Nvidia GeForce RTX 5080 leak hints at 16GB and 24GB GDDR7 variants in the works

There are rumors circulating about Nvidia's plan to introduce two editions of its RTX 5080 graphics card. One version will have a 16GB capacity, while the other will boast an impressive 24GB frame buffer.

Recently, leaked specifications for Nvidia’s RTX 5080 have surfaced. These specifications outline a 400W TDP, 10752 CUDA cores, and a 256-bit memory bus. It's important to note that the leak indicates the RTX 5080 will feature 16GB of GDDR7 memory. However, the possibility of a 24GB model should not be ruled out.

These rumors originated from a post by Chiphell forum user PolyMorph, who mentioned that the "5080 will be released in 24G later" (translated from Chinese using Google Translate). While this information has not been officially confirmed by Nvidia, PolyMorph has previously shared credible Nvidia leaks, such as posting a photo of the GeForce RTX 3090 GPU with the accurate "GA102-300-A1" codename prior to its release.

For Nvidia to achieve 16GB of GDDR7 memory, they would need to employ 2GB GDDR7 memory modules over a 256-bit memory bus. To provide users with 24GB of memory, Nvidia could opt for a broader 384-bit memory bus and additional 2GB memory chips. Another option would be to use 3GB GDDR7 memory chips over the same 256-bit memory bus. The latter option makes more sense for an alternative RTX 5080 model.

If Nvidia were to develop a GPU with a wider memory bus, it would no longer be considered an RTX 5080. A 50% wider memory bus would result in 50% more bandwidth for users (assuming memory speeds remain unchanged), effectively making this hypothetical RTX 5080 an entirely new GPU with different silicon. However, it is worth noting that the GDDR7 standard does include 3GB chip options. Yet, it is uncertain whether these 3GB chips are currently in production.

According to reports, Nvidia's rumored 24GB model of the RTX 5080 will be launched subsequent to the initial release of the 16GB version. This strategy makes sense, considering that only 2GB GDDR7 memory modules are currently available. Nvidia could target creators, workstation users, AI enthusiasts, and forward-thinking gamers with this GPU. This move is likely aimed at bridging what is expected to be a significant price gap between the 16GB RTX 5080 and the 32GB RTX 5090.

And as always, while waiting for more exciting news, remember to utilize our comparison tool to discover the best deal for your next favorite game.