Sony Patents a System That Splits Game Graphics Work Between Your Device and the Cloud
Every cloud gaming system picks one place to draw the picture: the server or your device. Sony's new patent describes a way to choose on the fly, frame by frame, based on what you're actually looking at.
How Sony's rendering switcher decides what draws your game
You're deep into a game and the scene suddenly fills with something your phone has never had to draw before, a massive crowd, a complex alien structure, a particle storm. Your device chokes, frame rates drop, and the moment falls apart.
Sony's patent describes a server that watches what you're looking at and asks: can your device handle what's coming? If yes, your phone draws the scene locally, which is fast and saves bandwidth. If no, the server takes over and streams the finished picture to you instead.
The switch happens based on whether the objects in your current or upcoming field of view are already familiar to your device. Think of it like a relay race where the baton passes to whoever can run fastest at that moment. The goal is that you never notice the handoff.
… extracts a target object included in a field of view of a user at present or in the near future on the basis of the internal state, determines whether rendering of the content space by a known object of the user terminal is possible on the basis of the target object …
Translation: The system predicts what you will look at soon and checks if your device can handle drawing it.
How the server reads your view and picks a render path
The system runs on a server that stays in constant communication with your device (the patent calls it a "user terminal"). The server continuously reads what it calls the internal state of the content space, essentially a live snapshot of the game world: where you are, which direction you're facing, and what objects are nearby or about to come into view.
From that snapshot, the server identifies the target objects, the specific 3D assets that will fill your screen right now or in the next few moments. It then checks whether your device already has the data it needs to render those objects locally (a "known object" in the patent's language means an asset the terminal can already process without help).
Based on that check, the server picks one of two paths:
- First rendering: your device draws the scene itself, keeping things fast and offline-friendly
- Second rendering: the server draws the scene and sends the finished image to your device, handling objects too complex or unfamiliar for local processing
The key distinction from standard cloud streaming is that this is a conditional decision made per scene state, not a blanket choice applied to the whole session. The patent doesn't specify how often the switch can happen, but the architecture is designed around that real-time prediction loop.
What this means for cloud gaming and device limits
Cloud gaming's central frustration is latency: when the server draws everything and streams it back, even a small network hiccup shows up as blur or stutter. Letting your device draw what it can, and only calling in the server for the hard parts, could reduce how often that lag is visible to you. That's a meaningful quality-of-life shift for anyone who has ever had a cloud game go choppy at the worst possible moment.
For Sony, Sony's steady investment in cloud-gaming infrastructure this sits at the intersection of PlayStation's cloud ambitions and its device ecosystem. A smarter handoff between local and remote rendering could make streaming viable on lower-powered hardware without asking users to buy a new console, which is a real business problem as subscription gaming grows.
That makes this Sony's 469th filing in our Sony coverage since May, adding to work like the phone controller idea and the avatar look patent.
Streaming games over the internet breaks the same way, every time: the connection stumbles for a moment, and suddenly the screen freezes or skips while you're in the middle of something that required your full attention. That failure has kept millions of people paying for expensive home consoles and gaming computers, because the alternative has never been reliable enough to trust.
Sony's patent tries to shrink how often the network gets a chance to cause that problem. By anticipating what a player is about to look at and handling those visuals directly on the player's own device, the system attempts to cover the gaps before they become visible.
The stakes here are not abstract. Every household that avoids game streaming because of past frustration represents a real ceiling on what this market can become, and an approach built around the network's worst moments rather than its best ones is aimed at exactly the right problem.
There are more where this came from
We read every patent application Big Tech publishes and send you the ones worth knowing. Plain English, free, every week.
The drawings
8 drawing sheets from US 2026/0263923 A1 · click any drawing to enlarge
Want this weekly breakdown for a company we don't cover? Patentlyze Pro →