AMD’s FidelityFX Super Resolution (FSR) conquered the world when it launched last summer; more than 80 games (including those on the eSports scene, subsequently on the betting scene at the best Neosurf bookmakers) have already announced support for performance-enhancing technology, thanks to easy plug-in and far-reaching hardware support (running on Nvidia and Intel GPUs, as well as on AMD’s Radeon graphics cards).
But while the simple spatial magnification provided by the FSR makes it universally attractive, the lack of multi-frame time tracking means that its visual quality, while good, is not as good as competing solutions like Nvidia’s DLSS. Still, the novel edition of FSR should be even better than what Nvidia offers.
What exactly is it about, read below…
FSR 2.0 Will Be Better Than FSR 1.0 in Everything
While the original FSR will never deliver better image quality than the original, AMD says FSR 2.0 can sometimes look better than the original rendering, reflecting claims Nvidia makes for DLSS. This is because its time increase in size uses frame color, depth, and motion vectors in the rendering pipeline, and uses information from past frames to create a very high-quality magnified output, according to AMD’s explanation. In other words, it’s pulling information out of multiple frames to create a cleaner magnified image that only simple spatial magnification can provide, similar to DLSS and Intel’s upcoming XeSS.
FSR 2.0, like its competitors, has its own anti-aliasing function. Additional technical hurdles mean that implementing FSR 2.0 requires a little more work for developers, although AMD says it should still be fairly simple if the game already supports DLSS or uses the Unreal Engine’s FSR 2.0 supplement. In these cases, FSR 2.0 can be added in, as noted, in less than 3 days.
FSR 2.0 Has Three Possible Quality Settings
The first generation FSR comes with four different preset qualities: Ultra Quality, Quality, Balanced, and Performance. Each internally rendered the game at a lower default resolution before enlarging the image to your final size (depth of focus is the key to enhancing performance). The better the mode of operation, the higher the basic resolution begins with the SFO.
FSR 2.0 presets include the last three presets of FSR’s first generation. They match the setting names for Nvidia’s DLSS, with a fourth (Ultra Performance) available to developers who want to include a mode designed to offer superior performance while keeping image quality representative of the original rendering.
FSR 2.0 Will Continue to Be Open-Source
Like FSR 1.0, FSR 2.0 remains open-source software, with the ability to run on graphics cards from competing GPU manufacturers. This differs from Radeon Super Resolution. DLSS only works on Nvidia’s RTX GPUs, and Intel’s upcoming XeSS, while open, works best on the company’s Arc graphics cards. This is because both Nvidia and Intel have filled their GPUs with dedicated, customized AI units that aid productivity but necessitate the use of custom firmware.
FSR 2.0 doesn’t require machine learning, which means it will run on a wide range of hardware. Actually, AMD has confirmed that the new generation FSR would be included in the Xbox GDK for game developers to use in their titles.
FSR 2.0 Works Better on Radeon RX 6000 Graphics Cards
Although FSR 2.0 is an open technology, it sounds like it will work best on AMD’s latest and greatest hardware. There are optimizations characteristic for RDNA 2 for the Lanczos algorithms which are based on the tech, according to “PC Gamer“, and FSR 2.0 can sometimes take advantage of Wave64 mode in the RDNA 2 shader.
FSR 2.0 can also rely heavily on memory cache, especially in 4K resolution, making much of the built-in Infinity Cache in Radeon RX 6000 series GPUs doubly practical. However, there are various alternatives and methods to reduce the load on other graphics cards.
