Patent · US Active

Compute near memory convolution accelerator

US11726950B2 · kind B2 · utility

1Cited by
1References
8Claims
0Family size

Assignee

Inventors

Key dates

Filing dateSep 28, 2019
Grant dateAug 15, 2023
Priority date
Expiry dateJul 17, 2041

Classification

  • Technology area (CPC Y)Emerging Cross-Sectional Technologies
  • CPC primaryY02D10/00
  • WIPO fieldComputer technology
  • WIPO sectorElectrical engineering

Abstract

A compute near memory (CNM) convolution accelerator enables a convolutional neural network (CNN) to use dedicated acceleration to achieve efficient in-place convolution operations with less impact on memory and energy consumption. A 2D convolution operation is reformulated as 1D row-wise convolution. The 1D row-wise convolution enables the CNM convolution accelerator to process input activations row-by-row, while using the weights one-by-one. Lightweight access circuits provide the ability to stream both weights and input rows as vectors to MAC units, which in turn enables modules of the CNM convolution accelerator to implement convolution for both [1×1] and chosen [n×n] sized filters.

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