Patent №
US 12,373,515
Granted
2025-07-29
Filed 2024
Owner
Texas Instruments Incorporated
Lab
—
AI components
0
Assignment
None on record
Dataset
AIPD
Application
18633703
A method for performing a fundamental computational primitive in a device is provided, where the device includes a processor and a matrix multiplication accelerator (MMA). The method includes configuring a streaming engine in the device to stream data for the fundamental computational primitive from memory, configuring the MMA to format the data, and executing the fundamental computational primitive by the device.
G06F 17/16G06N 3/045G06N 3/0464G06F 17/141G06N 3/063
Ownership
Texas Instruments Incorporated
From the same owner
MERGING MULTIPLE EXPOSURES TO GENERATE A HIGH DYNAMIC RANGE IMAGEUS 12,670,886 · 2026Sparsified Training of Convolutional Neural NetworksUS 12,645,921 · 2026NEURAL NETWORK TECHNIQUES FOR PREDICTING OBJECT POSES BASED ON PARAMETERIZED AND NON-PARAMETERIZED ROTATION AND TRANSLATION VECTORSUS 12,639,850 · 2026METHODS AND APPARATUS TO PROVIDE AN EFFICIENT SAFETY MECHANISM FOR SIGNAL PROCESSING HARDWAREUS 12,602,923 · 2026MACHINE LEARNING TECHNIQUES FOR CIRCUIT DESIGN VERIFICATIONUS 12,585,848 · 2026METHODS OF BATCH-BASED DNN PROCESSING FOR EFFICIENT ANALYTICSUS 12,586,153 · 2026BIAS SCALING FOR N-BIT CONSTRAINED HARDWARE ACCELERATIONUS 12,579,213 · 2026MATHEMATICAL CALCULATIONS WITH NUMERICAL INDICATORSUS 12,579,356 · 2026