GamingGPUsNews

Microsoft Working on Machine Learning-Based Rendering and Upscaling Techniques for Xbox

Microsoft is looking to leverage its DirectML API to improve the performance of next-gen games on its Xbox Series X|S consoles to cement its technical superiority against rival Sony. As demonstrated by Epic, NVIDIA, and AMD, cutting-edge upscaling techniques rather than sheer compute performance are going to be key to improving the visual fidelity of future titles. NVIDIA’s DLSS, Unreal Engine 5’s TSR, and (to an extent) AMD’s FSR are allowing gamers to experience their favorite worlds at incredible poly-counts with resolutions as high as 8K, all the while getting 60 or more FPS across the board.

In this regard, the Xbox team is looking to hire engineers well-versed in computer graphics and machine learning to explore ways to improve traditional rendering pipelines using convolution neural networks and other smart reconstruction algorithms.

Microsoft is looking into improving existing rendering solutions by leveraging DLSS-like upscaling techniques or by overhauling the entire pipeline to take greater advantage of machine learning, the likes of which we’ve only seen in research papers.

Ideal candidates would have at least 5 years experience in software development and some experience in machine learning or data science, as well as preferably experience in compute graphics and applied research in machine learning. Please use the following links based on your level of experience:

DirectML is a high-performance, hardware-accelerated DirectX 12 library for machine learning. DirectML provides GPU acceleration for common machine learning tasks across a broad range of supported hardware and drivers, including all DirectX 12-capable GPUs from vendors such as AMD, Intel, NVIDIA, and Qualcomm.

Areej

Computer Engineering dropout (3 years), writer, journalist, and amateur poet. I started my first technology blog, Techquila while in college to address my hardware passion. Although largely successful, it was a classic example of too many people trying out multiple different things but getting nothing done. Left in late 2019 and been working on Hardware Times ever since.
Back to top button