To quickly recap, the RTX 4070 Ti in its current form is actually a rebranding of NVIDIA’s RTX 4080 12GB, a card that the GPU brand tried to low-key “unlaunched“, in response to the backlash from fans, primarily due to the confusing nature of its naming scheme. Prior to the renaming, it planned on launching the RTX 4080 with two different GPUs, the AD103 and AD104. In any case, this is the first time that an image for the mid-range AD104 and more specifically, the full-fat AD104-400, has ever been seen and first impressions suggest that it is at least half the size of the top-tier AD102 GPU. In terms of measurements, that would put the GPU die at around 295mm2 in size.
— MEGAsizeGPU (@Zed__Wang) November 30, 2022 But that also means the AD104 gets the whole package that NVIDIA warrants: we’re talking 60 Streaming Processors (SM) enabled or 7680 CUDA cores, 240 Tensor Cores, and 60 RT Cores. However, because it is a lesser version of both the GPUs of the RTX 4090 and RTX 4080, it also means that it will have a smaller memory bus at 192-bit. Also, to rehash what has already been reported since its launch, the card will have 12GB GDDR6X graphics memory. Unfortunately, the tweet from MEGAsizeGPU (@Zed__Wang) didn’t have any other details accompanying the image of the AD104. However, there really isn’t that much left to discuss about the RTX 4070 Ti and its internals at this point. At least, not until its launch next year and we actually get the card into our labs for the usual in-depth testing and review that comes after. (Source: Twitter, Videocardz)
