NVIDIA's Unlaunched GeForce RTX 4080 12GB Is Allegedly Returning With A New Name

GeForce RTX 4080 that's been Photoshopped to say "4070 Ti"
NVIDIA ultimately ended up de-launching its GeForce RTX 4080 with 12GB of GDDR6X memory after outspoken enthusiasts complained about the specs disparity between it and the burlier 16GB model. While the scaled down card might be gone, it's not forgotten. It also might be making a comeback under a completely different model name: GeForce RTX 4070 Ti.

That's according to Twitter user @kopite7kimi, who is well-known for posting accurate GPU leaks (as well as shooting down leaks by others that miss the mark). You still have to take this with a pinch of salt because it's not official information, but for what it's worth, the leaker feels confident that the GeForce RTX 4080 12GB is headed for a rebranding rather than falling by the wayside completely.

Tweet claiming NVIDIA's canceled GeForce RTX 4080 12GB card will return as a GeForce RTX 4070 Ti.

"The original RTX 4080 12GB will become RTX 4070 Ti instead," @kopite7kimi wrote on Twitter. No pricing information was given, though the leakers suggested in a follow-up tweet that it will launch in January, 2023.

In case you missed it, NVIDIA drew some ire during its Ada Lovelace launch for announcing two GeForce RTX 4080 cards with very different specs. Here's how they compare (along with the GeForce RTX 4090 that was announced at the same time)...
  • GeForce RTX 4090 (AD102): 16,384 CUDA cores, 24GB GDDR6X, 384-bit bus
  • GeForce RTX 4080  16GB (AD103): 9,728 CUDA cores, 16GB GDDR6X, 256-bit bus
  • GeForce RTX 4080 12GB (AD104): 7,680 CUDA cores, 12GB GDDR6X, 192-bit bus
Had the only differences been the amount of memory and clocks, most people would not have batted an eye at NVIDIA launching two separate GeForce RTX 4080 cards. But the 12GB model sports a different GPU with fewer CUDA cores, slower memory (21Gbps versus 23Gbps), and a narrower memory bus. This sparked a backlash, and NVIDIA responded by "unlaunching" the lower-end model.

"The RTX 4080 12GB is a fantastic graphics card, but it’s not named right. Having two GPUs with the 4080 designation is confusing. So, we’re pressing the 'unlaunch' button on the 4080 12GB. The RTX 4080 16GB is amazing and on track to delight gamers everywhere on November 16th. If the lines around the block and enthusiasm for the 4090 is any indication, the reception for the 4080 will be awesome," NVIDIA stated last month.

If the GeForce RTX 4070 Ti rebranding rumor is true, then it will be the first GeForce RTX 40 series graphics card to bear the "Ti" badge (unless an inevitable Ti version of the GeForce RTX 4090 makes a surprise early appearance). According to the leaker, NVIDIA originally planned to launch the GeForce RTX 4080 12GB as the GeForce RTX 4070 Ti instead, before launching and then unlaunching the former model.

Assuming the specs are the same, the GeForce RTX 4070 Ti  will wield 7,680 CUDA cores, 12GB of GDDR6X memory (21Gbps) tied to a 192-bit memory bus for 504.2GB/s of memory bandwidth, 240 TMUs, 80 ROPs, 240 Tensor cores, 60 RT cores, 48MB of L2 cache, a 2,310MHz base clock, and 2,610MHz boost clock. The card will also deliver around 40 TFLOPS of performance, compared to a little shy of 49 TFLOPS on the GeForce RTX 4080.