Accession Number:

ADA581175

Title:

Foundations of Neuromorphic Computing

Descriptive Note:

Final rept. Sep 2009 - Sep 2012

Corporate Author:

AIR FORCE RESEARCH LAB ROME NY INFORMATION DIRECTORATE

Personal Author(s):

• Thiem, Clare
• Wysocki, Bryant
• Bishop, Morgan
• McDonald, Nathan
• Bohl, James

Report Date:

2013-05-01

Pagination or Media Count:

48.0

Abstract:

This neuromorphic computing research and development effort explored the design and implementation of computationally intelligent computer architectures and high performance computer software algorithms. The in-house research concentrated on the design of mathematical models, algorithms, computing architectures, and computational efficiencies for the advancement of neuromorphic computing and neuroprocessors. The software aspect of the research demonstrated the combination of computational power with human level cognitive functionality to create a system that can increase Department of Defense DoD operators and analysts ability to analyze textual and character data. The hardware aspect of the research explored memristor-based and zero instruction set computing technology to provide neuromorphic computing for size, weight, and power limited applications. Progress was made on both the hardware and software side of neuromorphic computing research setting the stage for future agile information systems.

Descriptors:

• *ANALOGS
• *CIRCUITS
• PERCEPTION
• REASONING
• SIMULATION

Subject Categories:

• Electrical and Electronic Equipment
• Theoretical Mathematics

Distribution Statement:

APPROVED FOR PUBLIC RELEASE