While today’s highlight was certainly AMD’s RX 6000 refresh, the company also announced plans to launch FidelityFX Super Resolution 2.0 (or FSR 2.0) on May 12th. AMD’s latest version of the upscaler will debut on Death Loop, a critically acclaimed first-person shooter released last year. With FSR 2.0, AMD hopes to catch up with Nvidia’s Deep Learning Super Sampling (or DLSS) scaling solution.
AMD had clarified for some time that Death Loop was going to be one of the first games to feature FSR 2.0, but we expected it to arrive this summer. In fact, news regarding FSR 2.0 was buried in a PR article that focused primarily on the new RX 6000 series refresh. come”, which means Death Loop will probably be the only FSR 2.0 compatible game for a while.
AMD has announced several more games that will receive FSR 2.0 support. Here is the full list:
- EVE online
- Farming Simulator 22
- Microsoft Flight Simulator
- Perfect World Remake
- Swordsman Remake
- Unknown 9: Awakening
It is crucial for AMD to close the gap with Nvidia when it comes to scaling technology. Nvidia launched DLSS in late 2018 alongside its RTX 20-series GPUs, and AMD didn’t release the first version of FSR until mid-2021. AMD in particular needed good scaling technology because its GPUs have a harder time ray tracing than Nvidia’s.
Although FSR 1.0 was a step in the right direction, it wasn’t enough to bring AMD to parity with Nvidia. However, FSR 2.0 might change that. One of the main improvements brought by the 2.0 update concerns temporal or time-based data. Combined with better anti-aliasing, it could provide image quality similar to DLSS.
Nvidia’s DLSS is based on machine learning, and while this can have a positive effect on image quality, it severely limits support. Only Nvidia RTX 20 and 30 series GPUs can use DLSS. In contrast, FSR can run on most GPUs made since 2017 (and even some since 2015 like the GTX 1070). AMD hopes its more open-source and supported solution will be more appealing to gamers and developers, even if FSR 2.0 can’t match the image quality of DLSS.