Invention Application
- Patent Title: Convolution Engine for Neural Networks
-
Application No.: US16685892Application Date: 2019-11-15
-
Publication No.: US20200159809A1Publication Date: 2020-05-21
- Inventor: Francky Catthoor , Praveen Raghavan , Dimitrios Rodopoulos , Mohit Dandekar
- Applicant: IMEC VZW
- Priority: com.zzzhc.datahub.patent.etl.us.BibliographicData$PriorityClaim@38aaf0fe
- Main IPC: G06F17/15
- IPC: G06F17/15 ; G06N3/04 ; G06N3/08

Abstract:
A method and hardware system for mapping an input map of a convolutional neural network layer to an output map are disclosed. An array of processing elements are interconnected to support unidirectional dataflows through the array along at least three different spatial directions. Each processing element is adapted to combine values of dataflows along different spatial directions into a new value for at least one of the supported dataflows. For each data entry in the output map, a plurality of products from pairs of weights of a selected convolution kernel and selected data entries in the input map is provided and arranged into a plurality of associated partial sums. Products associated with a same partial sum are accumulated on the array and accumulated on the array into at least one data entry in the output map.
Public/Granted literature
- US11475101B2 Convolution engine for neural networks Public/Granted day:2022-10-18
Information query