AMD hints at improving scaling with AI technology – hardware

AMD, Intel, and Nvidia each offer their own upscaling technologies using FidelityFX Supersolution (FSR), Xe Super Sampling (XeSS), and Deep Learning Super Sampling (DLSS). Microsoft is now also working on installing those with the new Windows API more directly into the system and thus making the work of game developers easier. Both Intel and Nvidia are working with hardware to drive scaling by supporting AI. And that's exactly what AMD appears to be planning for 2024.

Mark Peppermaster, AMD's chief technical officer (CTO), explicitly mentioned in an interview that his employer also wanted to use the AI ​​component of gaming to raise the bar in gaming this year. Although it doesn't explicitly mention FSR, it stands to reason that this is exactly where the upgrade should take place. Nvidia also uses dedicated Tensor cores for this purpose, which is why DLSS can also work with machine learning and the like.

With DLSS, Nvidia is also several years ahead of AMD, which it is now trying to catch up with. This will certainly be welcomed by Radeon owners. Thanks to the integration of AI, artifacts can be reduced during upscaling and the quality may slowly approach DLSS.

source:
No firsts (YouTube)