With FSR 2.0, AMD wants to offer an alternative to Nvidia’s DLSS upscaling technology. This is how technology works.
AMD’s FSR 2.0 can finally catch up with Nvidia’s DLSS.
Higher and higher resolutions demand more and more computing power from gaming computers. With upscaling techniques, this effort can be reduced somewhat by rendering the game scenes in a lower resolution and then upscaling them to the desired monitor resolution. With DLSS, Nvidia offers first-class technology for upscaling, and AMD now wants to follow suit with FSR 2.0 (FidelityFX Super Resolution).
“Deathloop” as the first test
The technology was announced in March. The first PC game that supports FSR 2.0 has now been released with “Deathloop”. Our colleagues at Techpowerup.com took a closer look at the alternative to DLSS for AMD graphics cards. Accordingly, little has changed in the basic technology since version 1.0. Compared to the predecessor of FSR 2.0, however, more data from several frames as well as motion vectors and depth information are included. This means that developers have more work than with DLSS, but the image quality should improve significantly. The result is impressive and comparable to DLSS 2.0. In some cases, FSR 2.0 would even achieve better results than Nvidia technology. The comparison between the two technologies is hardly possible with the naked eye. Display errors can only occur with FSR 2.0 for very fine lines such as fences. In addition, DLSS would prevent ghosting a little better.
More effort for developers
What FSR 2.0 is currently missing is a way to render the image with the native resolution and only use the image enhancement for anti-aliasing. This is possible with DLSS. It is to be hoped that AMD will provide this function later. It remains open, however, whether the game developers will shy away from the greater effort for FSR 2.0. According to AMD, the additional work is almost negligible for games that already use DLSS 2.0.
FSR 2.0: New Radeon driver with significantly better performance