Post by KostaAndreadis @ 03:34pm 24/03/20 | 0 Comments
With no loss in picture quality. In fact, in some cases clarity is actually improved over the native resolution. Yeah, this is insane and a game-changer for the RTX line. But first, a little bit about DLSS -- the AI rendering technique available in certain games. DLSS leverages the power of NVIDIA GeForce RTX via AI deep learning as a method to upscale lower resolution images and present them in 1080p, 1440p, or 4K.
The first version of DLSS rolled out over a year ago, with the AI process involving running individual titles through a super-computer and then utilising the AI Tensor cores on RTX cards to render the final DLSS image. With the goal being to present a picture that was identical to rendering in a native resolution, with the benefit of a sizeable 30-40% in increased performance.
The only problem was that in most cases the image displayed was noticeably softer than rendering in a native resolution. But, the performance gains were spot on.
DLSS 2.0 as seen in Wolfenstein: Youngblood and Deliver Us The Moon builds on over a year of AI expertise and does away with the need to run each individual game through giant AI computers. Here a single AI model developed by NVIDIA offers a broader range of DLSS options with the end results being identical and sometimes even crisper than native resolution.
And we're not talking about simply turning up sharpness settings. We're talking Skynet let's hope the AI doesn't turn on us.
Superior Image Quality - DLSS 2.0 offers image quality comparable to native resolution while rendering only one quarter to one half of the pixels. It employs new temporal feedback techniques for sharper image details and improved stability from frame to frame.
Great Scaling Across All GeForce RTX GPUs and Resolutions - A new AI network more efficiently uses Tensor Cores to execute 2X faster than the original. This improves frame rates and eliminates previous limitations on which GPUs, settings, and resolutions could be enabled.
One Network For All Games - The original DLSS required training the AI network for each new game. DLSS 2.0 trains using non-game-specific content, delivering a generalized network that works across games. This means faster game integrations, and ultimately more DLSS games.
Customizable Options - DLSS 2.0 offers users 3 image quality modes - Quality, Balanced, Performance - that control the game’s internal rendering resolution, with Performance mode enabling up to 4X super resolution (i.e. 1080p → 4K). This means more user choice, and even bigger performance boosts.
In testing DLSS 2.0 via Wolfenstein, MechWarrior 5, and an early look at Control's update -- we were blown away by its ability to take a 1080p image and present it in 4K, without any noticeable loss in quality. It's truly remarkable and groundbreaking feature that points to a future where raw-grunt isn't the be-all end-all. Having also demoed DLSS 2.0 using a GeForce RTX 2060, it also means that ray-tracing and running games with max-detail settings is no longer limited to high-end cards.
And with DLSS 2.0 no longer requiring a per-game implementation we should be seeing it crop up in many more PC titles this year.