GeForce RTX 5090: NVIDIA takes a more compact approach than the previous generation

Written by Guillaume
Publication date: {{ dayjs(1717430445*1000).local().format("L").toString()}}
Follow us
This article is an automatic translation

Year after year, graphics cards have accustomed us to "more and more", and the RTX 5090 may well mark a breakthrough.

In a few weeks, a few months at most, NVIDIA will be presenting its new graphics cards and bringing them to market. Rumors vary as to the exact timetable, but all indications are that the GeForce RTX 5080 and RTX 5090 will be the first in a new series, and the only ones to be launched before the end of 2024. At last count, the GeForce RTX 5080 would be the first to be distributed by NVIDIA, with the RTX 5090 following shortly afterwards.

It may only be second on the launch order, but it's the GeForce RTX 5090 - the future flagship of the new range - that is the talk of the town today, and even if precise data are still scarce, today's information is interesting. Indeed, it seems likely to mark a break in the trend observed in recent years for NVIDIA graphics cards. A few days ago, Overclock3D.net reported that the RTX 5090 would be even more voluminous, and would feature an imposing heatsink to cool the GPU. This heatsink would significantly thicken the card, which would then occupy four PCI-Express slots.

In the wake of this "information", kopite7kimi - one of NVIDIA's most reliable sources for rumours - has come forward to dispel the rumours. Not only is there no mention of a 4-slot PCI-Express heatsink for the GeForce RTX 5090, but "only" a 2-slot PCI-Express model. In other words, the RTX 5090 would be the occasion for a nice slimming down at NVIDIA, as it would be downright thinner than the RTX 4090 and its 3-slot heatsink. Of course, this information applies only to Founders Edition models, designed and sold directly by NVIDIA.