PlayStation AI frame generation was confirmed by Mark Cerny, lead architect of the PlayStation 5 and PS5 Pro. The machine-learning-based technology, co-developed with AMD, will use AI to imagine intermediate frames and smooth the image on future platforms.
What happened?
In an interview with Digital Foundry, Cerny explained that FSR frame generation is based on co-developed (or “co-engineered”) technology between Sony Interactive Entertainment and AMD in Project Amethyst. An equivalent library should appear on PlayStation platforms at some point. He made clear there are no further releases planned for this year.
The PS5 already offers frame generation via AMD FSR3 in some games, but without full machine-learning support. The new approach uses AI to predict what should be between truly rendered frames rather than simply interpolating.
Why does this matter?
AI frame generation allows a perception of higher fluidity without requiring the GPU to render every frame natively. On consoles this can help balance high resolution and high frame rates. Cerny did not specify whether it will reach the PS5 Pro or only a future PS6, but the collaboration with AMD on PSSR 2 already shows the path.
Sony’s patent for “Generation Super Sampling,” which takes controller inputs into account for more accurate prediction, reinforces the company’s interest in the technology.
What changes in practice?
PlayStation players can expect, at some future point, performance modes that use AI to insert frames and improve the feeling of smoothness, similar to what DLSS and FSR offer on PC. It should not arrive in 2026, according to Cerny. The technology continues to evolve alongside the ray-tracing and upscaling hardware already present on current platforms.
By Grok
