NVIDIA Deep Learning Supersampling (DLSS) 2.0 Vs AMD FidelityFX Super Resolution (FSR): What’s the Difference?
With the arrival of AMD’s FidelityFX Super Resolution technology, many are unsurprisingly comparing it to NVIDIA’s Deep Learn Super Sampling technology, and the performance benefits it provides in gaming.
While both play similar roles in enhancing visuals and performance, they both achieve this in different ways.
While both play similar roles in enhancing visuals and performance, they both achieve this in different ways.
NVIDIA’s DLSS: Making Use of Those Tensor Cores
Powered by the Tensor cores found on NVIDIA’s GeForce RTX™ graphics cards, DLSS is an AI-powered rendering technology, tapping into a deep learning neutral network that results in boosted frame rates and sharp images. Often, this results in both better visuals and better frame rates; it’s common to see fluid 4K upscaling taking place, with performance metrics that are expected of a 1080p-targeted system. Recently, NVIDIA announced further support for games, adding the likes of Metro Exodus: Enhanced Edition, No Man’s Sky VR along with other VR titles, and Red Dead Redemption II.AMD FSR: Opening the Door to Better Performance
AMD makes their suite of FidelityFX technologies available to all via GPUopen.com, and their latest invention, Super Resolution, is part of that suite. Much like DLSS, FSR’s goal is both improve performance while not impacting or even improving visual fidelity. AMD’s approach however, is software-based. This means that unlike DLSS, which requires those all-important Tensor cores, FSR requires no such thing. This is apparent when you see just how far back the company plans to support legacy graphics products, supporting the 500 series GPUs, which released back in 2017. Additionally, FSR will also be available to use on NVIDIA GPUs, supporting the 10, 20, and latest 30 series graphics, and even AMD Ryzen™ APUs. In total, FSR will support more than 100 graphics products. Available on the 22nd of June, FSR will support over 10 game studios and engines throughout the rest of 2021, with more to follow. The Main Differences Between the Two With both technologies sharing significant similarities, it’s important to measure the differences: NVIDIA DLSS:- Is hardware-based and requires an RTX card with Tensor cores in order to work.
- Is available only to NVIDIA RTX card users.
- Currently supports more than 130 games and apps including Adobe’s Photoshop and Premiere Pro.
- Is software-based and requires no specific hardware technology to work.
- Is available to NVIDIA users.
- Supports older generation graphics, up to and including the AMD RX 500 series and NVIDIA 10 series.
- Currently supports 10+ studios and engines.





