This month's cloud gaming and streaming category includes 6 filed patent applications from 3 companies: Sony (4), Samsung (1), and Intel (1).
Sony's filings cover a range of streaming pipeline techniques, including separating text draw calls from encoded scenes to preserve subtitle and UI readability during degraded connections, ranking game objects by priority to allocate frame rate and resolution more efficiently, capturing synchronized gameplay sessions for debugging purposes, and embedding per-frame display metadata alongside game streams. Samsung describes a gaming hub TV approach that compares screen states before and after a pause to determine which quick-panel tools are applicable to the changed display. Intel's application addresses latency by fetching game content in segments during level transitions and discarding what is no longer needed, avoiding a full installation.
Intel received 1 patent in this category, covering a method for managing game content on a client device by tying downloads and deletions to natural transition points in gameplay, such as level or scene changes. Rather than streaming a fully remote video feed or requiring a complete installation upfront, the approach keeps only the immediately relevant portion of the game on the device, fetching what comes next and discarding what has passed at each transition. The result is a local footprint that stays in step with where the player actually is in the game.
Sony received 4 patents spanning several distinct challenges in cloud game streaming. One filing addresses the problem of text and UI elements becoming unreadable when a connection degrades, proposing that text draw calls be separated from the main scene at the command-buffer level and transmitted outside the video encoding path entirely, so only the scene absorbs compression penalties rather than the full composited frame. A second patent tackles bandwidth and latency by assigning each game object or region its own frame rate and resolution allocation based on priority signals, including where players are looking, what they are interacting with via a controller, and flags set by the game engine itself. A third filing describes a method for embedding per-frame display metadata directly into an encoded gameplay stream, giving the client device the information it needs to make display decisions from the frames it receives rather than handling that externally. The fourth patent proposes a non-intrusive mechanism for tapping live cloud gaming sessions to collect synchronized video, audio, and controller input, labeled by in-game context such as boss encounters or loadout screens, so debugging teams can draw on real session data rather than dedicated test runs.
Samsung received 1 patent describing a gaming hub TV approach that determines which quick-panel overlay tools, such as a mini map, aim point, frame rate counter, or timer, are actually usable after a player pauses a game. Because different games respond to a pause in different ways, producing anything from a full-screen menu to a dimmed or partially obscured frame, a fixed overlay menu that lists every function regardless of context would often expose tools that cannot operate on the visible screen. The device instead captures the screen just before the pause and compares it with the frame the cloud server returns after the pause, then uses that pixel-level comparison to decide which functions remain applicable.
All data sourced from USPTO patent filings. Google Patents may take several weeks to index recent publications. If a link is unavailable, search for the patent number at USPTO Patent Public Search.