Patent · US Active

Data compaction and memory bandwidth reduction for sparse neural networks

US10096134B2 · kind B2 · utility

20Cited by
1References
21Claims
0Family size

Assignee

Inventors

Key dates

Filing dateFeb 1, 2017
Grant dateOct 9, 2018
Priority date
Expiry dateJun 2, 2037

Classification

  • Technology area (CPC G)Physics
  • CPC primaryG06N3/082
  • WIPO fieldComputer technology
  • WIPO sectorElectrical engineering

Abstract

A method, computer program product, and system for sparse convolutional neural networks that improves efficiency is described. Multi-bit data for input to a processing element is received at a compaction engine. The multi-bit data is determined to equal zero and a single bit signal is transmitted from the memory interface to the processing element in lieu of the multi-bit data, where the single bit signal indicates that the multi-bit data equals zero. A compacted data sequence for input to a processing element is received by a memory interface. The compacted data sequence is transmitted from the memory interface to an expansion engine. Non-zero values are extracted from the compacted data sequence and zeros are inserted between the non-zero values by the expansion engine to generate an expanded data sequence that is output to the processing element.

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