Epic Games: Unreal Engine 4 acquired Nvidia’s AI DLSS 2.0 software

Epic Games provides a new plug-in for Nvidia’s Deep Learning Super Sampling (DLSS) version 2.0 for game developers who rely on licensed Unreal Engine 4. Studios and indie developers can easily integrate upgrade technology into their games without requiring a direct connection with Nvidia.

DLSS 2.0 relies on a neural network trained by Nvidia, which learned from comparing images to display lower-resolution content with better quality. For example, 4K gameplay can be derived from a Full HD image with a resolution of 1920 x 1080 pixels. A time filter that counts frames together also increases image quality.

Year Implementation in the Unreal Engine 4 It has been working since updating to DLSS 2.0, because Nvidia no longer trains a single neural network per game, but is generally applicable with a variety of image content. The Unreal Engine 4 plug-in is a logical consequence.

For GeForce RTX graphics cards

DLSS image quality doesn’t always come close to the original resolution, but it is a welcome way to increase the frame rate, sometimes drastically, without sacrificing image quality. Prerequisites for use are the Tensor processing cores in Nvidia’s GeForce graphics cards for the RTX 2000 and RTX 3000 series, which power the Neural Network.

Epic Games introduces the DLSS 2.0 plugin for DirectX 11, DirectX 12, and Vulkan graphics APIs. It also includes the High Performance Mode, which produces 8K images with an internal display resolution of 2560 x 1440 pixels.

Immediate streaming of DLSS titles can’t be expected: Epic Games expansion builds on the Unreal Engine at 4.26 Published in December 2020 And it doesn’t even use all UE4 games currently in development.

(mma)