When NVIDIA unveiled the RTX 40 series, there was supposed to be two versions for the RTX 4080 in the form of a 12GB and 16GB variant. Just a little over a month after the announcement, NVIDIA made a surprise move of “unlaunching” the RTX 4080 12GB. The move may sound very strange, but it simply stems from NVIDIA receiving backfire for the poor naming scheme of the RTX 4080 12GB. This is because aside from having less memory, it has lower core counts compared to its 16GB variant.
To recall, the RTX 4080 16GB features 9728 CUDA cores with an AD103 GPU, a 256-bit wide bus, 22.6Gbps GDDR6X memory, and 717 GB/s of bandwidth. The RTX 4080 12GB, on the other hand uses an AD104 GPU with only 7680 CUDA cores, and is paired with 21Gbps GDDR6X operating on a 192-bit wide bus, yielding 504 GB/s of bandwidth. With those differences, NVIDIA admits a 30% performance difference between the two GPUs. This alone justified their move to pull the plug on the RTX 4080 12GB.
With that move, the RTX 4080 16GB will simply be known as RTX 4080, while we’re still waiting of there will be a replacement for the RTX 4080 12GB. While it is possible that the RTX 4080 12GB will be rebranded as an RTX 4070, it’s likely that it may go for an RTX 4070 Ti naming instead. This could also mean a lower price as well–which is a good thing considering the price hike of all RTX 40 series GPUs.
