24 GB of memory and higher power limits could be features of the NVIDIA GeForce RTX 5080 SUPER.

Rumored specifications of the GeForce RTX 5080 SUPER surfaced on VideoCardz shortly after those of the RTX 5070 SUPER and RTX 5070 Ti SUPER. It appears that even though NVIDIA maximized the “GB203” silicon to create the current RTX 5080, it will not use the larger “GB202” silicon to create the RTX 5080 SUPER. Giving the card more memory and power limits will be a slightly different strategy. Similar to the RTX 5070 Ti SUPER, the RTX 5080 SUPER will have 24 GB of GDDR7 memory spread across a 256-bit wide memory bus. 24 Gbit GDDR7 memory chips are a unique characteristic of RTX 50-series SUPER graphics cards.

In order for this SKU to have at least the same amount of bandwidth as the standard RTX 5080, which is 960 GB/s, NVIDIA may be able to use 30 Gbps memory speeds. Another intriguing feature of the RTX 5080 SUPER is anticipated to be its 15% boost in TGP (total graphics power) to 415 W over the RTX 5080’s 360 W TGP. In addition to supporting the greater density memory chips, this increase in TGP will enable NVIDIA to raise GPU clock speeds. Given that NVIDIA does not have the capacity to increase the number of shaders on “GB203,” this will probably be required in order to boost performance for games that do not require a lot of RAM.