Today we have a series of leaks giving us details of the theoretically upcoming GeForce RTX 4080 SUPER, 4070 Ti SUPER and 4070 SUPER technical specifications. These elements are pushed by the very reliable Kopite7kimi, which leads us to consider this information with a little more attention.
RTX 4080 SUPER details (rumor):
The next RTX 4080 SUPER is said to use the AD103-400 GPU. This GPU will use a 256-bit bus, which would suggest 16 GB GDDR6X memory, identical to the current RTX 4080. Does this mean that Nvidia isn’t planning to upgrade its 4080 to 20GB… But the greens are used to testing multiple configurations and unveiling the final version almost at the last minute.
RTX 4070 Ti SUPER details (rumor):
As the leaker notes, a possible RTX 4070 Ti SUPER is a very strange choice of name. But in terms of specifications, the card would feature an AD103-275 / AD102-175 GPU with the same 8,448 cores in both variants and 48 MB of L2 cache.
RTX 4070 SUPER details (rumor):
Finally, the RTX 4070 SUPER is said to be built around an AD104-350 or AD103-175 GPU with the same 7,168 cores in both variants. The GPU should feature 48 MB of L2 cache. For this version, we’re staying with a 12 GB GDDR6X memory configuration.
What are we to make of these details on the future RTX Super?
If we assume that these characteristics are credible, it means that NVIDIA wants to beef up its RTX 4070 segment even further. The SUPER version would be the most significant upgrade, and this could be explained by much stronger competition from the AMD RX 7800 XT and the arrival of Intel. On the other hand, the RTX 4080 SUPER would be a disappointment, as its big sister (the current 4080) has been selling rather poorly since its launch. On this basis, the Super version would be the smallest upgrade, with a 5% increase in the number of cores. The gulf between the 4080 and 4090 series remains, unless Nvidia decides to rebuild its price positioning by adding an RTX 4080 Ti to the range.