“So, we’re pressing the ‘unlaunch’ button on the 4080 12GB,” said the company in a statement announcing the unlaunch. “The RTX 4080 16GB is amazing and on track to delight gamers everywhere on November 16th.”

While the company hasn’t elaborated any further, it is quite likely that we’ll see the RTX 4080 12GB rebranded as the RTX 4070 and launched at a later date with a different price tag.

The RTX 4080 12GB was announced alongside the RTX 4080 16GB, and the flagship RTX 4090. Part of Nvidia’s new RTX 40-series lineup, the graphics cards make use of the new Ada Lovelace architecture, which is set to be twice as fast in rasterisation and four times faster in ray tracing than the previous generation’s Ampere architecture.

While the RTX 4080 12GB and RTX 4080 16GB bear the same name indicating that the only real difference would be the amount of VRAM the two GPUs would have, the reality is that both GPUs feature different clock speeds, different boost clocks, and different numbers of CUDA cores.

For more details, here’s everything you need to know about Nvidia’s RTX 40-series GPUs.

As for Portal with RTX, it also features high resolution physically based textures which are hand-crafted and high-poly models. Nvidia DLSS 3 is used to enhance performance, while latency is reduced using Nvidia Reflex. The ray tracing applies to every light source, from global illumination indirect lighting to volumetric ray-traced lighting that scatters in fog and smoke with perfect shadows.

It also features some new Nvidia tech, like RTX Direct Illumination (RTXDI) for “countless direct light sources, big and small, each casting light and shadows.” There’s also Reservoir Spatio Temporal Importance Resampling Global Illumination (ReSTIR GI) which “enhances indirect light, enabling it to bathe a scene and illuminate dark corners that are not lit directly.” This is supposed to work “smarter” than past ray-traced global illumination tech.

Check out the comparison trailer below to see it all in action.

The RTX 4080 starts at $900 for 12 GB G6X RAM, which features a boost clock of 2.61 GHz, and 7680 CUDA cores. The 16 GB GDDR6X RAM version costs $1200 with a boost clock of 2.51 GHz, and 9728 CUDA cores. It’s estimated to be two to four times faster than an RTX 3080 Ti, with both variants capable of outputting at 7680×4320 resolution.

The RTX 40 Series is built on a new architecture called Ada Lovelace, which is twice as fast in rasterization and four times faster in ray tracing than Ampere. The GPUs are also more energy efficient.

If that weren’t enough, Nvidia also announced DLSS 3. Along with massive performance increases, it uses AI for generating high-quality frames. A new video has been released, with comparisons for titles like Cyberpunk 2077 and Microsoft Flight Simulator with RTX On and RTX Off, and the differences are staggering.

The Nvidia GeForce RTX 4090 is available from October 12th, while the RTX 4080 is coming in November. Stay tuned for more details in the meantime.