Users with an Nvidia GeForce RTX Series are in luck, because the second generation of Deep Learning Super Sampling technology, known as DLSS 2.0 will be natively integrated into the Unreal Engine graphics engine from version 4.26, and this translates into promises of a great performance improvement when activating such technology.
Tom Looman, a video game developer working with the Unreal Engine 4, demonstrated how the DLSS 2.0 integration process works and the huge performance improvements it brings. In the blog post, he revealed that the integration with Unreal Engine 4.26 is simple, since it only requires you compile your project with a special branch of UE4 RTX, and after that you apply your AppID request on the website of Nvidia. After this, the DLSS 2.0 manages to increase the frame rate between 60 and 180% depending on the scene, now this is some serious problem for AMD.
“Last year Nvidia announced DLSS 2.0 with the ability to open this new antialiasing solution for all games. Game-specific Deep Learning is no longer required as was the case with previous iterations. For an indie developer this is especially exciting , since the possibility of entering the Nvidia supercomputing was quite slim.
In addition to the new game-agnostic algorithm, we also get greatly improved antialiasing results that honestly look pretty great and somewhat amazing in games like Control (see Digital Foundry’s DLSS comparison). You can see how DLSS works on the Unreal Engine below through my own experiments.
For testing, I used my open source SurvivalGame (available on GitHub) and the City Subway Train asset from Dekogon Studios. “