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.
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.
50 Comments on RTX 5090 FE Liquid Metal Swap: Thermal Paste Performs Just Fine
I have seen this statement many times from CPU delidders over at OCN that use LM and state it needs re-applying after several months.
I have also seen it happen on a 1080Ti FE that I have personally done. I opened it up and had to change it to a decent paste as the LM seemed to have dried up.
Me personally from what I have seen and tested will probably never use LM again. Paste is also way more practical to apply and clean up.
That risk just to have .5% better temperature
I think those who have issues with thermal paste that needs reapplied after a few months are doing something wrong or using the wrong product. I have build hundreds of computers over almost 30 years and never had an issue with machines in service even after 10 years, my own old machine was under the same cooling loop without changing the coolant, thermal paste for almost 8 years.
Red Devils even from Powercolor had the same issue. Paste leaks out and hotspot rises to 110.
ignoring trouble with application etc, as long as you do it a few times within a year, its fine after 3rd application.
its mainly a problem with new HS that never had it, and i remember a few brands/reviews recommending to apply/remove/reapply it ~3 times so it can saturate.
one the first cpu i used it i had the same problem, on the next one i did as i mentioned and it was still working without trouble almost 2y (sold cpu), and even block removal was easier (vs 1st time).
I used Céramique on my GPU since it was bare die when I installed my waterblock and the cards before that too, and I tried it on the CPU and MX4 worked better.
If they sell a million 7900XTX cards and 1000 users have an issue with improperly applied or defective thermal paste I wouldn't be surprised.
Don't know if it's because it was designed with MCDs, or because of the fact that you can run it at 460W with some cards.
PTM simply is better than a thermal paste so if I'm spending 1000$ or more on a GPU I would expect the card to use it, instead of thermal paste.
We never had GPUs with such high wattages, dual cards like 690 had 300W and that was spread to dual chips. So for new cards with those kinds of TDPs its better to use PTM.
LM makes a bigger difference in cases where the former is an issue eg. on Ryzens where the die is too small to transfer heat to the cooler which itself has headroom left for dissipating more watts.
I'm guessing PTM will perform close enough to LM in this case considering how a paste performs.
Also keep in mind they is A LOT of counterfeit PTM. Helios even, well regarded generally, is a clone and not real honeywell PTM. The clones vary from excellent to very bad.