Neural network accelerator with a configurable pipeline

ChangeBridge: Patent Grants - AI & Computing (G06N)

Published April 7th, 2026

Detected April 7th, 2026

Summary

USPTO granted Patent US12596919B2 to Imagination Technologies Limited for a neural network accelerator featuring a configurable hardware pipeline. The invention comprises multiple hardware processing units for accelerating neural network operations on data tensors, interconnected via a crossbar that selectively forms configurable processing pipelines. Inventors Javier Sanchez and Alan Vines secured 20 claims covering the configurable processing order and pipeline architecture.

View original document View source feed page

What changed

USPTO issued Patent US12596919B2 granting Imagination Technologies Limited exclusive rights to a neural network accelerator with configurable pipeline architecture. The patent covers hardware processing units capable of performing neural network operations on tensors, with a crossbar enabling dynamic pipeline configuration and selectable processing orders based on the formed pipeline topology.

Technology companies developing AI accelerators, neural processing units, or related hardware should review the patent claims to assess potential licensing needs or design-around considerations. The patent's broad coverage of configurable pipeline architecture in neural network hardware may affect product development strategies for competing implementations in AI acceleration technology.

What to do next

Monitor for updates
Review patent claims for infringement risk

Source document (simplified)

← USPTO Patent Grants

Neural network accelerator with a configurable pipeline

Grant US12596919B2 Kind: B2 Apr 07, 2026

Assignee

Imagination Technologies Limited

Inventors

Javier Sanchez, Alan Vines

Abstract

A neural network accelerator that has a configurable hardware pipeline includes a plurality of hardware processing units, each hardware processing unit comprising hardware to accelerate performing one or more neural network operations on a tensor of data; and a crossbar coupled to each hardware processing unit of the plurality of hardware processing units, the crossbar configured to selectively form, from a plurality of selectable pipelines, a pipeline from one or more of the hardware processing units of the plurality of hardware processing units to process input data to the neural network accelerator. At least one of the hardware processing units is configurable to transmit or receive a tensor via the crossbar in a selected processing order of a plurality of selectable processing orders, and the selected processing order is based on the pipeline formed by the crossbar.