/Patent Tracker
NvidiaPatent applicationChips & hardwareInterest 6/10

Nvidia patents method to split GPU tasks into parallel groups

Nvidia describes a way to organize computer instructions called threads into smaller groups that run at the same time on different parts of a graphics processor. This technique, known as cooperative thread groups, allows the hardware to handle large tasks more efficiently by breaking them down into manageable pieces.

Why it mattersThis filing details how Nvidia's graphics chips manage complex calculations, which is fundamental to how modern gaming and AI hardware works.

In the filing’s own words

Apparatuses, systems, and techniques to implement cooperative thread groups. In at least one embodiment, two or more groups of synchronized threads are synchronized and executed in parallel for efficient execution on a GPU.

What exactly they claim (claim 1)

1. One or more processors, comprising: circuitry to: in response to an application programming interface (API) call, distribute a first group of threads into two or more subgroups and cause each subgroup of the two or more subgroups to be performed in a separate processor core in parallel, wherein the API indicates a total number of subgroups in which the first group of threads is to be distributed, wherein the total number of subgroups to be distributed is based, at least in part, on a size of each of the two or more subgroups.

Facts

Company
Nvidia (NVIDIA Corporation)
Official title
PARALLEL THREAD SYNCHRONIZATION
Publication number
US20260310524A1
Application number
19717428
Filed
Jun 23, 2026
Published
Oct 8, 2026
Inventors
4
Claims
20

Classified under

G06F 9/522 Physics & computingG06F 9/544 Physics & computing

More from Nvidia in Chips & hardware

Summaries are simplified from the filing’s own abstract and first claim. A patent filing is not a product announcement.
Patent data and drawings: U.S. Patent and Trademark Office (USPTO). Nebula is not affiliated with or endorsed by the USPTO.