Question For i9-9900K, does it matter whether I use Noctua NT-H1 or NT-H2?

[anandtechreader]

Hello, I think there is a NT-H1 thermal paste that came with the Noctua NH-D15S. For i9-9900K, is it better to use the new NT-H2? It looks like the improvement is not that much for this CPU platform. Is it risky to use a completely new product that has not been tested by real customers and reviewers?

https://noctua.at/en/performance-comparison-nt-h1-vs-nt-h2?___from_store=de

 

[anandtechreader]

I think mine looks more like the cap.

 

[DrMrLordX]

@anandtechreader

those spills are from using too much paste. If you had good temps then I suggest you continue using the Kryonaut. Just use less of it.

 

[anandtechreader]

@anandtechreader

those spills are from using too much paste. If you had good temps then I suggest you continue using the Kryonaut. Just use less of it.

Thanks. Shall I continue to use the full spread method but cover only, say 75-80% of the area and leave the edges auto-filled by the pressure from the heatsink? Alternatively, shall I just apply a small dot or a X-pattern in the center?

 

[DrMrLordX]

Honestly I would go with the dot, though if you're having problems with full spread and too much thickness, then doing 75% of the heat spreader centered in the middle would be a good alternative. Just be careful to avoid introducing air bubbles.

 

[anandtechreader]

I asked "When applying it on the Intel i9-9900K CPU and Noctua NH-D15S heatsink, should I just apply a small dot, a X-shape pattern or a small dot and then spread the paste myself using a plastic spreader?"

This is what they recommended:

"If you use a 9900k you can use all 3 Methods, i personally prefer the X Shape or spreading the paste. While spreading the pase you have to use a bit of pressure and spread it slowly from one side to the other side, with one run lasting about 5 seconds ."

When asked to explain what the "5 seconds" mean, they replied:

"we have currently no videos or photos available, we are working on that. If you check "der8auer" on Youtube you can see him applying our paste.
By 5 seconds i mean when you are using the spatula to distribute the paste on the cpu it should take around 5 seconds from one side to another on the cpu.
You do not have to use, but we recommend cleaning before."


Why 5 seconds from one side to another?

 

[UsandThem]

Just remember, you are only filling microscopic holes. You don't need much (and a whole lot less than you used last time).

Small dab in the middle. Then tighten each side of the cooler down, alternating sides, so the paste spreads out evenly. I usually do 3 turns of the included screwdriver tool Noctua gives you before switching to the other side. Do it this until both sides are tightened.

If you still manage to make a mess again like you did in your picture, maybe buy something like this instead: https://www.amazon.com/Innovation-Cooling-Graphite-Thermal-Pad/dp/B07CKVW18G

 

[anandtechreader]

Just remember, you are only filling microscopic holes. You don't need much (and a whole lot less than you used last time).

Small dab in the middle. Then tighten each side of the cooler down, alternating sides, so the paste spreads out evenly. I usually do 3 turns of the included screwdriver tool Noctua gives you before switching to the other side. Do it this until both sides are tightened.

If you still manage to make a mess again like you did in your picture, maybe buy something like this instead: https://www.amazon.com/Innovation-Cooling-Graphite-Thermal-Pad/dp/B07CKVW18G

Thanks.

Did not notice about the screwdriver tool in the Noctua kit. I just used a screwdriver. Any advantage in using the Noctua screwdriver vs. conventional screwdriver?

Is the thermal pad better than thermal paste?

 

[UsandThem]

Thanks.

Did not notice about the screwdriver tool in the Noctua kit. I just used a screwdriver. Any advantage in using the Noctua screwdriver vs. conventional screwdriver?

It's just a differently shaped screwdriver is all. If a regular one works for you, that's fine.

Is the thermal pad better than thermal paste?

Roughly the same based on the reviews I've seen. I've always used paste because it's cheap (or comes with a retail cooler), but I've also never made a mess quite like you did.

If you use too much paste, then your temps would run a little warmer. (and / or make a mess all around your socket).

 

[UsandThem]

I also checked the Noctua cooler you bought, and it should have come with the screwdriver as well.

https://www.guru3d.com/articles-pages/noctua-nh-d15s-cpu-cooler-review,3.html

 

[DrMrLordX]

Roughly the same based on the reviews I've seen.

Interesting . . . AMD used a similar product on Radeon VII's GPU.

 

[anandtechreader]

It's just a differently shaped screwdriver is all. If a regular one works for you, that's fine.



Roughly the same based on the reviews I've seen. I've always used paste because it's cheap (or comes with a retail cooler), but I've also never made a mess quite like you did.

If you use too much paste, then your temps would run a little warmer. (and / or make a mess all around your socket).

So the thermal pad performs similarly than the paste, don't make a mess but just a bit expensive? How are these ranked from the best?

1. Paste that comes with Noctua NH-D15S heatsink
2. Thermal pad and
3. Kryonaut

 

[anandtechreader]

I watched "der8auer" on Youtube just before I applied the Kryonaut. From that video, it seems that he also applied quite a lot like I did.

For the Innovation Cooling Graphite Thermal Pad user reviews, it seems that if I am not careful due to the slippery of the thermal pad, it could damage the components nearby since it is connective. How likely does this happen? Is it worth the risk compared with using Kryonaut or the stock Noctua NH-D15S thermal paste?

 

[UsandThem]

So the thermal pad performs similarly than the paste, don't make a mess but just a bit expensive? How are these ranked from the best?

You really could rank them this way as they are all so close, and all will give you very similar results:

1A. Paste that comes with Noctua NH-D15S heatsink
1B. Thermal pad
1. Kryonaut

 

[UsandThem]

I watched "der8auer" on Youtube just before I applied the Kryonaut. From that video, it seems that he also applied quite a lot like I did.

For the Innovation Cooling Graphite Thermal Pad user reviews, it seems that if I am not careful due to the slippery of the thermal pad, it could damage the components nearby since it is connective. How likely does this happen? Is it worth the risk compared with using Kryonaut or the stock Noctua NH-D15S thermal paste?

It all comes down to how careful you are, and if you can use [A LOT LESS] thermal paste without making a mess again. You have to be careful building a PC, so I don't think carefully laying down a thermal pad without it touching your socket area or motherboard should be that difficult.

 

[YouriP]

I got Noctua NH-D15S and I9-9900k. The included NT-H1 didn't work for me well. I re-applied it 3 times and every time got temps around 90c in Handbrake with stock clock. Yesterday I decided to apply MX-4 I always used in my previous PC builds. Temp went down for 7c. Great! I have 83c with full AVX load now.

 

[UsandThem]

I got Noctua NH-D15S and I9-9900k. The included NT-H1 didn't work for me well. I re-applied it 3 times and every time got temps around 90c in Handbrake with stock clock. Yesterday I decided to apply MX-4 I always used in my previous PC builds. Temp went down for 7c. Great! I have 83c with full AVX load now.

You might have applied too much.

Your results do not match the results in thermal paste tests.

https://www.guru3d.com/articles_pages/guru3d_thermal_paste_roundup_2019,10.html

https://www.tomshardware.com/reviews/thermal-paste-comparison,5108-9.html

 

[YouriP]

You might have applied too much.

Your results do not match the results in thermal paste tests.

https://www.guru3d.com/articles_pages/guru3d_thermal_paste_roundup_2019,10.html

https://www.tomshardware.com/reviews/thermal-paste-comparison,5108-9.html

I tested with a stock clock 4.7 not with 5.0

 

[UsandThem]

I tested with a stock clock 4.7 not with 5.0

You still would have not gotten a 7c difference between Noctua's paste and MX-4.