he'd be better off using the Noctua products since Artic products have not been high performers since forever.PeachNCream - Friday, February 1, 2019 - link
I like MX-4 as a good middle ground sort of thermal compound. It does what sonny73n is saying by lasting many years. It also has the right viscosity balance that makes application easy, but limits movement after application. It isn't electrically conductive so a bit of compound applied someplace where it shouldn't isn't going to cause a problem and it cleans up easily. The price per quantity is very good too and balances out its lower than liquid metal TIM performance. However, given that the person in question already has Noctua paste on-hand, it makes sense to just use what is available and not worry about it.mode_13h - Wednesday, January 30, 2019 - link
> Noctua hasn't highlighted which materials have been used.Shouldn't there be a Material Safety Data Sheet, or some EU-equivalent?
Did they say anything about the suitability of their wipes for 3rd party compounds?
I've typically used rubbing alcohol and paper towels, for coarse removal. Then, I use a 2-stage cleaning/prep system from Arctic Silver (probably just mineral spirits and rubbing alcohol) + lint-free cotton cloth.
5-years is the minimum I would consider. I have 3 PCs I'm still using beyond that point (though I'm getting ready to replace one). One big reason I've stuck with Arctic Cooling's MX-4 is its longevity.
Hxx - Wednesday, February 6, 2019 - link
1% improvement for 100% price hike. awesome