By Jon Martindale
Publication Date: 2026-03-23 11:00:00
Nvidia “GeForce Evangelist” Jacob Freeman spoke with YouTuber Daniel Owen late last week about the company’s new DLSS 5 technology. He explained a little more about how the tech works, its limitations, and why some strange artifacts appeared in its early demonstration. While making it clear that this is an early look at the technology, he confirmed that DLSS 5 simply analyzes a 2D scene and infers everything from that, along with motion vector data.
That’s probably why it looked so weird.
Earlier this week, Nvidia set the gaming and PC enthusiast community ablaze with a first look at its next-generation DLSS 5 technology. It purports to adjust lighting and color grading to deliver a more realistic-looking image in games, and the effect can be dramatic. But gamers hated the “yassified” and “AI slop” look of the uprated visuals, and it took an entire second RTX 5090 to even run it.
To try to get ahead of the growing problem, Nvidia CEO Jensen Huang told gamers they were “absolutely wrong” about how DLSS 5 worked. The company suggested it was an incredible technology that might require Nvidia graphics cards, but would unlock more realistic gaming visuals than ever before. Gamers still aren’t so sure, which is why Owen has been asking some interesting questions and receiving some interesting answers.
When pressed on how it works, Freeman said, “Yes, DLSS 5 takes a 2D frame plus motion…
