NVIDIA Deep Learning Super Sampling is a deep learning technique developed by NVIDIA that uses AI to intelligently upscale rendered frames. This technology which must be individually added game by game allows for significant performance boosts by running the game at a lower resolution and upscaling the output to a nearly identical level of detail. This technology relies on the Tensor Cores found in RTX 20/30 series GPUs and allows for a performance boost of up to 70%.
In a recent Q&A thread, NVIDIA announced that the new DLSS 2.1 will bring VR compatibility along with 8K, and dynamic resolution support. The introduction of DLSS 2.1 will help enable more immersive VR titles on next generation headsets. VR headsets typically expect a consistent frame rate of 90 FPS at resolutions up to 2880x1600 with some headsets such as the Valve Index taking that up to 144hz. DLSS technology will allow VR games to be natively rendered at half resolution and intelligently upscaled to ensure high frame rates and quality.
View at TechPowerUp Main Site
In a recent Q&A thread, NVIDIA announced that the new DLSS 2.1 will bring VR compatibility along with 8K, and dynamic resolution support. The introduction of DLSS 2.1 will help enable more immersive VR titles on next generation headsets. VR headsets typically expect a consistent frame rate of 90 FPS at resolutions up to 2880x1600 with some headsets such as the Valve Index taking that up to 144hz. DLSS technology will allow VR games to be natively rendered at half resolution and intelligently upscaled to ensure high frame rates and quality.
View at TechPowerUp Main Site