Intel might use AMD’s FSR with their upcoming Xe-HPG series
At Computex, AMD made big news by announcing its own alternative to NVIDIA DLSS, a resolution upscaling algorithm.
A long-awaited response to NVIDIA DLSS has been announced by Scott Herkelman during AMD’s Computex keynote. This technology does not rely on artificial intelligence or machine learning, but a spatial upscaling technique. Unlike similar super-resolution algorithms, AMD’s FidelityFX SR does not rely on motion vectors or history buffers, which means that performance boost with Tensors alike is not necessary for the technology to work.
AMD FSR will be available not just to the latest GPUs but also to the processors based on older architectures, such as AMD’s Polaris or NVIDIA Pascal. Being a cross-platform technology supported by DirectX12, DirectX11, and Vulkan APIs, it will also work on Intel GPUs.
On Twitter, in a response to Kyle Bennett, Raja Koduri confirmed that Intel is already looking into giving FSR a try on the upcoming Xe-HPG based GPUs. Those are the next-generation DG2 discrete gaming GPUs, Koduri says that Intel will try to align with open approaches, which AMD FSR certainly will be as it will be released under MIT license.
Definitely looking at it – the DL capabilities of Xe HPG architecture do lend to approaches that achieve better quality and performance. We will definitely try to align with open approaches to make ISVs job easier..
— Raja Koduri (@Rajaontheedge) June 2, 2021
So far NVIDIA has made no comments on the new AMD technology, which is understandable as the official reveal will take place on June 22nd. AMD has not confirmed which games will support the technology at launch with an exception of Godfall.
At Computex, AMD demonstrated its competitor’s most popular graphics card on Steam (GeForce GTX 1060) successfully running with FSR enabled. AMD fans were busy analyzing the quality of FSR in action side by side ever since. However, AMD is yet to demonstrate the technology in detail later this month.