Search International and National Patent Collections

1. (WO2018127422) AREA-EFFICIENT, RECONFIGURABLE, ENERGY-EFFICIENT, SPEED-EFFICIENT NEURAL NETWORK SUBSTRATE

Pub. No.:    WO/2018/127422    International Application No.:    PCT/EP2017/083881
Publication Date: Fri Jul 13 01:59:59 CEST 2018 International Filing Date: Thu Dec 21 00:59:59 CET 2017
IPC: G06N 3/063
G06F 8/41
Applicants: INTERNATIONAL BUSINESS MACHINES CORPORATION
IBM UNITED KINGDOM LIMITED
Inventors: MODHA, Dharmendra
Title: AREA-EFFICIENT, RECONFIGURABLE, ENERGY-EFFICIENT, SPEED-EFFICIENT NEURAL NETWORK SUBSTRATE
Abstract:
Architectures for multicore neuromorphic systems are provided. In various embodiments, a neural network description is read. The neural network description describes a plurality of logical cores. A plurality of precedence relationships are determined among the plurality of logical cores. Based on the plurality of precedence relationships, a schedule is generated that assigns the plurality of logical cores to a plurality of physical cores at a plurality of time slices. Based on the schedule, the plurality of logical cores of the neural network description are executed on the plurality of physical cores.