Thermal shot of the Palit RTX 5080 Gaming Pro OC. Image source: igorslab.de
The root cause of the problem lies in the design of the graphics card’s power delivery systems: MOSFETs, coils, drivers, and the traces that connect them are placed too close to each other, creating temperatures that can degrade the performance of the power delivery system. As a result, graphics cards can begin to fail within just a few years of use. The printed circuit board (PCB) consists of several thin copper layers connected to the power planes, causing high heat density on the board, especially around the voltage converters. The components responsible for powering the GPU’s rails (including the aforementioned MOSFETs, coils, and drivers) are often placed close together to keep the design compact, but at the expense of thermal performance.
PNY GeForce RTX 5070 OC Thermal Shot
Localized Overheating in NVVDD Area on Palit RTX 5080 Gaming Pro OC
Thermal paste on the back plate
Palit RTX 5080 Gaming Pro OC with thermal paste
On many graphics cards, hot spots are found around the VRM area, which can cause the temperature on the GPU die to rise. A reading of around 80°C is on the verge of long-term electromigration, a conductor-induced ion drift effect that can cause premature failure of the graphics card.