PRODUCT-SUM OPERATION DEVICE, NEUROMORPHIC DEVICE, AND METHOD FOR USING PRODUCT-SUM OPERATION DEVICE

WIPO Patent Application WO/2019/176540

A1

This product-sum operation device (1) is provided with a product calculation unit (10) and a sum calculation unit (11). The product calculation part (10) is provided with a plurality of product calculation elements (10AA, 10AB), which are variable resistance elements, and at least one reference element (10AMAX, 10AMIN). The sum calculation unit (11) is provided with an output detector (11A) which detects the sum value of at least the outputs from the plurality of product calculation elements. Each of the plurality of product calculation elements is a magnetoresistance effect element, which has: a magnetized free layer having magnetic domain walls; a magnetized fixed layer having a fixed magnetization direction; and a non-magnetic layer sandwiched between the magnetized free layer and the magnetized fixed layer. The reference element is a reference magnetoresistance effect element having a magnetized free layer that does not have magnetic domain walls.

JP2856259	[Title of Invention] Stabilization Method of Self-Learning System
JPH06187006	DISTURBANCE COMPENSATING DEVICE
JPH03129489	OPTIMIZING METHOD FOR DATA PROCESSOR

SHIBATA TATSUO (JP)
SASAKI TOMOYUKI (JP)

PCT/JP2019/007557

September 19, 2019

February 27, 2019

Click for automatic bibliography generation  

TDK CORP (JP)

G06G7/60; G06N3/063; H01L21/822; H01L21/8239; H01L27/04; H01L27/105; H01L29/82

WO2017183573A1	2017-10-26
WO2017183573A1	2017-10-26

JP2009282782A	2009-12-03
JP2016004589A	2016-01-12
JP2018049766A	2018-03-29

ZHE CHEN ET AL.: "Optimized Learning Scheme for Grayscale Image Recognition in a RRAM-Based Analog Neuromorphic System", 2015, IEEE, pages: 17.7.1 - 17.7.4

TANAI Sumio et al. (JP)

View/Download PDF      PDF Help

Previous Patent: CONNECTION TERMINAL AND TERMINAL CONNECTION BODY

Next Patent: INFORMATION PROCESSING DEVICE, DISPLAY SYSTEM, AND DISPLAY PROGRAM