SILICON BRAIN

WIPO Patent Application WO/2023/228869

A1

The standard for computing storage capacity in modern computing is a bit, and the number of transistors (elements) that are to be nodes, namely, the number of bits is a quantitative unit of modern information communication. In contrast, the brain of a human is composed of cerebral neural circuits, and the storage capacity of a human is not based on the number of cerebral nerves (node number). The circuit complexity in the same number of nodes is much larger than the bit capacity. Modern artificial intelligence, which intends to reproduce a human brain from computing based on bits, basically handles different information, and thus is inherently wasteful. In addition, the computing based on the number of bits is always accompanied by a limit to the degree of integration. [Problem] To provide an information storage system which simulates a neural network on a silicon chip and is independent of bits. [Solution] An information processing mechanism of a human brain is produced by switching electrical connections of non-volatile memory cells distributed as an array on a chip.

WO/2017/200088	NEURAL NETWORK CIRCUIT AND NEURAL NETWORK INTEGRATED CIRCUIT
WO/2023/179619	NEURAL NETWORK CACHING METHOD, SYSTEM, AND DEVICE AND STORAGE MEDIUM
WO/2020/053148	REMOVING NOISE CAUSED BY VEHICULAR MOVEMENT FROM SENSOR SIGNALS USING DEEP NEURAL NETWORKS

WATANABE HIROSHI (JP)

PCT/JP2023/018673

November 30, 2023

May 19, 2023

Click for automatic bibliography generation  

WATANABE HIROSHI (JP)

G06N3/063; G11C11/54; G11C13/00; G11C16/04

JP2009080892A	2009-04-16
JP2018521400A	2018-08-02

View/Download PDF      PDF Help

Previous Patent: SUBSTRATE FOR ELECTRONIC DEVICE, AND MANUFACTURING METHOD THEREFOR

Next Patent: COMPOSITION FOR COLORING KERATIN FIBERS