Microsoft is working on AI-based game upscaling technology

General
Microsoft is working on AI-based game upscaling technology

Resolution upscaling used to be a taboo subject only spoken of in the dim recesses of the gaming community. Today, companies are all working on it: Nvidia's Deep Learning Super Sampling (DLSS), AMD's FidelityFX Super Resolution (FSR), and Epic's Temporal Super Resolution (TSR). And now Microsoft is dipping its hand into the Fidelity pot.

Microsoft has realized that for next-generation games to excel in both graphic fidelity and popularity, they need the ability to light up 4K TVs without being locked into low frame rates. *The ability to enjoy games on 4K TVs without being locked into low frame rates. As such, users of the Xbox Series X and S consoles may soon get a taste of their own high-resolution upscaling solution in the form of machine learning-based software (via Hardware Times).

This does not mean that PC users will not benefit from this technology at some point, but as with DirectStorage, the Xbox may be first in line.

Microsoft is preparing to fill the Xbox team with talented people. They are looking for a senior principal software engineer specializing in computer graphics and machine learning so that the dream can become a reality and, of course, beat Sony to the punch.

"Xbox is leveraging machine learning to make traditional rendering algorithms more efficient and provide better alternatives. The Xbox graphics team is looking for engineers to implement machine learning algorithms in their graphics software to delight millions of gamers. They will work closely with partners to develop software for future machine learning hardware. Make a big impact on computer graphics and gaming.

By leveraging Microsoft's DirectML API, the team will have access to a treasure trove of machine learning wisdom and the chance to make console games look great in high resolution. This could emerge as a DLSS-like solution, or it could lead to a complete overhaul of the process and create something entirely new.

Categories