Nvidia‘s new RTX 50 graphics cards, from the RTX 5060 Ti to the powerful RTX 5090, may have a major issue affecting their long-term performance. According to an investigation by Igor’s Lab, these GPUs suffer from “hotspots” in their power sections, which can lead to temperatures exceeding 100°C. This issue could harm the cards’ lifespan, especially with heavy use over time.
Critical Temperature Issues in Thermal Tests
Igor’s Lab conducted thermal tests using infrared cameras and found a PNY RTX 5070 reaching temperatures as high as 107.3°C in the VRM (voltage regulator module) area, while the GPU core stayed at around 69.7°C. This large temperature gap shows that the cooling system is not doing enough to manage heat in such a sensitive part of the card. The Palit RTX 5080 Gaming Pro OC also showed a hotspot close to 80°C, despite having better component distribution on the PCB.
The issue? Neither of the two cards used thermal pads on the back to help dissipate heat from the power section. By simply adding conductive material, the temperature of the RTX 5080 dropped from 80.5°C to 70.3°C, and the RTX 5070’s temperature dropped below 95°C. These improvements highlight how poorly the thermal design was thought out in these GPUs.
A Cheap Design Hurting Durability
The problem seems to stem from Nvidia’s design limits, known as the “Thermal Design Guide,” which provides guidelines based on ideal conditions. In real-world use, these conditions don’t apply, meaning the cooling of the VRMs was not prioritized during the design process.
Igor’s Lab suggests that more heat-resistant materials, like those used in server or industrial GPUs, would be a better choice. However, these materials are costly, making them impractical for consumer-grade GPUs. As a result, Nvidia’s design compromises may hurt the cards’ durability in the long run.
