Last June AMD responded to NVIDIA and its DLSS with an upscaling technique renamed FidelityFX Super Resolution, FSR for short. In a nutshell (but if you want to learn more you can read this article) FSR uses spatial upscaling technologies in order to guarantee the performance improvement while preserving high image quality.
DLSS leverages the Tensor cores of the GeForce RTX 2000 and 3000 series GPUs to perform superior upscaling by leveraging a combination of motion vectors, temporal feedback (i.e. data from previous frames) and an advanced AI algorithm. The FidelityFX Super Resolution, on the other hand, it does not rely on hardware units or artificial intelligencehence a more conventional upscaling algorithm.
Its strengths are that easily integrated into games and agnostic, that is, it can also work on competing GPUs and even on next-generation consoles based on AMD-designed SoCs. Quite effective as we have seen in several articles, especially in the higher quality modes, FSR is ready to take a new step forward.
During the second quarter AMD intends to introduce FidelityFX Super Resolution 2.0 (FSR 2.0), a major improvement over the first version. AMD promises not only to maintain the agnostic character of the FSR, but also to provide a “image quality similar to or better than native using temporal data and including optimized anti-aliasing“.
AMD promises better image quality than FSR 1.0 at “all resolutions and quality settings” by adding the time data found in NVIDIA’s DLSS to achieve a more detailed and less imperfect image. The company will talk about FSR 2.0 at the Game Developers Conference2022, the conference that brings together those involved in graphics and video games.