Patent · US Active

Inner product convolutional neural network accelerator

US12131250B2 · kind B2 · utility

0Cited by

3References

25Claims

0Family size

Assignee

Intel Corporation · US

Inventors

Ehud Cohen · Kiryat Motzkin, IL
Moshe Maor · Kiryat Shmona, IL
Ashutosh Parkhi · Kanchinakote, IN
Michael Behar · Jerusalem, IL
Yaniv Fais · Tel Aviv-Yafo, IL

Key dates

Filing date	Sep 29, 2017
Grant date	Oct 29, 2024
Priority date	—
Expiry date	Dec 7, 2041

Classification

Technology area (CPC G)Physics
CPC primaryG06V10/955
WIPO fieldComputer technology
WIPO sectorElectrical engineering

Abstract

A convolutional neural network (CNN) accelerator, including: a CNN circuit for performing a multiple-layer CNN computation, wherein the multiple layers are to receive an input feature according to an input feature map (IFM) and a weight matrix per output feature, wherein an output of a first layer provides an input for a next layer; and a mapping circuit to access a three-dimensional input matrix stored as a Z-major matrix; wherein the CNN circuit is to perform an inner-product direct convolution on the Z-major matrix, wherein the direct convolution lacks a lowering operation.

Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.