Traditional thermal paste or liquid metal paste?

I have problems with temperatures, specifically after few days after repasting (Artic MX-4), where the temperatures on load rise from high 70s/low 80s to mid 90s. (The CPU is also undervolted by -0,80mv)

I read this from another forum from a guy with the same proble as mine,(thought his temperatures lasted longer):

I already ordered a tube of TG Kryounaut ,but Im wondering if I should order Liquid metal paste. The only con I see about the liquid one (besides you have to be careful to apply it), is that isnt suited for uneven heatsinks usually found in laptops (thats what I read in one of the threads of this forum).

So what do you guys suggest?

 

Knowing model/CPU would help to put your temps in context

Kryonaut is a better paste than MX4

Any TIM will benefit from a good fitting heatsink. The best fit is the one that requires the least TIM - flat and close to the die. LM amplifies the risk by moving about when there's big gobs of it, the more you have to use to gap fill the more risky it becomes. When you repaste check your heatsink alignment, make sure you didn't make it worse by bending a heatpipe out of alignment or folding over a thermal pad

 

Its a MSI GE62 2QD Apache Pro with a i7 5700HQ.

I have applied the thermal paste multiple times in different ways, the problem is that the lower temperature lasts few days. I move the laptop like once or twice a day, but just to a different room in the same house.

 

Phoenix had terrible results with Kryonaut. Oddly enough, I also noticed a 5C temp rise with Kryonaut after a week of high temp stress testing as well. I believe that is because of slightly imbalanced heatsink pressure, however.

If you are having a problem with pumpout this quickly with MX-4, then Kryonaut may not help you. I suggest VERY HIGHLY a very vicious thick paste, like Noctua NT-H1 (which most overclockers STILL swear by; as this has a very low rate of degradation). Other options are IC Diamond (but that company has done some very shady things, including lying about defective batches which scratched dies and heatsink surfaces; there was a huge angry thread over here long ago about that), and Phobya Nanogrease Extreme. Mr Fox had good results with Nanogrease, while I had bad results. Your Mileage May Vary.

 

What about liquid metal paste? Not recommended?

 

If there is already a high pump out then the fit of the HS to chip is most likely faulty. In a bad fitting situation LM is not recommended as it not only will still pump out but a conductive material is being pumped out.

 

Thanks for the replies, acording to this chart,there are more viscous thermal pastes:

http://www.tomshardware.com/reviews/thermal-paste-comparison,5108-12.html

Should I buy one of those instead?

 

Well, I already answered you.
Noctua NT-H1, Phobya Nanogrease Extreme, IC Diamond and Gelid Extreme are all good choices.

 

Hard to believe ICD was so low on that link/test.

I have a Delidded i7-8700k with Liquid ultra under the IHS and ICD 24 on top between the Evga AIO (Single fan radiator) and don't get over 60c while playing PUBG for an hour at a time. That's at stock clocks though.

Would've used Kryonaut but ran out of it. Was getting better temps while using that.

Charlie

 

@Temp1234453 I've got some overheating issues too with Coollaboratory Metal Pad ( http://www.coollaboratory.com/product/coollaboratory-liquid-metalpad/) and with Thermal Grizzly Conductonaut (Liquid Metal). The heat was going so high despite advertised thermal conductivity... and i couldn't figure it out...

It turns on the pressure on the CPU DIE was not big enough. You see, the thermal paste that DELL put on the CPU was quite thick originally (around 0.5mm) and that was adding extra tension on the spring-loaded heatspreader plate. So, when I went with a very thin layer of TIM, it just didn't work.
What I had to do, was to ad a copper plate as spacer between CPU DIE and heatsink and believe me, once you do it right, Liquid Metal is just crazy... plus you won't have to replace it (it won't dry, won't change its properties). Just beware to not drop on the circuits, as it's metal and forms very little droplets which are hard to see, but can cause real problem in there...

 

Thanks for the recommendations, I finally bought Gelid Extreme.

Havent thought of this. Did you have bad temperatures just after you repasted? Or they became bad after some time?
Anyways, the idea of adding a copper plate in between is good,I'll try if I still get the same problems.

 

I like IC Diamond, it's safe to apply for noobs like me. But I don't like it scratching the die, it doesn't affect or degrade CPU performance. Gelid Extreme is also a good choice, let us know of your results.

 

@Temp1234453 At first I used Coollaboratory Metal Pad (a very thin metallic foil, I guess an Indium alloy, which would melt at 60 degrees and act like a thermal bridge), which worked well for about 3 weeks at first, and less than one week after I opened the machine and reapplied. Second time I skipped the copper spacers, thinking that was the issue. I was wrong. Half of the CPU die was not even in contact with the heatspreader anymore. All metal had gathered on the other half, creating a thick "island".

Thinking that metal pad is not the right thing to use, I decided to try liquid metal, but I had the exact same problems (this time the problems were more immediate and easier to follow). So, after one full day of experimenting I figured it out what was the cause.
Point is you can get overheating immediately, or after a while, after things start to move a bit. All parts are expanding during heat, then they will contract and the process will repeat many times. So, if there is something that can get loose, it will happen. Air may be trapped in between, surface contact may be broken. I even read about uneven heatsink plate, which was pressing more on one side of the DIE and less on the other end, making the cores dissipate heat unevenly (having maybe 5 degrees or more difference between cores).

 

Yeah, most overclockers on *desktops*. Both me and @Mr. Fox tried Noctua NT-H1 on oc'd 4930Mx (Haswell) with awful results after only a few days with benching. A No go from me.

@Temp1234453 forget Liquid metal if you have problems with pump out. I would use Phobya NanoGrease Extreme or ICD. Both should work well with a bad fitting heatsink. FYI. Artic MX-4 is trash on laptops worse cooling.

 

It's strange, @Papusan I was reading overclock.net where people were BASHING Kryonaut (on the 8700K CPU thread) and praising Noctua because Kryonaut kept having to be repasted every 2 weeks, while Noctua had higher temps but was stable for months....

I guess you just can't win

 

And you talk about laptops? Have been there, one week and new repaste. All know laptops have ****y cooling vs. desktops. And 4930Mx ain't easy to cool even stock.

 

No. this was the 8700K thread (or the 7700K but i THINK it was the 8700K thread) on OCN. No laptops there. All desktop users. A lot were on water cooling and complaining about Kryonaut drying out (even on their de-lidded processors with LM inside the IHS and Kryonaut on the outside).

 

Exactly

 

d00d, Kryonaut last 3 days on my taptop before the temps start going up so I immediately remove that garbage paste.

 

Phoenix, did you have higher CPU core temps on cores 1 and 3, and lower core temps on cores 2 and 4? Or was it too far back to remember?

 

I fix it for you Not all HS is TRIPOD
http://forum.notebookreview.com/threads/thermal-grizzly-kryonaut @Phoenix

 

That was on his old Clevo. Was referring to his GT73.

 

Then I expect ICD → Liquid metal

 

I just noticed something. My CPU heatsink seems to be made of alluminium:


While some heatsinks for are made of copper (found on Ebay), seems like an upgraded version:


Would it make a big difference if I bought a copper one? Theyre about ~50/60€

 

If it's aluminium, drop it, forget it. A flawed HS. And for Gods name don't use Liquid metal if this is what you have.

 

Yes I know about alluminium and liquid metal paste.
Im saying its alluminium because its silver color. Is there silver colored copper?

 

I don't think so, if not the HS has a kind of metal alloy.

 

Could be, doesnt feels like alluminium but Im not an expert


At minute 12:00 you can see, but dont know if you can tell.

 

Aluminium. What a flawed design Why not use copper, same way as they use on GPU coold plate? Oh'well, maybe they saved 20-50 cent. At least they didn't went for TRIPOD design on top.

 

I thought Liquid Metal would not be affected by this, as it's in a very thin layer and seems to adhere very good to metal surface. Could you explain it in more detail?
Thanks.

 

Pump out is already explained in previous posts. Means your heatsink has bad fit on IHS/die. High risk for leakage.

 

I got that, but I assumed it would happen only on TIMs that are relatively thick and fluid. Whilst LM is fluid, the layer is very thin compared with any thermal paste, so I assumed it wouldn't be pushed out by thermal cycling. This is the reason I tried to get a more detailed explanation. Maybe an example.
I'm only asking because I used Liquid Metal for my CPU and I want to understand the phenomenon better and eventually to find solutions, in case it will happen.

 

None talk about pump out for Liquid metal.
1.) If you have a bad heatsink fit... You will need more Liquid metal aka big risk for leakage. Second... Heat and exposed with air due bad heatsink fit will make that the Liquid metal will react more on the copper. It will dry out faster than necessary aka Liquid metal will be brittle.

Liquid metal ain’t a good option if the fit is bad aka you have a uneven/Warped heatsink.

 

You should assume any "silver" color on a heatsink plate is aluminum so that rules out LM.

For laptops you should use thermal paste designed for low clamping pressure. There are some "shootouts" here that give you some clues. You have some recommendations from the expereienced guys here.

For laptops with imperfect heatsinks, I have used both IC Diamond and and Gelid Extreme.

- The IC diamond scratched a copper heatsink a bit and lasted a few years before temps started to ramp up. It was in poor condition after a couple of years based on temps and how it looked.

- The Gelid Extreme was on a Dell XPS laptop for almost 2 years with no issues and virtually no rise in temps. I have not inspected the paste but the performance speaks for itself. I got about 6*C cooling improvement with Gelid Extreme, which was consistent with others on the XPS forums.

I recently applied Gelid Extreme on a 10-year old MacBook Pro with a slightly warped heatsink; that worked very well (partial credit to the Milwaulkee Hole Dozer and two large holes in the case bottom for air intake)

 

If there is pump out of any TIM then 99% of the time there is poor contact and or bad fit of the HS to the die.

 

Is there any liquid metal paste that has a thicker structure for laptops without perfect heatsinks?

 

I think No. LM needs perfect HSF. Try ICD or Phobya.

 

Laready have ICD7 and Phobya nanogrease, but I want better. But pretty much all laptops these days have in some way imperfect heatsinks.

It would be very beneficial for everyone if there was a thicker galium based paste.

 

Gallium melts in your hand and at room temps, so looking at high viscous LM is unlikely to be found.
Did you try Cool Lab's Metal pad made especially for Notebooks? http://www.coollaboratory.com/product/coollaboratory-liquid-metalpad-notebook/

 

There's coolaboratory liquid copper. A medium filled with copper material.

Seems to be close to the actual LMs in test.

 

I heard a lot of bad stories about this for laptops. Seems to dry up quite quickly with an uneven heatsink.

But forgot that about galium, Too bad though I want a good solution for in notebooks.

 

While we're on the subject of destroying your laptops with conductive balls of doom, just bought 30g of this for $32.
Worth it or not?

https://www.ebay.com/itm/Galinstan-...sh=item4b38602455:g:488AAOSwzBJaf3aT#shpCntId

 

I wonder if it even works as a thermal paste, even though it's advertised as one.

Please test and report back

 

It's coming from Russia. Will be awhile.
I'll probably throw it on my 290x video card first.

 

Nope. It looks like Coolaboratory changed the composition of Liquid metal. Not the same as before. Moe like the other LM. be sure HS work as intended and use Conductonaut.

Coolaboratory liquid copper is already tested. Trash on laptop cooling!!

@t456 used it with with success if I remember correct.

 

Oh well :|

 

Don't look you blind on reviews (better desktop cooling - or customized test bench)

 

Hmm the liquid copper paste still catches my interest though, maybe first will try it in my old desktop and or old Vaio Z that I still have laying around here somewhere.

But here cool labs state its paste like, does this hold true? http://www.coollaboratory.com/product/coollaboratory-liquid-ultra/

 

Not sure what you mean regarding the paste.
People who used Liquid Ultra has mowed on over to Conductonaut. Silicon lottery which delidding and sell binned chips as well.

Coolaboratory copper paste is also tested on notebook graphics with not good results. Skip the idea.

 

Hmm i thought if cool labs ultra was mlre thick compared to conductonaut it would be a better candidate for non perfect heatsink situations.

Too bad the copper paste isnt that great. Would have loved to have something better than phobya.