Elysian Realm, aka @KittyYYuko on Twitter, filtered through which aims to be a unit GeForce RTX 4080 Founders Editionthe top model (although not the maximum) of the next generation of NVIDIA graphics: Ada Lovelace.

A photo posted by @KittyYYuko shows the alleged RTX 4080 Founders Edition inside the protective plastic, making it difficult to check some of the product details, although the model name, RTX 4080, is perfectly legible. The type of cooler it will use remains under wraps, but there will apparently be an important change from the Founders Edition RTX 3080.

The design of the RTX 4080 Founders Edition aims to be dual slot and 7-blade fan would be bigger than the one included in the same RTX 3080 and RTX 3090 versions, whose respective fans have nine blades. Fortunately, the Founders Edition left that reputation very hot in a few generations thanks to the inclusion of more competent dispersion mechanisms from NVIDIA.

NVIDIA GeForce RTX 4080 Founders Edition.

At the feature level, the RTX 4080 is expected to use a “reduced” version of the AD103-300 GPU with 9728 CUDA cores, 48 ​​MB L2 cache and that it would be based on the PG136/139-SKU360 PCB. This is in contrast to the full version of the processor, which supports 10,240 CUDA cores and 64 MB of L2 cache.

Other features of the expected final version of the RTX 4080 are 16 GB GDDR6X memory, 23 Gbps speed and 256-bit bus interface. Bandwidth is expected to be 736 GB/s, which would be slightly less than the RTX 3080’s 760 GB/s, which uses a 320-bit bus interface and 10 GB of VRAM. TBP should now be 340 watts.

Both the RTX 4080 and RTX 4070 could be among the first Ada Lovelace graphics cards to hit the market after or alongside the RTX 4090, which is slated to launch in October 2022. We’ll see how NVIDIA embraces the next generation by looking at the economic context in which we live, but on the other hand, on some fronts and in the long term, the challenge is to take advantage of the hardware rather than to increase its performance.