Back to results
US 12,730,633 B1Granted
Storing information for matrix multiply-accumulate operations
AssigneeNVIDIA Corporation
InventorsHarold Carter Edwards, Vijay Harshad Thakkar, Gokul Ramaswamy Hirisave Chandra Shekhara, Edward H. Gornish, Rishkul Kulkarni, Maciej Piotr Tyrlik, Sean Jeffrey Treichler, Chao Li
Life of this patent
Estimating the term…
Abstract
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.
Claims20
One or more processors, comprising:
- circuitry to:
- perform a matrix multiply accumulate (MMA) instruction to exclusively store in a specific location of memory one or more results generated as a result of performing the MMA instruction;
- and
- use a single cooperative thread array of a single processor core to control a plurality of processor cores to perform the MMA instruction based, at least in part, on an instruction that identifies an operating mode.
1 more element
A system comprising:
- one or more processors having one or more circuits to:
- perform a matrix multiply accumulate (MMA) instruction to exclusively store in a specific location of memory one or more results generated as a result of performing the MMA instruction;
- and
- use a single cooperative thread array of a single processor core to control a plurality of processor cores to perform the MMA instruction based, at least in part, on an instruction that identifies an operating mode.
1 more element
A method comprising:
- performing a matrix multiply accumulate (MMA) instruction to exclusively store in a specific location of memory one or more results generated as a result of performing the MMA instruction;
- and
- using a single cooperative thread array of a single processor core to control a plurality of processor cores to perform the MMA instruction based, at least in part, on an instruction that identifies an operating mode.
Dependencies are inferred from the claim text.