Dlss 5 revealed at GTC as Nvidia pushes real-time AI deeper into games
dlss 5 was unveiled by Nvidia at GTC, with the company framing it as a step toward bringing real-time AI into video games. The announcement lands as Nvidia also continues to spotlight path tracing and neural rendering as core pillars of its modern graphics push. The update comes amid a wider drumbeat of messaging around faster path tracing performance and expanding DLSS-related features for new games.
What Nvidia announced and why it matters now
Nvidia presented dlss 5 at GTC as part of its effort to bring real-time AI into games. Beyond the headline reveal, Nvidia’s broader message has been consistent: AI and “neural rendering” are central to lifting visual fidelity without giving up performance, especially as path tracing becomes a bigger focus for cutting-edge graphics.
In a separate on-stage discussion at GDC 2026, John Spitzer, Developer & Performance Vice President at Nvidia, described how the company tracks ray tracing and path tracing progress across GPU generations and tied those gains to dedicated hardware blocks used for AI workloads. Spitzer argued that silicon improvements alone will not be enough to reach photorealistic visuals, and positioned AI advances as the catalyst for major leaps in rendering capability.
Path tracing performance claims and the “neural rendering default” message
At GDC 2026, Spitzer said Nvidia’s internal comparisons show path tracing performance on today’s Blackwell gaming GPUs (RTX 50) is “10, 000 times” improved over Pascal-era GPUs (RTX 10). He attributed the gains largely to hardware-accelerated neural rendering enabled by RT and Tensor cores, and said features like DLSS rely on AI models trained on Nvidia supercomputers to reconstruct and interpolate frame data more accurately.
Spitzer also outlined a far more ambitious target for the future: “1, 000, 000 times” better path tracing performance compared with the RTX 10 series, arguing newer and more efficient hardware blocks will make neural rendering the default. Nvidia CEO Jensen Huang has also publicly claimed neural rendering is the default going forward, reinforcing the company’s strategic direction.
On timelines, Spitzer pointed to next-generation Rubin GPUs, which he said are slated to launch sometime between 2027 and 2028. He suggested these future GPUs could help deliver the “1-million-times” path tracing reality described in the presentation.
Expanding game support and new technologies highlighted
As Nvidia presses its case for AI-driven graphics, it has also highlighted the growing list of games with path tracing support, naming Resident Evil Requiem as the latest addition. During the same GDC 2026 presentation, Nvidia referenced additional path tracing technologies including ReSTIR (spatiotemporal resampling algorithms) and RTX Mega Geometry.
To demonstrate what it says is possible with these approaches, Nvidia showed a tech demo for Witcher 4 featuring “over two trillion triangles” in a scene, aiming to illustrate realistic foliage and lighting at the same time.
Separately, Nvidia has also promoted upcoming DLSS-related game additions, including 007 First Light, CONTROL Resonant, and Tides of Annihilation, while stating that “DLSS 4. 5 Dynamic Multi Frame Gen” is available March 31. Nvidia did not provide additional details in the provided materials beyond naming those titles and the March 31 availability note.
What’s next
In the near term, Nvidia’s next visible milestone is March 31 for “DLSS 4. 5 Dynamic Multi Frame Gen, ” while the company’s longer arc now explicitly includes dlss 5 as part of its real-time AI gaming roadmap. Looking further out, Nvidia’s stated Rubin GPU window of 2027–2028 will be the next major checkpoint for whether the company’s path tracing performance targets and “neural rendering default” strategy translate into real-world gaming hardware and broader developer adoption.
