thermal paste Which Thermal Paste to buy and apply (Traditional and Liquid Metal)

[Etern4l]

Deepcool G40 huh? Has that been rated in a single reputable review?

[raptorddd]

honeywell paste just arrive now. will try it on later tonight.

am shocked it took 2 weeks to arrive. i had buy on ali express back in 2019 gpu fan took 2 months...

 

w/mk  its just a reference number. in my opinion. or different set up of PC. differenet way of pasting weather ect.

[JeanLegi]

2 minutes ago, raptorddd said:

honeywell paste just arrive now. will try it on later tonight.

am shocked it took 2 weeks to arrive. i had buy on ali express back in 2019 gpu fan took 2 months...

 

w/mk  its just a reference number. in my opinion. or different set up of PC. differenet way of pasting weather ect.

Hopefully positiv shocked :D? 

[Papusan]

10 hours ago, Rofa1234 said:

W/m*K is mainly a marketing number, it really doesn't tell much and nobody knows if it's real or just a fake number. That's why some manufacturer won't even share it. Is there a W/m*K number for Corsair XTM70? I haven't found. Some of my best pastes have a low W/m*K number, for example Deepcool G40 or Akasa T5 ProGrade Plus and some with a high W/m*K are poor performer.

Put it the other way. If they don’t want to post W/m*k numbers they do it for a reason;)

[raptorddd]

not sure if it needs break in. but i did notice -5 celsius

with honeywell

got higher score not sure if its the paste or if it was second run.

 

 

i have only used MX4 so applying honeywell was different not as wet and runny as mx4.

[jaybee83]

18 minutes ago, raptorddd said:

not sure if it needs break in. but i did notice -5 celsius

with honeywell

got higher score not sure if its the paste or if it was second run.

 

 

i have only used MX4 so applying honeywell was different not as wet and runny as mx4.

impressive stuff. one more happy customer with honeywell then...

 

yeah well, W/mK values are about as valuable and informative as nm designations for process nodes, ms response times in display marketing or sequential bandwidth numbers for SSDs 😋

[Rofa1234]

13 hours ago, Papusan said:

Put it the other way. If they don’t want to post W/m*k numbers they do it for a reason;)

 

This is the reason:

 

FAQ: Why doesn’t Noctua specify thermal conductivity or thermal resistance of NT-H1 and NT-H2?

Quote

While there are some standardised methods for testing the thermal conductivity or thermal resistance of thermal compounds, they do not always guarantee comparable results. What's more, many vendors have started to claim thermal resistance or thermal conductivity specifications that appear simply unrealistic in the light of standardised testing. Last but not least, real world performance of thermal compounds is affected by various other factors such as spreading properties, bond line thickness, mounting pressure or contact quality, so even if there was a common and reliable industry standard for testing the thermal conductivity or resistance of thermal pastes, it would not give customers a clear indication regarding the performance they can expect. For these reasons, Noctua has chosen not to publish specifications for thermal conductivity and resistance. Taking part in a game of claiming exaggerated specifications is not an option for us and we neither want to give specifications that, in the context of this practice, could appear misleading.

https://noctua.at/en/why-doesnt-noctua-specify-thermal-conductivity-or-thermal-resistance-of-nt-h1-and-nt-h2

 

 

[Rofa1234]

I cannot test Kooling Monster KOLD-01 because they scammed me, there was no paste inside the package. This is really disappointing because I was curious about this paste  I'm currently testing Arctic MX-6 and Corsair XTM70.

[Etern4l]

2 hours ago, Rofa1234 said:

 

This is the reason:

 

FAQ: Why doesn’t Noctua specify thermal conductivity or thermal resistance of NT-H1 and NT-H2?

https://noctua.at/en/why-doesnt-noctua-specify-thermal-conductivity-or-thermal-resistance-of-nt-h1-and-nt-h2

 

 

 

Sounds like an elaborate excuse to me. Simply state the number, and give precise information regarding the testing conditions. Be a market leader, and other manufacturers will be expected to follow. This would be better than no information. As @Papusan said, withholding the information suggests there is an underlying performance issue. Case in point,  I would not think Noctua pastes are market leading.

 

Ideally though, we would have some independent testing for this. The benefit would be that we'd have a number representing expected thermal performance independent of application scenario and conditions.

[Etern4l]

8 hours ago, raptorddd said:

not sure if it needs break in. but i did notice -5 celsius

with honeywell

got higher score not sure if its the paste or if it was second run.

 

 

i have only used MX4 so applying honeywell was different not as wet and runny as mx4.

 

Sorry, catching up: is this the one?

 

https://www.moddiy.com/products/Honeywell-PTM7950-SP-Super-Highly-Thermally-Conductive-PCM-Paste.html

 

Apparently that's a stock paste used by Lenovo... Hmm.

[Rofa1234]

59 minutes ago, Etern4l said:

 

Sounds like an elaborate excuse to me. Simply state the number, and give precise information regarding the testing conditions. Be a market leader, and other manufacturers will be expected to follow. This would be better than no information. As @Papusan said, withholding the information suggests there is an underlying performance issue. Case in point,  I would not think Noctua pastes are market leading.

 

Ideally though, we would have some independent testing for this. The benefit would be that we'd have a number representing expected thermal performance independent of application scenario and conditions.

 

 

Why is it an excuse, did you even understand what they said? Basically you are saying they could give you a fake W/mK and then you are happy, that's all you need. They can easily specifiy a high thermal conductivity, this is not the problem. You are also claiming a proven paste like Noctua NT-H2 and some other famous pastes suffer from underlying performance issues just because they don't specify a thermal conductivity, it's nonsense. Honeywell PTM7950 paste is specified only with 8.5 W/mK, it should be much worse than all the 13, 16, 17 W/mK pastes right? Sorry it's not that easy.

[Etern4l]

1 hour ago, Rofa1234 said:

 

 

Why is it an excuse, did you even understand what they said? Basically you are saying they could give you a fake W/mK and then you are happy, that's all you need. They can easily specifiy a high thermal conductivity, this is not the problem. You are also claiming a proven paste like Noctua NT-H2 and some other famous pastes suffer from underlying performance issues just because they don't specify a thermal conductivity, it's nonsense. Honeywell PTM7950 paste is specified only with 8.5 W/mK, it should be much worse than all the 13, 16, 17 W/mK pastes right? Sorry it's not that easy.

 

I don't think you have understood what I've been saying. While it's clear that the W/mK numbers provided by manufacturers today are noisy, they are not necessarily useless. The relationship is not that simple and linear though. However, by way of a trivial example, when I see a 5W/mK paste like the dirt cheap "Deepcool G40", on which I get Google hits mostly in India, I have enough data to be deeply skeptical. Not having any W/mK data at all on some pastes if of course worse than having the noisy data.

 

[Rofa1234]

Deepcool G40 works great on my Dell Inspiron 15 7506, it is one of the best I have ever tried. I finished my new test and also included 5 reference pastes including Deepcool G40. MX-6 is a big improvement over MX-4/MX-5, although it's not reaching the very best but it's a solid paste and cheap. I'm sure it will be more competitive on most other devices. Corsair XTM70 is better but cannot reach my best pastes.

 

 

Arctic MX-6 21.5
73/79/75/75

 

Corsair XTM70 21.2
73/77/74/75

 

Deepcool G40 21.3
71/75/72/73

 

Thermalright TF9 21.3
72/76/73/74

 

Alphacool Apex 21.4
73/77/74/75

 

Arctic MX-4 21.5
79/82/81/80

 

Akasa T5 ProGrade Plus 21.5
73/76/74/74

[Etern4l]

13 minutes ago, Rofa1234 said:

Deepcool G40 works great on my Dell Inspiron 15 7506, it is one of the best I have ever tried. I finished my new test and also included 5 reference pastes including Deepcool G40. MX-6 is a big improvement over MX-4/MX-5, although it's not reaching the very best but it's a solid paste and cheap. I'm sure it will be more competitive on most other devices. Corsair XTM70 is better but cannot reach my best pastes.

 

 

Arctic MX-6 21.5
73/79/75/75

 

Corsair XTM70 21.2
73/77/74/75

 

Deepcool G40 21.3
71/75/72/73

 

Thermalright TF9 21.3
72/76/73/74

 

Alphacool Apex 21.4
73/77/74/75

 

Arctic MX-4 21.5
79/82/81/80

 

Akasa T5 ProGrade Plus 21.5
73/76/74/74

 

Thanks for sharing although sadly it would be hard for people in US/Europe to validate this. No reputable reviews, and looks like limited availability outside of India. The only thing to go by are some anecdotal reports/Amazon reviews from Indian customers. Perhaps the product is very new, and/or the company is not very good at accurately measuring W/mK (but kudos to them for posting the 5 result anyway).

[raptorddd]

5 hours ago, Etern4l said:

 

Sorry, catching up: is this the one?

 

https://www.moddiy.com/products/Honeywell-PTM7950-SP-Super-Highly-Thermally-Conductive-PCM-Paste.html

 

Apparently that's a stock paste used by Lenovo... Hmm.

yes but i got it from china too k 2 weeks to deliver to mexico.

here

https://www.ebuy7.com/item/641009621899

[KING19]

On 11/12/2022 at 9:56 AM, Etern4l said:

 

Sorry, catching up: is this the one?

 

https://www.moddiy.com/products/Honeywell-PTM7950-SP-Super-Highly-Thermally-Conductive-PCM-Paste.html

 

Apparently that's a stock paste used by Lenovo... Hmm.

 

Lenovo uses a special version of it called Honeywell 7954 or something like that. I dont know what's the difference between them.

 

Here another link of the pad version as the previous link is currently sold out.

https://www.ebuy7.com/index.php?route=product/product&product_id=632208664079

 

The pad version will have better results than the paste version because its already cured.

[jaybee83]

On 11/12/2022 at 1:21 PM, Rofa1234 said:

 

This is the reason:

 

FAQ: Why doesn’t Noctua specify thermal conductivity or thermal resistance of NT-H1 and NT-H2?

https://noctua.at/en/why-doesnt-noctua-specify-thermal-conductivity-or-thermal-resistance-of-nt-h1-and-nt-h2

 

 

 

On 11/12/2022 at 3:50 PM, Etern4l said:

 

Sounds like an elaborate excuse to me. Simply state the number, and give precise information regarding the testing conditions. Be a market leader, and other manufacturers will be expected to follow. This would be better than no information. As @Papusan said, withholding the information suggests there is an underlying performance issue. Case in point,  I would not think Noctua pastes are market leading.

 

Ideally though, we would have some independent testing for this. The benefit would be that we'd have a number representing expected thermal performance independent of application scenario and conditions.

 

ha, im not surprised that smth like this would some from a highly technical vendor like Noctua tbh 😄 and yes, in theory they might as well just throw out their own W/mK value, but im sure theyre aware that customers would be too easily drawn in and just go by that number alone. as we all know by now, that doesnt tell the whole story though. so its tricky going either way, do you let the product speak for itself and risk making customers think u have smth to hide or do u throw out some quick numbers to try and draw in lazy customers? 

 

On 11/12/2022 at 3:09 PM, Rofa1234 said:

I cannot test Kooling Monster KOLD-01 because they scammed me, there was no paste inside the package. This is really disappointing because I was curious about this paste  I'm currently testing Arctic MX-6 and Corsair XTM70.

um....wtf? can u contact them and try for a refund or replacement?

[Rofa1234]

On 11/12/2022 at 6:16 PM, Etern4l said:

 

Thanks for sharing although sadly it would be hard for people in US/Europe to validate this. No reputable reviews, and looks like limited availability outside of India. The only thing to go by are some anecdotal reports/Amazon reviews from Indian customers. Perhaps the product is very new, and/or the company is not very good at accurately measuring W/mK (but kudos to them for posting the 5 result anyway).

 

 

There are some Deepcool G40/G15 reviews like here and here but they are all tested on a desktop CPU afaik. There is no real big thermal paste test with a notebook I'm aware of (excluding me). Deepcool launched Z10 and EX750 last year, they are quite good as well but couldn't match G40 in my testing. All Deepcool pastes have a low W/m*K and some other companies as well...Akasa, Honeywell, Shin-Etsu.

 

jaybee83, I will try.

[Etern4l]

1 hour ago, Rofa1234 said:

 

 

There are some Deepcool G40/G15 reviews like here and here but they are all tested on a desktop CPU afaik. There is no real big thermal paste test with a notebook I'm aware of (excluding me). Deepcool launched Z10 and EX750 last year, they are quite good as well but couldn't match G40 in my testing. All Deepcool pastes have a low W/m*K and some other companies as well...Akasa, Honeywell, Shin-Etsu.

 

jaybee83, I will try.

 

That's better than nothing I guess, but if a 5 W/mK paste is a top performer then indeed the number is meaningless with respect to traditional pastes (and 74 vs 79 W/m2K probably doesn't matter either). BTW I wonder why W/mK and not W/m^2*K - I guess they wanted to keep it simple or something.

My guess is that a lot of the "difference between different pastes" we see in reviews is just due to application variance, especially when it comes to laptops with their poor heatsink geometry and pressure. 

 

Anyway, I doubt any relatively recent traditional paste is significantly better than say SYY or Phobya. The question is whether to go with a decent traditional paste or LM.

[radioactiv]

On 11/12/2022 at 12:08 PM, Rofa1234 said:

Deepcool G40 works great on my Dell Inspiron 15 7506, it is one of the best I have ever tried. I finished my new test and also included 5 reference pastes including Deepcool G40. MX-6 is a big improvement over MX-4/MX-5, although it's not reaching the very best but it's a solid paste and cheap. I'm sure it will be more competitive on most other devices. Corsair XTM70 is better but cannot reach my best pastes.

 

 

Arctic MX-6 21.5
73/79/75/75

 

Corsair XTM70 21.2
73/77/74/75

 

Deepcool G40 21.3
71/75/72/73

 

Thermalright TF9 21.3
72/76/73/74

 

Alphacool Apex 21.4
73/77/74/75

 

Arctic MX-4 21.5
79/82/81/80

 

Akasa T5 ProGrade Plus 21.5
73/76/74/74

How do you rank Thermalright TFX and  FuzeIce Plus with mx6? Any issue  or big difference? How about long term usage of each? Any drying or temp increases?

[aldarxt]

Which thermal paste am I using; here

https://www.amazon.com/Gelid-Solutions-GC-Extreme-Thermal-Paste/dp/B008S1QC0K/ref=sr_1_1?crid=1W1MJVKNLQ8WS&keywords=Gelid+Solutions+GC-Extreme+10g&qid=1668496939&sprefix=gelid+solutions+gc-extreme+10g%2Caps%2C472&sr=8-1

at $23.99 for 10g I could do alota swappin!!!

[jaybee83]

1 hour ago, aldarxt said:

Which thermal paste am I using; here

https://www.amazon.com/Gelid-Solutions-GC-Extreme-Thermal-Paste/dp/B008S1QC0K/ref=sr_1_1?crid=1W1MJVKNLQ8WS&keywords=Gelid+Solutions+GC-Extreme+10g&qid=1668496939&sprefix=gelid+solutions+gc-extreme+10g%2Caps%2C472&sr=8-1

at $23.99 for 10g I could do alota swappin!!!

 

thats a classic, for sure 🙂 

[Rofa1234]

On 11/15/2022 at 12:21 AM, radioactiv said:

How do you rank Thermalright TFX and  FuzeIce Plus with mx6? Any issue  or big difference? How about long term usage of each? Any drying or temp increases?

 

 

If I use G40 and Akasa T5 ProGrade Plus as reference I would say MX-6 performs something between 15-21 in this test because it's 1-3 degrees behind these two. Deepcool G40 performs similar as FuzeIce Plus and Thermalright TFX on my notebook at the end of the day. No idea about long term performance. MX-6 is much more viscous than MX-4/MX-5 which is probably the main reason why it performs so much better, for a direct die chip it's better.

[raptorddd]

On 11/14/2022 at 12:22 AM, KING19 said:

 

Lenovo uses a special version of it called Honeywell 7954 or something like that. I dont know what's the difference between them.

 

Here another link of the pad version as the previous link is currently sold out.

https://www.ebuy7.com/index.php?route=product/product&product_id=632208664079

 

The pad version will have better results than the paste version because its already cured.

are you going to keep using honeywell or will try other paste.? have you tried SSY.?

[KING19]

On 11/22/2022 at 9:43 PM, raptorddd said:

are you going to keep using honeywell or will try other paste.? have you tried SSY.?

 

Yea ofc since its the only paste that works well with my Laptop and its great paste for all gaming laptops, Its almost been 6 months since the repaste and it hasnt degraded one bit!.

 

I tried out SYY before but it didnt last long, I have the 2020 version and i heard the 2021 and 2022 versions improved the paste formula with decent results.