NVidia's "Deep Learning Super Sampling," also known as DLSS, is already a great feature when it comes to improving image quality. Now, the company has improved DLSS even further. In other words, version 2.0 has been released.
DLSS 2.0 runs on RTX cards and uses a new AI model and machine learning to improve game resolution. According to Nvidia, the new AI model in DLSS 2.0 is twice as fast as the previous version.
“The new AI model uses the Tensor Cores even more efficiently, resulting in them running 2x faster than the original. In this way, the new DLSS improves the performance of the card,” says Matt Wuebbling, Nvidia’s head of marketing.
DLSS works by upscaling images to high-resolution ones using machine learning. Until now, Nvidia had to “train” the AI for each game that supported DLSS, a time-consuming process.
However, with version 2.0, things are easier. Now, Nvidia doesn't have to "train" A for each game individually. From now on, the new algorithm will work automatically on supported games, allowing more companies to use the technology in their games.
