An Overview of Deep Learning Super Sampling (DLSS) [MiniTool Wiki]

  • Linkedin
  • Reddit

Quick Navigation :

What Is DLSS

Deep learning super sampling (DLSS) is an image upscaling technology that utilizes artificial intelligence and machine learning to create an image. The image generated by this technology looks like a higher-resolution image without the rendering overhead. To obtain more details about DLSS, keep reading this post of MiniTool.

NVIDIA DLSS is enabled by training a neural network on thousands of images from a specific video game both with and without anti-aliasing enabled. Then the network learns how to improve the image to anti-aliasing quality without requiring any performance overhead.

The trained neural network will be packaged into a video game driver and then run during the game play. At present, DLSS is enabled on NVIDIA RTX20 Series GPU card with a number of video games.

Tip: DLSS is not perfect. It also has some drawbacks. For instance, it can cause blurred textures and some loss of detail. Though the latest DLSS 2.0 version has fixed many issues, you can still encounter them in some cases.

The Development of DLSS

When the GeForce RTX20 series GPUs were released in September 2018, NVIDIA announced DLSS as a vital key feature them. However, DLSS only supported a few video games like Battlefield V and Metro Exodus at that time.

It is because that the algorithm had to be trained specifically on the applied game and the results were not as good as simple resolution upscaling. In 2019, the video game Control came with ray tracing – an improved version of DLSS that didn’t utilize the Tensor Cores.

NVIDIA released a new version called DLSS 2.0 in April 2020. This version supported a few existing games including Control and Wolfenstein: Youngblood. It used the Tensor Cores and the AI didn’t need to be trained specifically on each game. But DLSS 2.0 cannot work well with anti-aliasing techniques like MSAA and TSAA.

What’s more, the performance of DLSS will be affected when the techniques are enabled on top of DLSS. Since April 2020, DLSS 2.0 has become a must element on a per-game basis.

Top recommendation: What Is VSync Technology and Whether You Should Use It

The Working Theory of DLSS

NVIDIA DLSS forces games to render at a low resolution (usually 1440p) and utilize its trained artificial intelligence algorithm to infer what it would look like if it were rendered at a higher one (typically 4K).

DLSS does this by making using of some anti-aliasing effects and some automated sharpening. It can also eliminate visual artifacts that may not be displayed at higher resolutions and even infer details that should be present in the image.

Artificial intelligence algorithm is trained to look at certain games at extremely high resolutions (64x super-sampling). Before being added to the latest NVIDIA driver and becoming available to all the games, it is distilled down to something having a size of a few megabytes.

Previously, NVIDIA must go through this process game by game. With the existence of DLSS 2.0, A.I. model doesn’t need to be trained for each game.

The Function of DLSS

DLSS is the final result of NVIDIA’s A.I. algorithm detailed teaching process that is aimed at generating better-looking games. After rendering the game at a lower resolution, DLSS concludes information from its knowledge of super-resolution image training to produce an image that looks like it was running at a higher resolution.

The purpose of this operation is to make games rendered at 1440p look like they’re running at 4k, or 1080p games to look like 1440p. With DLSS 2.0, you can render games at 1080p while outputting them at 4k. Some conventional super-resolution techniques can cause artifacts and bugs in the eventual picture. NVIDIA DLSS is built to handle those errors to generate a better-looking image.

Based on that, DLSS NVIDIA can also deliver substantial performance uplifts without affecting the look and feel of a game. Now, DLSS supports many games including Anthem, Battlefield V, Bright Memory, Call of Duty: Black Op Cold War, Control, Cyberpunk 2077, Death Stranding, Deliver Us The Moon, Fortnite, Justice, Watch Dogs Legion, etc.

According to investigation, more and more games will be supported by DLSS in the soon future like Amid Evil, Atomic Heart, Boundary, Call Of Duty: Warzone, Edge Of Eternity, and so forth.

What is DLSS? You may have learned the answer from the post now. Here comes the end of the post!

  • Linkedin
  • Reddit