I replaced the single pipe heatsink on my T490 for one with dual heat pipes as the single couldn't sufficiently cool the CPU (was hitting 97C out of the box and after several re-pastes). So I set out on a journey to find a good thermal paste to use. However, finding a good thermal paste for laptops (more specifically on die usage) isn't as straightforward as one would think. A lot of the high end thermal pastes work best for desktop use, i.e. between a heatsink and an IHS, whereas on laptops you don't have an IHS so it's between the heatsink and the die. These high end thermal pastes usually all succumb to the same problem called the pump-out effect [1], where after a number of thermal cycles the performance of the paste begins to fall and it needs to be re-pasted.
I ended up going with Prolimatech PK-3 as it's a fairly viscous thermal paste, and it has held up well for the past six months. Although, I have noticed slightly higher max temperatures than when I originally applied it, but nothing alarming.
I ended up going with Thermal Grizzly's Kryonaut when I repasted my MBP (Apple does such a horrible job on their own). Due to how thick it is, pump-out basically can't happen. It's the thickest paste on the market.
It's designed for sub-zero overclocking (thus, the name), and has thinner siblings, but I chose it specifically how thick it is, and how it has no curing time. Sub-zero overclocks has the worst pump-out effect due to the vast range of temps you can travel.
The pump-out on my MBP with stock paste was absolutely absurd (hardly any paste left on the dies, almost entirely shoved to the very edges of it), but given the past year, it has had less thermal degredation than my tube AS5 historically has (which is near-zero, I just have to repaste it every 5-8 years, or "once per life of device").
Apple's own paste didn't survive 3 years, or basically went to shit right after Applecare ended. Suspect this is on purpose, but, obviously, can't prove that.
Edit: By the way, thicker pastes require a little bit more pressure to get good contact. I suggest everyone get a torque screwdriver that can be set to 0.6 Nm/5-6 lbf to screw their heatsink retaining screws on.
Intel and AMD have pretty consistent recommendations, as do GPU manufactures, and they range from 0.8 to 1.2 maximum. 0.6 is enough to not risk overtightening, and is also enough to make thicker pastes happy; 0.6 is also Noctua's universal recommendation.
Also, Arctic Cooling's MX4 is another popular thick paste (usually for GPUs, not CPUs, seems to be more consistent on large dies) that likes high pressure mounts.
Interesting, I had heard somewhat of a different take from Der8auer (who is associated with Thermal Grizzly). In [2] he explains that Kryonaut is a fairly wet thermal compound and that it does indeed suffer from the pump-out effect, and that in direct die applications like GPUs they recommend replacing it once per year.
Yeah, I watch his stuff already. He specifically mentions because GPUs are large hot dies. I'd never use Kryonaut on a GPU (I've had very good luck with MX4 on GPUs Ive had to repaste or replace stock coolers on).
Bare die CPUs (laptops, delidded CPUs, etc) don't seem to be big enough to cause issues. I imagine, if I have to repaste the laptop, if it even lives that long (it is from 2012 as it is), it's going to be another 4 years down the road.
Any readers understand how the pump out effect works?
In my imagination, it seems like the less paste you have between whatever is emitting heat, and the heat sink, the better- so long as there aren't air gaps.
I see there are many forum posts where people show that it does not, but it feels like the closer proximity of the heat sink to the hot object should _improve_ conduction!
In theory, yes, you would want just enough thermal paste to fill any gaps between the heatsink and die as the thermal conductivity of the heatsink (e.g. copper) is much higher than that of the thermal paste. Unfortunately, I haven't been able to find much on the internet beyond forum posts and anecdotal evidence regarding the pump-out effect. It would be interesting to see a more scientific study of this.
Thinkpads with a dedicated GPU come with a dual heat pipe heatsink, whereas ones without the dedicated GPU don't. So, you can take the heatsink for the former and use it on the latter, given that you make sure to insulate the part of the heatsink that is supposed to cool the GPU (this can easily be done by using a quality electrical tape like super 88 on the bottom side). You can buy the part aftermarket or see if Lenovo is willing to replace yours. The part number for the heatsink I put on my T490 is 01YU186.
I never understood why they don’t just put the dual heat pipe on all models. I run my t460p with gpu disabled, and the dual heatpipe cools the cpu so effectively that the fan is inaudible most of the time.
>Those who know me know that I am a bit fan of the oldschool Lenovo ThinkPad laptops
May be just me, as I was expecting oldschool to be IBM Thinkpad.
Anyway you could get $100+ worth of performance simply by using better thermal paste. Unfortunately most Laptop Vendors refuse to do it for whatever reason. May be it is trade off between how long the paste would last?
And while we have trillion dollar worth of silicon improvement in the last decade, are there any recent innovation in thermal paste or thermal cooling?
Moving heat around is pretty well understood science. Heat pipes are probably the most recent innovation, in that we started using them about 20 years ago in PCs, although they've existed in other applications for 50-60 years.
Regardless of how well you move heat around, all heat is waste. If the goal of better cooling is better mobility, then your bigger limiting factor is that you don't have sufficient battery technology to afford wasting that energy as heat, not that you don't have the technology to sufficiently cool the device. The best innovations in cooling are the improvements in silicon. In the past decade, we now have commonplace consumer PCs that don't need active cooling at all.
30 years from now we're going to look back at the people who tried really hard to cool their silicon better like the people who tried really hard to make their vacuum tubes last longer.
> And while we have trillion dollar worth of silicon improvement in the last decade, are there any recent innovation in thermal paste or thermal cooling?
This is the real crime. Its 2020, can't we have a better physical interface than a flat surface?
>This is the real crime. Its 2020, can't we have a better physical interface than a flat surface?
Carbon nanotubes!
But they're rather fragile, expensive to manufacture on a commercial scale and you'd be fighting 30+ years of legacy products, marketing and manufacturing infrastructure.
Like OP I also have a t420s. It always had overheating issues. A few months after purchase, back in 2012, I even sent if for warranty repair because it routinely got "too hot to touch" and the fan kept going at insane speeds and the CPU throttling. My version only has an integrated GPU (Intel HD3000). When it came back from warranty I guess they replaced the thermal paste and installed some better Fan driver or something. The laptop still overheats on heavy loads occasionally but I have used it ever since (I can only spend on a new laptop every 10 years max).
This year (2020) I completely dissembled it and replaced the thermal paste with new good quality one. It still overheats and throttles.
Back then (2012) I chose this laptop mostly going on ThinkPad reputation for easy upgradeability/ repairability. Boy did I get unlucky... what a bad choice. It's the hardest laptop to disassemble I have ever had the displeasure to work on. To get to the fan assembly you literally have to remove every other component.
In retrospect the reason for this is obvious. The t420s was supposed to be a slim version of the t420 (which by then was already considered thick as a tank). I guess in a away the t420s was Lenovo's first attempt at a proper slim portable laptop. Unfortunately for me the result was a slim, hard to disassemble, short battery life, overheating mess...
I changed the fan and heatsink unit in my HP elitebook convertible tablet thing. I also had to remove every other component. Never did get Bluetooth back afterwards.
The manufacture paste isn't supposed to expire and is designed for a long long life (5-10 years). That said, pump-out still happens.
Pump out is the result of thousands of thermal cycles of the heat source and the heat destination on the thermal interface material.
You might be able to fix your problem by replacing the thermal paste, but caveat is that "performance" thermal paste is MORE susceptible to pump-out and would need to be replace more frequently.
Manufactures choose "less performant" thermal pastes because they have longevity. You don't tell users to re-paste their CPUs on a year/bi-yearly basis.
DIY people choose the "performant" options because benchmarks happen only on the first day and non-performance means it's time to upgrade.
For you, replacing the thermal paste will improve the transfer of heat to the heat sink, but if your CPU/GPU generates more heat than the system can get rid of, it's MAX fan time. It's really a function of efficient your system is at generating and dissipating heat and what you are doing on the laptop.
I've had to do this on much newer thinkpads. The fan on my T470 was on constantly until I went in there and found that the manufacturer applied paste was done badly. Grr.
This is just my experience, but I had observed a wild temperature difference at idle and load on two identical MacBook Pros.
I’d already recently (within days) replaced the thermal paste with the same paste but still one of them was getting 10-15 deg C hotter.
The one that was cooler had been through AppleCare and it appears the copper surface had been scratched up intentionally as well as some dents on the fins.
The worse performing heat sink looked immaculate aside from some dust over the years. The copper surface was polished smooth.
Just for the heck of it I swapped the heat sinks and now the temps were reversed.
In the end I figured maybe the heat pipes were damaged or there was some performance differences of the heat pipes during manufacturing.
> If this is a sponsored blog entry? Unfortunately not, but maybe in the future …
Still, somehow the article reads like a promo for that paste.
I was curious whether the original factory paste indeed gets older to such degree that it causes the overheating. Instead, the only mentioning was that the old paste came off easily.
I replaced the paste on a few old ThinkPads. Mostly because I was cleaning/replacing the fan. For paste I used some no-name white goo, the kind you get in those pill-sized squeeze pouches. Usual procedure, nothing special, just wiped off the old thin layer; nothing crusty or peeling old paste. Other as old ThinkPads hum along well without any re-pasting.
Perhaps, this story is just a case of reseating the fan and reapplying the paste, than ThinkPad paste got old.
Paste certainly does have a shelf-life. The stuff they use in the factory even on modern laptops optimises longevity over performance- that’s why you get thermal benefits by re-pasting new laptops.
But the paste still ages and dries out, even liquid metal will age (by virtue of it eating the metal of the heatsink over time).
Good rule of thumb is repaste after 4 years if it’s a factory paste, and every 18 months if it’s an aftermarket performance paste (like arctic silver).
It will still perform decently; it’s not like your machine will die, but it’s not ideal.
Last weekend I replaced the motherboard of my i5 x220 and took the chance to change thermal paste with Thermal Grizzly Kryonaut. I cannot say how it is performing since I'm only using the laptop with a minimal Debian installation in tty mode but everything seems ok so far.
My next project is to replace the fan/heatsink of my i7 x220 with the x230 one, which is supposed to be just as efficient but noticeably quieter.
I have a couple of X201 Tablets that I picked up cheap a few years ago, and they could definitely use this. Running BOINC/Rosetta@Home, they often get into the high 90s.
I should have thought about thermal paste - these machines don't have clogged fans.
Laptop wise I've changed the paste on 2 macbook pros and 1 macbook. 2/3 times CPU throttling got better. I will probably never do it again, smartest move is to buy a new laptop and sell the old one on ebay.
I ended up going with Prolimatech PK-3 as it's a fairly viscous thermal paste, and it has held up well for the past six months. Although, I have noticed slightly higher max temperatures than when I originally applied it, but nothing alarming.
[1] https://www.overclock.net/forum/28099300-post10.html