The first time I ever saw Nvidia’s AI-infused DLSS 2 technology in action, I said it works like black magic. Leveraging the dedicated AI tensor cores in GeForce RTX graphics cards let Nvidia run supported games much faster than before with little to no visual downgrade. Today, Nvidia is unleashing those same tensor cores on web videos with RTX Video Super Resolution after unveiling the feature at CES 2023.
I spent the weekend playing around with the technology and I’m thrilled to report that while there are a few minor DLSS-like caveats to be aware of, RTX Video Super Resolution is an absolute game changer on par with DLSS. If you watch web videos regularly, it’s a compelling reason to consider Nvidia over AMD if you’re shopping for a new graphics card and use Chrome or Edge. But better yet, it’s coming to existing RTX 30-series GPUs as well as Nvidia’s new RTX 40-series offerings.
WHAT IS RTX VIDEO SUPER RESOLUTION?
Let’s start with a quick recap of how RTX Video Super Resolution—introduced today in Nvidia’s new Game Ready Driver (531.14)—works.
As PCWorld’s Keith May explains here, after you activate RTX Video Super Resolution in the Nvidia Control Panel (more on that later), it runs any video displayed in Chrome or Edge through an AI algorithm designed to drastically improve the final image.
“It runs [your video] through an algorithm that softens the darks and the blacks,” Keith explained. “The final image isn’t going to be a true representation of what [native 4K video] would look like, because you’re working with a 1080p source, but it’s going to look substantially better.”
