/Patent Tracker
TeslaGranted patentChips & hardwareInterest 4/10

Tesla patents chip design that picks math tasks based on network position

Patent drawing
Patent drawing from the official filing.

This patent describes a processor system that combines a matrix processor for neural network convolutions with SIMD processors. The SIMD units choose which group of operations to run from a list of programs based on the current position in the neural network processing, using a program counter with start and end pointers to limit execution.

Why it mattersThis improves how Tesla's hardware handles neural network calculations by dynamically selecting specific instruction sets, potentially making their AI processing more efficient.

In the filing’s own words

Systems and methods for efficient selection of single instruction multiple data operations for neural processing units. An example processor system comprises a matrix processor configured to perform convolutions associated with a neural network and single instruction multiple data (SIMD) processors in communication with the matrix processors, with the SIMD processors being configured to execute a group of operations based on a current position associated with processing the neural network, and with the group of operations being selected from multiple SIMD programs, and with the group of operations being selected from the SIMD programs according to the current position.

What exactly they claim (claim 1)

1. A processor system comprising: a matrix processor configured to perform convolutions associated with a neural network; and one or more single instruction multiple data (SIMD) processors in communication with the matrix processor, wherein the SIMD processors are configured to execute a group of operations based on a current position associated with processing the neural network, wherein the group of operations is selected from a plurality of SIMD programs, wherein the group of operations is selected from the SIMD programs according to the current position, wherein the group of operations is selected using a program counter, and wherein the program counter is used to limit execution between a beginning pointer and an ending pointer, and wherein the beginning pointer and the ending pointer identify the group of operations from the SIMD programs.

Facts

Company
Tesla (Tesla, Inc.)
Official title
Efficient selection of single instruction multiple data operations for neural processing units
Patent number
US12748723B2
Application number
18861756
Filed
May 9, 2023
Granted
Sep 29, 2026
Inventors
3
Claims
17

Classified under

G06F 15/8007 Physics & computingG06F 9/3887 Physics & computingG06N 3/0464 Physics & computing
Read the full filing (official PDF) ↗ Open in the Patent Tracker

More from Tesla 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.