The next Escape from Tarkov patch is set to wipe your account, but it does everything it should, adding AMD FSR to the mix. It’s much more affordable for your gaming PC than Nvidia DLSS, but Deep Learning Super Sampling probably has an edge considering we’re only getting AMD’s first-gen FSR and not FSR 2.0.
Three months after developer Battlestate Games finished implementing Nvidia DLSS, FidelityFX Super Resolution (FSR) is coming to the game. This gives those using older GPUs like the Radeon RX 400 series or GeForce GTX 1000 series the opportunity to increase the frame rate. . Also, DLSS is only available to those who own a GeForce RTX graphics card.
It comes with four modes that do what they say on the tin, including ultra, quality, balanced, and performance options.
AMD FSR Parameters
Scalers are designed to reduce the load on your hardware by rendering an image to the screen at a lower resolution before scaling it to a number of pixels of your choosing. With this in mind, the FSR EFT modes have certain limitations, such as Ultra High Quality, Quality, and Balance, which are only available when gaming at 1080p or higher, with performance reserved for 4K.
Battlestate Games reveals that you can expect a frame rate increase of around 41% when using the AMD FSR performance preset, which is pretty impressive, though this is a select sample that you should take with a pinch of salt. We need to see how image quality compares with DLSS before we can judge, as the two methods use very different methods.
Like Nvidia DLSS, AMD FSR 2.0 uses time scaling, and some say enabling it could be even better than native resolution in some games. However, AMD FSR 1.0 uses a much more inconsistent spatial scaling method that does not provide as high image quality.
When it comes to demanding games like EFT, anything is better than nothing, but it’s a shame there’s no sign of AMD FSR 2.0 yet. Fortunately, the Cyberpunk 2077 AMD FSR 2.0 mod offers hope that the community can implement it even if Battlestate Games doesn’t.
Source : PC Gamesn
