Nvidia has updated its Deep Learning Super Sampling (DLSS) technology. This technology is only available to owners of RTX GeForce graphics cards to make the critical Tensor core do the hard work, but this genuinely seems like an option that is finally worth turning on.

There are several behind-the-scenes technologies driving DLSS 2.0, including improved AI networks that are 2x faster, improved target resolution with additional temporal feedback, and a move to using one model for all games. This last point is important and should encourage more developers to use DLSS 2.0. DLSS 2.0 still needs access to buffers, so it will need to be integrated on a game-by-game basis rather than magically working for all games, but it is still definitely a step in the right direction.

The first game to receive the new treatment is MechWarrior 5: Mercenaries, where DLSS 2.0 has the dual effect of improving image quality and performance. This game showed that using DLSS 2.0 improves rendering of fine detail over native rendering. By the time you read this, new drivers specifically for MechWarrior 5: Mercenaries will be available from Nvidia.

As for DLSS, there are problems with non-deterministic effects (smoke, sparks, etc.). Those who have used DLSS with controls may have witnessed large fans spinning behind meshes in certain sections, creating horrible artifacts. These problems seem to have been resolved in DLSS 2.0, so hopefully DLSS will be easier to use in the future.

Aside from the fact that there is now a single model for all games, there is also a new SDK that takes care of much of the heavy lifting, which should make it much easier for developers to add to their titles than the first generation technology. Nvidia DLSS 2.0 essentially gives players a performance boost at no cost to them, without sacrificing image quality.

"Nvidia DLSS 2.0 essentially gives players a performance boost at no cost to them, without sacrificing image quality.