USPTO Patent Grant for Analog Resistive Processing Unit System

ChangeBridge: Patent Grants - AI & Computing (G06N)

Published March 24th, 2026

Detected March 25th, 2026

Summary

The USPTO has granted patent US12585940B2 to International Business Machines Corporation for an analog resistive processing unit system. The patent covers techniques for learning static bound management parameters for such systems, which are configured for neuromorphic computing and artificial intelligence applications.

View original document View source feed page

What changed

The United States Patent and Trademark Office (USPTO) has issued patent US12585940B2 to International Business Machines Corporation. This patent grants protection for "Learning static bound management parameters for analog resistive processing unit system," specifically detailing techniques for configuring analog resistive processing unit systems for neuromorphic computing. The system involves training artificial neural network models using matrix-vector compute operations influenced by bound management parameters, leading to a refined model with learned parameters.

This patent grant is primarily an intellectual property matter and does not impose direct regulatory obligations or compliance deadlines on regulated entities. However, it signifies innovation in the field of AI hardware and neuromorphic computing, which may influence future technological developments and potentially lead to new industry standards or regulatory considerations in the long term. Companies operating in the AI and semiconductor sectors should be aware of this granted patent as it pertains to their research and development in related areas.

Source document (simplified)

← USPTO Patent Grants

Learning static bound management parameters for analog resistive processing unit system

Grant US12585940B2 Kind: B2 Mar 24, 2026

Assignee

International Business Machines Corporation

Inventors

Malte Johannes Rasch, Manuel Le Gallo-Bourdeau, HsinYu Tsai, Charles Mackin, Nandakumar Sasidharan Rajalekshmi, An Chen

Abstract

Techniques are provided for learning static bound management parameters for an analog resistive processing unit system which is configured for neuromorphic computing. For example, a system comprises one or more processors which are configured to: perform a first training process to train a first artificial neural network model; perform a second training process to retrain the first artificial neural network model using matrix-vector compute operations which are a function of bound management parameters of an analog resistive processing unit system, to thereby generate a second artificial neural network model with learned static bound management parameters; and configure the resistive processing unit system to implement the second artificial neural network model and the learned static bound management parameters.