The Official Thermal Interface Material thread

[Simone]

Check

The Official Thermal Interface Material thread

I am looking for an end solution for my 7900 XTX air cooler as the hotspot temp has gotten worse. I have already been recommended PTM7950 in the past but I still want to push the card to its limits without it going beyond 95C. What does this have to do with not choosing PTM7950? It can handle...

www.techpowerup.com

Ho'fuck me that's exACTLY what I've been clawing for, thank you so much. Awesome.

I think I'm just going to change everything I own to PTM7950 once I make sure I can source a genuine sheet of it.

Liquid Metal is just too much of a hassle.. especially in my Scar 18, the 2023 edition I had I barely owned it for 4 months before I saw the oxidation in the centre happening. Absolutely hopeless for Conductonaut and all other liquid metals out there.

Unless I'm wrong somewhere and glossed over something, there literally isn't any reason to NOT use PTM7950 over Liquid Metal or anything else, right?

My only worry is whether or not mounting pressure is a factor to worry about, most of my intended applications are going to be on laptops.

 

[R-T-B]

Unless I'm wrong somewhere and glossed over something, there literally isn't any reason to NOT use PTM7950 over Liquid Metal or anything else, right?

Technically the only reason to go LM over PTM7950 would be a slight initial performance edge, though longevity is an issue as you pointed out.

 

[Simone]

Technically the only reason to go LM over PTM7950 would be a slight initial performance edge, though longevity is an issue as you pointed out.

The weight of the pros and cons in that regard are heavily leaning toward PTM7950 for me then.

I don't want to have to deal with respreading, cleaning oxidisation, risk of pumpout, hotspots etc.. and the 'slight initial performance edge' sounds only useful for getting the highest benchmark numbers out of your machine for bragging rights.

I think I'm sold and will move to PTM7950 now on everything.

For other things like VRMs and so on what would one recommend? I used to use MX-4 but I'd rather something that lasts a long time and doesn't get crusty if possible.

If anyone thinks I'm doing the wrong thing though please save me. ';d

Also, the funny thing is, I've overheard that PTM7950 only gets better over time as it gets thinner and thinner from mounting pressure, given it isn't a liquid, not prone to capillery and doesn't pump out, (to my logic) so the numbers in that review seem interesting, was the 'curing time' taken into consideration when comparing to liquid metal?

 

[InVasMani]

I've been pretty hopeful about these pad based thermal interfaces the last handful of years maturing to the point that they can readily replace the need for paste/liquid metal options easily with no significant downside. I don't mind a small deviation in terms of cooling between them and liquid metal or a good paste prior to dry out, but them reaching a point where that's less and less significant is all these pads really need to supplement and replace those other options pretty much outright for most people. Set it and forget it and no ideally a material that's non conductive at the same time. Hopefully reusable too, but even if it isn't it's not a huge problem. Clean up is another factor if it's generally easier to remove and cleanup than the other options that's a plus.

 

[wNotyarD]

I think I'm just going to change everything I own to PTM7950 once I make sure I can source a genuine sheet of it.

If you can find it more easily, you can get some Thermalright Heilos PTM pads.

 

[FreedomEclipse]

Might be here on Friday

It should be rebranded PTM7950.

 

[freeagent]

I am using SYY-157 on my CPU and GPU, works pretty good.

 

[FreedomEclipse]

And now we wait

The heat cycling begins. PTM7950 has a burn in time of 48hrs at room temp so im assuming a good few heat cycles before results.

:EDIT:

According to reddit. best results comes after about a week of heat cycling.

 

[srich_2m]

Artic Silver since COMPUSA was still around. Easy to apply and still is reasonably priced. Grizzly is too proud of its products for me to try. Cost is high, and most likely will perform better than good ol' silver.

 

[KLiKzg]

Artic Silver since COMPUSA was still around. Easy to apply and still is reasonably priced. Grizzly is too proud of its products for me to try. Cost is high, and most likely will perform better than good ol' silver.

Kryonaut is only paste I use, for 24/7 BOINC use.

 

[Pictus]

Ho'fuck me that's exACTLY what I've been clawing for, thank you so much. Awesome.

You are welcome.

I think I'm just going to change everything I own to PTM7950 once I make sure I can source a genuine sheet of it.

Check

Honeywell PTM7950 SP Super Highly Thermally Conductive PCM Pad

Buy Honeywell PTM7950 SP Super Highly Thermally Conductive PCM Pad for $5.99 with Free Shipping Worldwide (In Stock)

www.moddiy.com

Liquid Metal is just too much of a hassle.. especially in my Scar 18, the 2023 edition I had I barely owned it for 4 months before I saw the oxidation in the centre happening. Absolutely hopeless for Conductonaut and all other liquid metals out there.

Unless I'm wrong somewhere and glossed over something, there literally isn't any reason to NOT use PTM7950 over Liquid Metal or anything else, right?
My only worry is whether or not mounting pressure is a factor to worry about, most of my intended applications are going to be on laptops.

Low mounting pressure and/or uneven surfaces is a problem for PTM, but laptops usually have enough pressure for the PTM.

For other things like VRMs and so on what would one recommend? I used to use MX-4 but I'd rather something that lasts a long time and doesn't get crusty if possible.

Maybe this

Shin Etsu MicroSi Thermal Interface Material X23 8117 3g

Buy Shin Etsu MicroSi Thermal Interface Material X23 8117 3g for $14.99 with Free Shipping Worldwide (In Stock)

www.moddiy.com

 

[StarlightGalaxi]

MX-6 is not very good on direct die from my experience I applied it to my TUF 4070 Ti and it always pumped out, intially the temps were fine between core and hotspot was a 10-11C delta now I'm seeing as high as 17-22C just after 1 week of it being applied and I did the spread method made sure the die was completely covered etc, now I'm waiting on PTM7950 that I purchased I should recieve it in a couple days or tomorrow, I've heard good things about this stuff and how resiliant it is to pump out effect but so far I am highly disappointed in MX-6 and will never use it on a GPU ever again.

 

[lexluthermiester]

MX-6 is not very good on direct die from my experience

I've had no issues with it on the laptops I've repasted.

I applied it to my TUF 4070 Ti and it always pumped out

I hate to seem rude, but pictures or it didn't happen. Everyone keeps crying "pump-out". It's a myth for the most part.

 

[StarlightGalaxi]

I've had no issues with it on the laptops I've repasted.

I hate to seem rude, but pictures or it didn't happen. Everyone keeps crying "pump-out". It's a myth for the most part.

If pump out is a "myth" then why did my hotspot temps increase after just 1 week of it being applied? clearly pump out happened, I'll take pics when I take the card apart again to apply PTM7950.

 

[lexluthermiester]

If pump out is a "myth" then why did my hotspot temps increase after just 1 week of it being applied?

There could be a number of reasons why that might happen. Blaming it on "pump-out" is not a valid catch-all excuse.

 

[StarlightGalaxi]

There could be a number of reasons why that might happen. Blaming it on "pump-out" is not a valid catch-all excuse.

I highly doubt that.

 

[lexluthermiester]

I highly doubt that.

How about these reasons:
1. Improperly applied paste.
2. Improper amount of paste used.
3. Heatsink not fully contacting the die after being reassembled.
4. Paste not the issue at all and hot-spot is due to silicon lottery.
5. Heatsink improperly manufactured(imperfections in contact surfaces).
6. Heatsink not well suited for the job being asked of it(this is not likely in your case, that card's heatsink is huge).

I could keep going down that list, but you get the idea. "Pump-out" is moose-muffins until proven, and it has yet to be definitively proven by ANYONE. It's a damned myth. Figure out what is really going on with your situation and be objective about it. Don't use the "pump-out" scapegoat.

 

[StarlightGalaxi]

How about these reasons:
1. Improperly applied paste.
2. Improper amount of paste used.
3. Heatsink not fully contacting the die after being reassembled.
4. Paste not the issue at all and hot-spot is due to silicon lottery.
5. Heatsink improperly manufactured(imperfections in contact surfaces).
6. Heatsink not well suited for the job being asked of it(this is not likely in your case, that card's heatsink is huge).

I could keep going down that list, but you get the idea. "Pump-out" is moose-muffins until proven, and it has yet to be definitively proven by ANYONE. It's a damned myth. Figure out what is really going on with your situation and be objective about it. Don't use the "pump-out" scapegoat.

Still doesn't explain why my hotspot temps are literally 10-11C from the core temps for a few days after applying the MX-6 then they gradually increase over time if it was everything you have stated then I would have bad temps from the get go and not after the first few days, if the die wasn't being properly contacting the heatsink/cooler I would have overheating right after being reassembled which is not the case here. and I have repasted this card 3 times same situation temps started to climb after the intial few days and everytime I took the card apart part of the die was being exposed which is creating the hotspot increase even though it was fully covered and yes I used a good amount of paste on the GPU, you can keep to your beliefs and say it's a myth but from what I'm experiencing this is a pump out issue. and if PTM7950 doesn't exibit this bevaviour then my assumption is correct.

 

[lexluthermiester]

Still doesn't explain why my hotspot temps are literally 10-11C from the core temps for a few days after applying the MX-6 then they gradually increase over time if it was everything you have stated then I would have bad temps from the get go and not after the first few days, if the die wasn't being properly contacting the heatsink/cooler I would have overheating right after being reassembled which is not the case here. and I have repasted this card 3 times same situation temps started to climb after the intial few days and everytime I took the card apart part of the die was being exposed which is creating the hotspot increase even though it was fully covered and yes I used a good amount of paste on the GPU, you can keep to your beliefs and say it's a myth but from what I'm experiencing this is a pump out issue. and if PTM7950 doesn't exibit this bevaviour then my assumption is correct.

Ok, you do you.

 

[FreedomEclipse]

Here are my thoughts on the Thermalright Heilos so far....

I have noticed when I boot my laptop up after a while. Idle temps run a little high at 55'c and as high as 63'c if its doing something in the background like a small Antivirus update. -- this massively confused me and I immediately thought there was something wrong with the Heilos as I had seen the temps idle around 35-38'c on previous occasions.

I ran some guildwars 2 about a few days ago to give it some burn in time and temps maxed out around 74-85'c around that time but we were in a midst of a heatwave which had ambient temps between 27-31'c.

So The Heilos being a phase change pad and noticing the idle temps were higher than usual, I decided to start the laptop up this afternoon and put a little heat through it by loading up GW2 again. Temps averaged out at 62-65'c while the highest was 72'c while in game which proved to me that the Heilos was fine and still doing its job. It needs a bit of heat before it will do its best work. (current ambient temp is 21'c)

Quit out of the game and started using the laptop for some web browsing and youtube etc. temps are back to idling as low as 34'c but averaging around 36-37'c. I had never seen the laptop idle at 34'c with MasterGel Maker.


Im not sure if the Heilos is currently at its peak or it still needs more curing but its worked great for me so far.


::EDIT::

Yes. I can confirm that idle temps are higher If the laptop is left turned off for any amount of time where its allowed to completely cool off. I booted my laptop up to double check the idle temps and test some headphones that I found. idle temps are back to being 40-48'c. So its confirmed that PTM needs some heat before it does its best work.

 

[wNotyarD]

Here are my thoughts on the Thermalright Heilos so far....

I have noticed when I boot my laptop up after a while. Idle temps run a little high at 55'c and as high as 63'c if its doing something in the background like a small Antivirus update. -- this massively confused me and I immediately thought there was something wrong with the Heilos as I had seen the temps idle around 35-38'c on previous occasions.

I ran some guildwars 2 about a few days ago to give it some burn in time and temps maxed out around 74-85'c around that time but we were in a midst of a heatwave which had ambient temps between 27-31'c.

So The Heilos being a phase change pad and noticing the idle temps were higher than usual, I decided to start the laptop up this afternoon and put a little heat through it by loading up GW2 again. Temps averaged out at 62-65'c while the highest was 72'c while in game which proved to me that the Heilos was fine and still doing its job. It needs a bit of heat before it will do its best work. (current ambient temp is 21'c)

Quit out of the game and started using the laptop for some web browsing and youtube etc. temps are back to idling as low as 34'c but averaging around 36-37'c. I had never seen the laptop idle at 34'c with MasterGel Maker.


Im not sure if the Heilos is currently at its peak or it still needs more curing but its worked great for me so far.

Makes me want to go the PTM route on my lappy, even if its i5 isn't direct die.

 

[FreedomEclipse]

Makes me want to go the PTM route on my lappy, even if its i5 isn't direct die.

How old is your laptop if it still has an IHS???? laptop chips have been direct die for the longest time.

 

[wNotyarD]

How old is your laptop if it still has an IHS???? laptop chips have been direct die for the longest time.

Not that old, actually. Some 5 years. Check my sig, it has a bonafide LGA1151 i5-9400.

 

[FreedomEclipse]

Ah yeah. MobileWorkstation(tm)

PTM should still get good results. although not as good as direct die unless youre happy about delidding.

If the Heilos is available in Brazil. It could be a cheaper alternative then ordering authentic PTM from Moddiy and the gamut of fake randoms on ebay and amazon that ship from china. I got my Heilos from Thermalright's amazon store for £5.89. Moddiy is $5.99 but it will probably cost the same in shipping.

 

[oiawshboiwsahnoidaws]

Makes me want to go the PTM route on my lappy, even if its i5 isn't direct die.

I have ptm on my 12700k and it works great but don't expect miracles with an IHS, i saw 2-3°C improvement compared to best paste i had on hand (which granted is pretty cool already)
Oh i also tried paste version of ptm and works as good