Friday, January 24th 2025

RTX 5090 FE Liquid Metal Swap: Thermal Paste Performs Just Fine

Did you catch our launch review of the NVIDIA GeForce RTX 5090 Founders Edition graphics card? If you did, you'd have noticed the hoops we had to jump through to disassemble the card without breaking anything. As we carefully peeled the card's aesthetic finery and worked out way down to the GPU, we found that NVIDIA is using a liquid-metal thermal interface material (TIM) between the "GB202" GPU and the unified vapor chamber plate of the Double Flow Through cooling solution. After our photography the problem arose of putting the card back together, and we wondered if using a mainstream thermal paste would be a suitable replacement since we didn't have the patience to toy with liquid metal given our review volume. So we pulled out our trusty tube of Arctic MX6, gave it a suitable application, skipped the gasket, and put the card back together.

Here's what we found—the GPU temperatures rise by about 2°C on average over the stock liquid metal TIM. In the graph below, you will see both TIM applications compete with each other over a 370-second stress from a benchmark run. In the chart the start temperatures are slightly different, this does not mean that idle temperatures are higher with liquid metal. At the end of test the maximum temperature reached with the stock liquid metal TIM is around 77.6°C. The Arctic MX6, on the other hand settles to 79.4°C. This +1.8°C temperature increase really isn't significant at all—room temperature changes between summer and winter will cause bigger swings. We also tested performance, and it was spot on, the same as pre-disassembly—not a hint of thermal throttling. Both values are safely below the 90°C thermal threshold for the RTX 5090—that's right, NVIDIA raised the thermal throttle point, it's not 83°C anymore like on the RTX 40-series Ada. Taking the RTX 5090 apart and putting it back together was a challenging experience, but we're glad we didn't have to do a liquid metal application to ensure trouble-free operation. This is good news for all DIYers—don't bother with replacing the liquid metal—a thin thermal paste application works fine, too.
Add your own comment

62 Comments on RTX 5090 FE Liquid Metal Swap: Thermal Paste Performs Just Fine

#51
R-T-B
remekraDon't know if it's because it was designed with MCDs
This. It's the uneven surface. PTM deals with it just fine though due to how it melts and reforms constantly.
Posted on Reply
#52
Arkz
FiRe"just fine", then proceeds to show ~2 degree temperature difference. In that case I might as well use cotton wool just as well if 2.3% is classed as "fine"
Can you not read?

"stock liquid metal TIM is around 77.6°C. The Arctic MX6, on the other hand settles to 79.4°C. This +1.8°C temperature increase really isn't significant at all—room temperature changes between summer and winter will cause bigger swings. We also tested performance, and it was spot on, the same as pre-disassembly—not a hint of thermal throttling. Both values are safely below the 90°C thermal threshold for the RTX 5090"

If it was a device that ran at 8'C then a 2 degree rise would be significant. This isn't.
Posted on Reply
#53
ypsylon
blufeb95Copper is better than aluminum at dissipating heat, the biggest problem is copper is 3 times as dense as aluminum and is impractically heavy on a big heatpipe radiator and while it cools better it's typically within a few C and not worth making the card weight twice to three times as much.
Not to mention price of copper vs aluminium which would add significant chunk of $ to the card itself for like 1-2C difference plus then painting copper black to reduce evaporation potential makes no sense.

Anyway, it's good to know that TIM paste will do just as good for normal operation as LM while being much safer long term DIY.
Posted on Reply
#54
Arkz
LycanwolfenWhat thermal paste was used I wonder. I am using thermal grizzly extreme on few of my card it lowered the temps lower than liquid metal.
Arctic MX6. It's only two short paragraphs man, read it.
Posted on Reply
#55
Veseleil
MX-6, seriously? :laugh: That thing will perform fine for a couple of days, maybe weeks, depending on type of load and usage time.
Then it will pump out, the leftovers will dry out (burn), and you'll figure what junk that paste actually is. Meanwhile, my 6900xt LM application (conductonaut) is still going strong with the same temps after а year, without reapplying (ever).
And this GPU was even OC'ed to 2.86GHz on air, tested and benchmarked until I got bored to death. If anyone is interested about my experience with this POS paste (that came with €250 order from Arctic webshop btw) on my 6900xt, you can start from here. Cheers.

P.S. Beauty of LM.
Posted on Reply
#56
freeagent
VeseleilMX-6, seriously?
Just my opinion, I have always felt Arctics stuff was a bit sub par, that's why I do not run anything from them.

These are 20 year old feelings so.. nothing new to me lol :laugh:
Posted on Reply
#57
Onasi
@freeagent
It’s fine for CPUs and most mainstream GPUs. Easy to apply, can be found basically anywhere - a no-brainer choice. Then again, I don’t fuck with any CPU that goes above 150 or GPU that goes above 250, so that mostly is what informing my experience. On a ridiculous space heater like the 5090 I probably wouldn’t really trust any paste long term and would run LM or PTM. That being said, our 3090s at work are still running whatever original paste NV put on the FEs and I’ve yet to see any thermal degradation that would be statistically significant, so… eh.
Posted on Reply
#58
freeagent
I like SYY-157. No name brand, awesome TIM. My latest tube is 40g.. but I have bought a lot more than that over the last few years. Swapping CPUs and coolers all the time.. its good stuff.

And yeah.. I saw TF7 did better in the list at Elmors.. cant say I agree.

But yeah.. back to the show I guess..
Posted on Reply
#59
Veseleil
freeagentJust my opinion, I have always felt Arctics stuff was a bit sub par, that's why I do not run anything from them.

These are 20 year old feelings so.. nothing new to me lol :laugh:
TBH, this whole thread/headline sounds like a bad joke of an Arctic marketing team.

Now compare this to that "hotspot temp reading no more" thread. I swear to God, this site/forum is going places lately... Such extremeness...
And... what if I told you that Arctic fans (fans like ventilators, not brand fans like people - in these crazy times I really need to clarify) are a cheap junk that starts to show its low QC and build quality in just a couple of years, less than a year, and sometimes right out of the box (especially the 5-pack box ones!)?
I bet some TPU members would crucify me, trying to defend that multi-mil$ company, and their purchases from that same company.

One day when I retire all these crappy P12, P14, F12, F14 still spinning in my rigs, I'll show you what I'm talking about. They sound awful at all of their usable RPMs.
People are delusional, because cheap fans can't sound good if they're moving any decent amount of air. For that you need expensive materials (T30), and expensive QC...
F series sound much better than P, but they're not optimized for pressure, and even a standard honeycomb grille without filters will choke them enough. P series have good pressure, but they produce terrible noise and are resonating in the most annoying ways possible.
Good amount of Arctic fans I bought came with unbalanced blades, it's like a bad joke from technical perspective, and then what to expect... The only thing that performs really well so far are their TP-3 thermal pads, and they are not cheap.
Posted on Reply
#60
freeagent
VeseleilTBH, this whole thread/headline sounds like a bad joke of an Arctic marketing team.

Now compare this to that "hotspot temp reading no more" thread. I swear to God, this site/forum is going places lately... Such extremeness...
Its crazy man, so much fake rage, so much fake. Its everywhere and its gross.
Posted on Reply
#61
Onasi
I feel like we’ve gone full roundabout from the times when people debated various TIMs and their merits in the 00s to “TIM doesn’t matter, use whatever decent stuff you’ve got” of the 2010s and now, with GPUs hitting silly TBPs and some CPU weirdness like the AM5 THICC IHS and Intel cranking the power on their pre-Ultra chips to stupid degrees, we are back again to the same spot. Time truly is a flat circle.
Posted on Reply
#62
freeagent
No.. I had my favourites. With my X5690 ES and running Linpack, the system would be doing 400-500w with just that running, just like my 5900X lol. I prefer a thicker TIM, with a thin application.

I thought the stuff that came with my D14 was terrible, still stands out as some of the worst I have tried lol..

But I am a savage, I will exploit anything on my system that will allow me to. That is how I can get a 5900X to do over 260w with PBO. 300w without lol..
Posted on Reply
Add your own comment
Jan 27th, 2025 00:01 EST change timezone

New Forum Posts

Popular Reviews

Controversial News Posts