NvidiaPatent applicationChips & hardwareInterest 6/10

Nvidia patents chip design that keeps math results in special memory

Patent drawing
Patent drawing from the official filing.

Nvidia describes a processor design that handles large math calculations called matrix multiply-accumulate operations. It uses a specific type of memory to store the results of these calculations, keeping them separate from other data.

Why it mattersThis could make computer chips faster at handling the heavy math used in graphics and data analysis.

In the filing’s own words

Apparatuses, systems, and techniques to perform a matrix multiply-accumulate (MMA) instruction to exclusively store information to be used by the MMA instruction. In at least one embodiment, a processor retrieves matrix information from a memory that exclusively stores and performs a multiplication computation using said matrix information.

What exactly they claim (claim 1)

1. One or more processors, comprising: circuitry to: perform a matrix multiply accumulate (MMA) instruction to: perform an MMA operation on at least one operand having multiple elements; and store one or more results generated as a result of performing the MMA operation in memory dedicated to storage of information used in MMA operations.

Facts

Company
Nvidia (NVIDIA Corporation)
Official title
STORING INFORMATION FOR MATRIX MULTIPLY-ACCUMULATE OPERATIONS
Publication number
US20260359045A1
Application number
19746257
Filed
Jul 17, 2026
Published
Nov 12, 2026
Inventors
8
Claims
20

Classified under

G06F 9/3001 Physics & computingG06F 9/30036 Physics & computingG06F 9/3887 Physics & computingG06F 9/3893 Physics & computingG06F 7/5443 Physics & computingG06F 17/16 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.