Mark Cerny, PlayStation architect, has confirmed that Sony will implement frame generation technology based on machine learning in its future consoles. This system uses artificial intelligence to create new intermediate frames, aiming to multiply perceived smoothness. Although it promises significant advances, it also introduces challenges such as added latency. The question remains whether it will arrive first on the rumored PS5 Pro or if it will be a pillar of the PS6. 🎮
Beyond upscaling: how it works and how it differs 🤖
This technology should not be confused with resolution upscaling, like the PSSR on the PS5 Pro, which improves image clarity. Nor is it the same as frame interpolation from FSR3 or NVIDIA DLSS3, which combine consecutive real frames. Sony's proposal goes one step further: AI generates new and plausible visual content between two frames rendered by the GPU. This requires specialized AI hardware and a very stable base performance, as it is not a solution for low frame rates, but a smoothness multiplier starting from a solid base.
Implications for development and gaming experience ⚙️
For developers, this technology could simplify the goal of achieving high refresh rates, delegating part of the work to specialized hardware. However, it also imposes a new optimization variable: ensuring that stable base performance and minimizing the inherent latency of the process. For the player, the promise is a sensation of extreme smoothness, but with the potential risk of visual artifacts in fast motion or a slightly less immediate response. Its success will depend on a perfect balance between AI silicon and software optimization.
How could AI frame generation on future PlayStation consoles redefine the limits of visual fidelity and performance in video game development?
(P.S.: a game developer is someone who spends 1000 hours making a game that people complete in 2)