Power/Area-Efficient I/O and Write Circuits for Memristor Crossbar Array-Based Neuromorphic Computing System

Abstract

The overarching research goal is to develop a high-efficiency neuromorphic computing system based on the recently-developed memristor crossbar array technology. While the primary research effort focuses on the system-level implementation and the operating algorithms, the system’s input and output (I/O) and write circuits have not been systematically optimized to achieve high computation efficiency with minimal chip area and power consumption. The research objectives of this project are to design and test power/area-efficient I/O and write circuits used in the memristor crossbar array-based neuromorphic computing system, and seamlessly integrate them with the rest of the system to demonstrate a high-efficiency neuromorphic computer.

Document Details

Document Type
DoD Grant Award
Publication Date
Feb 11, 2016
Source ID
FA87501510069

Entities

People

  • Hao Jiang

Organizations

  • Rome Laboratory
  • San Francisco State University
  • United States Air Force

Tags

Readers

  • Integrated Circuit Design and Technology.