We’ve known for a while that AMD has been working on RSR (Radeon Super Resolution), since CES 2022 in fact. Now the time has finally come for release. AMD announced that RSR will launch today with the latest 22.3.1 drivers for Radeon RX 5000 Series graphics cards and newer.
That’s not all though, as AMD also announced that FSR 2.0 (FidelityFX Super Resolution) will arrive in Q2 2022. This is a little surprising, as many expected AMD to release the successor to FSR 1.0 later in the year. Seeing that the implementation method for FSR 2.0 functions differently than the original, it’ll be interesting to see how it stacks up to other upscaling solutions when it releases.
RSR goes places FSR can’t
AMD’s Radeon Super Resolution utilizes the FSR 1.0 algorithm. And while that may seem strange with FSR 2.0 right around the corner, there’s still the chance that AMD could upgrade it in the future. The key to RSR is that it uses in-driver spatial upscaling through Radeon graphics cards. This allows the upscaler to work across almost every game, instead of a select few that require support from the developer. It’s worth noting that RSR only works on RX 5000 Series and future AMD graphics cards.
RSR’s major advantage over other upscaling solutions is that it has widespread support for thousands of games. However, the more general application means that performance and visual quality may not be as good as FSR. This will largely depend on the game though, and results may vary from title to title. Radeon Super Resolution should still be a good solution for games that don’t support FSR but require some kind of performance boost. Radeon Super Resolution is available to use today via AMD Software: Adrenalin Edition.
FSR 2.0 is on the way
AMD’s FSR 2.0 is arguably the more exciting thing to discuss, and that’s for good reason. This image upscaler from AMD directly competes against Nvidia’s DLSS, and soon Intel’s XeSS as well. There’s also the fact that FSR and DLSS work in completely different ways to accomplish the same sort of goal. And while FSR 1.0 was certainly a good step forward for AMD, its next iteration could be even better.
The biggest change coming to FSR 2.0 is that uses a temporal upscaler instead of spatial scaling. Hopefully, this means that image quality will improve alongside performance. AMD’s will also forego using machine learning algorithms again for FSR 2.0, unlike Nvidia’s DLSS. Beyond that, AMD hasn’t shared much more about what we can expect from FSR 2.0. However, the company did release a video showing off its capabilities with 2021’s Deathloop.
While we can’t say for sure how big of an upgrade FSR 2.0 will be over its earlier iteration, the hope is that it will produce less blur/smudging at lower resolutions and quality settings. We won’t need to wait too much longer either way, since Q2 2022 is right around the corner. Plus, AMD has a GDC presentation on March 23 where we’re sure to learn some more information. It’s certainly a busy time for AMD since it recently announced its expanded Ryzen 5000 Series lineup as well.
