A GeForce RTX 5090 AORUS owner says they opened their graphics card after more than a year of use and found a large amount of factory-applied thermal material inside the cooler. The claim has not been independently verified, but the images and description are likely to draw attention from high-end GPU buyers watching for cooling issues on expensive RTX 50-series cards.
The owner said the card had been mounted horizontally, a setup that should, in theory, reduce the chance of thermal material shifting under gravity compared with a vertical mount. Even so, the user claimed the material had pooled heavily across part of the vapor chamber, while the lower half of the GPU core appeared to have little or no coverage after disassembly.
What the owner claims they found
The Reddit user, posting under the name Correct_Difference72, described the factory compound as a liquid-metal-style composite that appeared susceptible to pump-out, though that diagnosis has not been independently confirmed. Pump-out is a known thermal-interface concern in PC hardware, where repeated heating and cooling cycles can gradually push material away from the hottest contact area.
In this case, the owner claimed the thermal material had collected on the upper half of the vapor chamber and left visible discoloration where it pooled. They also said the GPU temperature they normally saw did not rise above 76 degrees Celsius, but fan speeds were allegedly jumping to around 2,000 RPM during games. Those temperature and fan readings are user-reported and were not backed by a full log of hotspot or memory junction temperatures.
That missing detail matters. GPU edge temperature is only one part of the thermal picture on modern graphics cards. Hotspot temperature, memory junction temperature, fan curve behavior, case airflow, workload, BIOS mode, ambient room temperature, and monitoring software can all change how serious a cooling issue looks. Without before-and-after measurements, it is difficult to say whether the material pattern directly caused the faster fan behavior.
Why the thermal material question matters
AORUS RTX 5090 cards sit in the most expensive tier of consumer graphics hardware, so even a single teardown claim can create concern among owners. The practical question is not just whether the inside of the cooler looks messy. It is whether the factory thermal interface is still covering the GPU die well enough to move heat into the vapor chamber under sustained load.
The owner said they planned to replace the original material with PTM7950, a phase-change thermal pad often discussed in GPU modding and laptop repair communities. That replacement plan has not been independently verified either, and opening a graphics card can carry warranty, damage, and reassembly risks. On a card as costly as an RTX 5090, even a routine repaste is not a casual job.
There is also a difference between excess material around a contact area and inadequate coverage on the die itself. Extra gel outside the GPU package may look alarming but not necessarily affect performance. Poor contact over the actual GPU core is more serious, especially if it creates uneven pressure or leaves part of the die without a proper thermal path.
Not the first thermal gel complaint around RTX 50 cards
This is not the first time Gigabyte RTX 50-series cards have been discussed in connection with thermal gel application. Earlier user complaints focused on excess material, including concerns that gel could move when large cards were mounted vertically. Claims that gel could reach PCIe slots or nearby motherboard areas have circulated, but confirmed damage from those reports has not been established in the material available here.
Gigabyte has previously been linked to statements framing visible differences in factory-applied gel as cosmetic variance, while maintaining that the material should not damage components. Because the latest teardown claim is based on a user account, it should be treated as a warning sign rather than proof of a broad defect across all AORUS RTX 5090 cards.
Still, the story lands in an uncomfortable place for premium GPU makers. Flagship cards are heavier, hotter, more complex, and more expensive than ever, and buyers increasingly expect factory cooling assemblies to hold up for years without requiring inspection or rework. When users see thermal material outside the expected contact zone, it raises a simple question: is the cooler merely untidy inside, or is it slowly losing effective contact where it matters?
Owners should be careful before opening a card
The most useful next step for owners is measurement, not immediate disassembly. Anyone worried about an AORUS RTX 5090 should check normal GPU temperature, hotspot temperature, memory junction readings where available, fan speed, and noise under repeatable workloads before deciding anything is wrong. A sudden change in fan behavior can come from thermal interface problems, but it can also come from dust buildup, case airflow changes, driver behavior, room temperature, or a more aggressive fan profile.
Opening a GPU cooler can damage pads, disturb factory pressure, strip screws, or complicate warranty support. It may also make a minor cosmetic issue worse if the cooler is not reassembled with the correct pad thickness and mounting pressure. For users still inside warranty coverage, contacting the board maker or retailer remains the lower-risk path.
The teardown claim is worth watching because it involves a high-end AORUS RTX 5090 and a specific allegation: factory-applied thermal material may have shifted away from part of the GPU contact area after long-term use. What it does not prove is that every similar card has the same issue, that the fan behavior was definitely caused by the gel pattern, or that owners should immediately open working cards.
For now, the practical takeaway is narrower: unusual fan speeds on a premium GPU deserve proper thermal logging, and any visible thermal material concerns should be handled carefully before turning a suspected cooling issue into a much more expensive repair.