Skip to content

DLSS 5 explained, with Linus Tech Tips

The Vergecast

Hosted by David Pierce

36m
Transcript

Original podcast

Summary

Written by a model from the transcript; not yet reviewed.

Nvidia's release of DLSS 5 on September 3rd has sparked intense debate across the gaming community and developer circles. Unlike past versions focused on traditional upscaling and frame generation, DLSS 5 uses neural rendering to generate elements, textures, and colors that were not originally in the scene. In this episode, guest Linus Sebastian and David Gautier from Linus Tech Tips join the host to break down what DLSS 5 actually does, why it currently requires extremely high-end hardware like a 5090 while taking a heavy performance hit, and whether this technology represents the future of gaming or simply a tool for rushed developers. The discussion also explores the broader implications of Nvidia's hardware ecosystem, pricing pressures on PC gamers, and how the industry weighs automated enhancements against handcrafted art styles.

  • Nvidia's DLSS 5 moves beyond traditional upscaling by using trained AI models to add missing elements and colors to game scenes. 5:02
  • The current performance cost of running DLSS 5 is extremely high, requiring top-tier hardware like an Nvidia 5090. 12:25
  • DLSS 5 relies heavily on vector data to function properly, meaning simple filters applied to webcams or video inputs will fail. 33:38
  • Nvidia has announced plans to work on a 4000 series version of DLSS 5 to address community backlash and hardware limitations. 16:44

Chapters

  1. 0:02
    Vergecast Intro and Host Welcome

    The hosts introduce the episode, announce the release of Nvidia’s DLSS 5, and set the stage for a deep dive into the new technology.

  2. 0:41
    90‑Second News Segment

    A rapid rundown of tech news, including Beats headphones, Peloton treadmill, Microsoft layoffs, and Apple’s new venue, delivered in a concise format.

  3. 3:53
    DLSS 5 Overview and Host Questions

    The hosts ask Linus Tech Tips’ Linus Sebastian and David Gautier to explain what DLSS 5 is and why it differs from previous DLSS versions.

  4. 5:02
    Technical Deep Dive into Neural Rendering

    Linus explains how DLSS 5 uses machine learning to add new scene elements and colors, moving beyond simple upscaling to true neural rendering.

  5. 9:07
    Artist Intent and Game‑Specific Impact

    The conversation turns to how DLSS 5 affects artistic vision, with examples from Expedition 33 and concerns about stylized games losing their intended look.

  6. 12:06
    Performance, Ecosystem, and Competitive Landscape

    Discussion covers the high performance cost on current GPUs, Nvidia’s strategy, and how AMD/Intel might respond with competing features.

  7. 35:50
    Future Outlook and Community Reactions

    Guests speculate on DLSS 5’s evolution, potential for wider adoption, and how the community’s backlash might shape future releases.

  8. 39:39
    Wrap‑Up and Listener Takeaways

    The hosts thank the guests, recap key points, and provide contact and support information for listeners.

Highlights

  • 5:02

    Explaining DLSS 5 Neural Renderinginsight

    The guest explains that DLSS 5 goes beyond traditional pixel interpolation by using a trained AI model to add elements and colors into a scene that were not actually rendered.

  • 7:40

    Limitations in Actual Game Testinginsight

    The guest notes that when tested on games without developer optimization, DLSS 5 falls short of magic photorealism and leaves an unmistakable AI look.

  • 12:25

    Heavy Performance Hit on Hardwaredata

    The host and guest discuss how DLSS 5 carries a massive 50% performance cost, making it currently unusable for most standard PC gamers.

  • 27:36

    The Debate Over Optimized Gamesdebate

    The speakers debate whether massive handcrafted projects like GTA 6 demonstrate that developers should focus on optimization rather than relying on AI tools.

  • 33:38

    Why Webcams Cannot Use DLSS 5strategy

    The guest explains why unauthorized attempts to use DLSS 5 filters on webcams or capture cards fail: the technology requires specific vector data to avoid glitches.

Quotes

  • “DLSS 5 takes that, and instead of just, you know, adding pixels that weren't there, it seems to be adding elements of the scene that weren't there, or adding colors that weren't there.”
    Unnamed speaker · 5:02
  • “I think it's a little disingenuous for people to say, 'Well, it's just an Instagram filter.'”
    Unnamed speaker · 7:23
  • “I think the DLSS 5 that we're going to have in a year is going to be so markedly different than the one we have now.”
    Unnamed speaker · 38:50
  • Amazon Prime 2:00–2:18
  • Amazon Prime 22:40–23:20

About this episode

Nvidia's new DLSS 5 (Deep Learning Super Sampling 5) rendering technology uses AI to (theoretically) improve video game graphics. When it was announced, social media declared that it was "yassifying" video game characters, changing faces in a way game developers didn't intend. Linus Tech Tips' own Linus Sebastian and David Gauthier join guest host David Imel to break down the the hype and controversy. Further reading: The Beats 360 are what I’d bring to the gym if I hated earbuds Peloton is back with a ‘cheaper’ folding treadmill Whoop and Peloton integration is finally here Xbox is barely Xbox anymore Apple’s first live music venue is opening beneath its UK headquarters Nvidia’s DLSS 5, explained DLSS 5: Has Nvidia’s AI graphics technology gone too far? DLSS 5 looks like a real-time generative AI filter for video games Subscribe to The Verge for unlimited access to theverge.com, subscriber-exclusive newsletters, and our ad-free podcast feed. We love hearing from you! Email your questions and thoughts to [email protected] or call us at 866-VERGE11. Learn more about your ad choices. Visit podcastchoices.com/adchoices

Transcript

0:00
36:35

6,954 words · 6 speakers

Speaker names marked are inferred from speaking time and the show’s known hosts — they are not voice-verified.

Topics

  • Nvidia
  • DLSS 5
  • Prime Big Deal Days
  • The Home Depot
  • Peloton
  • Apple
  • AMEX Corporate
  • TreadFlex
  • Whoop
  • Microsoft

Page updated · transcript and summary are machine-generated