New information allows us to clarify a little more Nvidia’s timetable for the launch of its powerful GeForce RTX 5080. We learn in particular that the latter would not arrive on the market before the end of January… despite an announcement that the ‘we know very close.
The RTX 5080 will likely be one of the first GeForce RTX “Blackwell” graphics cards announced in a few days at CES, and it could be the first marketed by Nvidia. However, we learn that a delay of 2 weeks would have been planned between the announcement of the new GeForce, and the actual marketing of the RTX 5080.
This information comes to us from leaker HKEPC, on X, whose remarks are relayed by WCCFTech. The person concerned estimates that the RTX 5080 will not arrive on the market before January 21, 2025, or approximately 15 days after its official presentation, anticipated during the conference that Nvidia will hold in Vegas on January 6.
The RTX 5080 available before the 5090, on January 21?
According to other sources, the firm would also launch its RTX 5080 before the RTX 5090, which will nevertheless appear as flagship for the “Blackwell” range. This inconsistency cannot be verified at this stage. We currently have no clue allowing us to decide on a marketing niche for the RTX 5090.
The latter could as well arrive at the same time as the RTX 5080, or actually be marketed a little later in January. Impossible to know immediately.
Regardless, the technical specifications of the RTX 5080 no longer hold much of a secret. The chip would be based on a GB203 GPU bringing together 10,752 CUDA cores. We would also find 16 GB of VRAM in GDDR7.
The video memory of this RTX 5080 was at 30 Gbps, and would use a 256-bit bus allowing it to display, theoretically, 960 GB/s of memory bandwidth… or 34% more than an RTX 4080 per hour current.
The other side of the coin, the RTX 5080 would be more energy intensive than its predecessor, notably with 80 W more TGP according to the latest rumors.
