/Patent Tracker
NvidiaPatent applicationAI & softwareInterest 6/10

Nvidia patents method to run only parts of AI models based on input

Patent drawing
Patent drawing from the official filing.

Nvidia describes a system that chooses specific sections of a neural network to process data, rather than running the entire model every time. It loads only the needed parts into memory to finish the calculation, sharing the initial results across these selected sections.

Why it mattersThis approach could make AI processing more efficient by avoiding unnecessary calculations, which is useful for devices with limited power or memory.

In the filing’s own words

Apparatuses, systems, and techniques to perform inferencing with neural networks. In at least one embodiment, portions of one or more neural networks are selected for use in inferencing based, at least in part, on information to be inferenced by the one or more neural networks.

What exactly they claim (claim 1)

1. A system, comprising: a memory to store network parameters of one or more layers of one or more neural networks to generate output values from a portion of inferencing based, at least in part, on input data; and one or more processors to: select one or more sub-networks from among a plurality of sub-networks of the one or more neural networks to complete the inferencing based, at least in part, on the input data or the output values from the one or more layers; load the selected one or more sub-networks to the memory; and provide the output values from the one or more layers to the selected one or more sub-networks to complete the inferencing, wherein the output values are duplicated to each selected sub-network such that the portion of inferencing is common to the plurality of sub-networks.

Facts

Company
Nvidia (NVIDIA Corporation)
Official title
PARTIAL NEURAL NETWORK ACTIVATION
Publication number
US20260321048A1
Application number
19718934
Filed
Jun 24, 2026
Published
Oct 15, 2026
Inventors
2
Claims
1

Classified under

G06V 30/19113 Physics & computing

More from Nvidia in AI & software

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.