NVIDIA recently unveiled two models of its GeForce RTX 4080 graphics card, a 12GB model and a 16GB model. While both GPUs are great, we don’t think NVIDIA got the name of the lower-spec model right.
The 12GB version of the GeForce RTX 4080 has a lower clock rate than the 16GB model, fewer CUDA cores, and a slower memory bus. As a result, NVIDIA is not releasing it anytime soon. The company may decide to launch a new, higher-end model later, or rename the existing 12GB card.
While the 16GB model had more than twice as many CUDA cores, the 12GB model is considered “lazy” compared to its more powerful counterpart. The 12GB version had less memory than the 16GB model, and had a narrower 192-bit memory bus. Moreover, it’s 30 percent slower than the 16GB version.
The 16GB version of the RTX 4080 GPU will be available on November 16. As for the 12GB version, it will not be available at this time. It is still unclear what Nvidia plans to do with the thousands of ready-to-ship units.
NVIDIA has acknowledged that the GeForce RTX 4080 12 GB GPU is not the correct name. They are confusing consumers by having two different graphics cards with the same 4080 designation. According to the company’s blog post, the 12GB version of the card has less CUDA cores and a slower memory interface, but still retains its name. As such, it will probably be renamed to GeForce RTX 4070 or something similar.