Control And MechWarrior 5 Add Some Nice New Graphical Features For Nvidia Users
Nvidia is continuing to tweak and improve its proprietary software tools for its graphics cards, and with the newest update it introduces a new version of its most widely applicable one. DLSS, or Deep Learning Super Sampling, is going 2.0, with Control and MechWarrior 5: Mercenaries the first to support it.
DLSS is a method of rendering a game lower than its target resolution, and then using a combination of Deep Learning models and the powerful Tensor cores on Nvidia’s RTX line to process them and reconstruct a single game frame as it’s rendered, filling in the missing detail and producing a cleaner image with a massive performance boost. Its initial implementation delivered mostly on the performance side, but often lead to image quality that was blurry and unappealing. DLSS 2.0 aims to not only improve that, but produce better images than even natively rendered one.
The trick is in new learning algorithms that Nvidia has used, which not only improves image quality beyond native frames but also ditches the requirement for its models to be trained on a per game basis. This means that DLSS could be improved and implemented in more games faster, with the latest Game-Ready driver targeting two for now.
Control is getting an update on March 26 for its second bit of DLC, its story-focused expansion The Foundation, which will also introduce DLSS 2.0. In the video above it’s clear that the previous implementation’s performance gains now include sharper image quality, improving legibility on a lot of text when applied. MechWarrior 5: Mercenaries doesn’t have a specific update for DLSS 2.0, but support is included in Nvidia’s update today.
The driver also includes performance enhancements for Half-Life: Alyx (which is out today and very good according to GameSpot’s 9/10 review), as well as the imminent Resident Evil 3 Remake and the big title update for Ghost Recon: Breakpoint. You can download it here now or through your Geforce Experience client.
Source: Read Full Article