Advanced Education Platform for Multifunctional Intelligent Systems with Self-programming Functionality and High Efficiency
Abstract
This Defense University Instrumentation Program (DURIP) proposal aims to establish key instruments for an “advanced education platform for multifunctional intelligent systems with self-programming functionality and high efficiency.� The world’s fastest supercomputer, Fugaku, may have a computing capacity comparable to that of the human brain and simulate the intelligent functions of the human brain, but Summit consumes the equivalent power of 7000 homes (~15 MW), and the brain only consumes a power of a light bulb (~20 W). This fundamentally restrains computers from real-time learning efficiently, and limits the developments of multifunctional intelligent systems such as selfpiloted unmanned aerial vehicles (SPUAVs). Existing AI systems based on computers lack brain-like real-time learning functionality; adaptability to complex, dynamic environments; high energy efficiency; and general intelligence for versatile applications. By contrast, the brain simultaneously processes and learns from “big data� via trillions of synapses and neurons in analog parallel mode, and facilitates parallel processing and real-time learning with an energy efficiency more than five orders of magnitude superior to that of the supercomputer. In an AFOSR Multidisciplinary University Research Initiative (MURI) project, entitled “Braininspired networks for multifunctional intelligent systems in aerial vehicles�, we have invented a synaptic resistor (synstor) to emulate biological synapses. A self-programming neuromorphic integrated circuit (SNIC) based on synstors have been developed to execute inference and learning algorithms concurrently in real-time with an energy efficiency more than six-orders of magnitudes higher than those of computers. In this DURIP project, we plan to establish key instruments, including (1) SNIC interface circuits; (2) SNIC testing sockets; (3) SNIC interface printed circuit boards; (4) a capacitancevoltage measurement unit; and (5) an ion source. The key instruments will be integrated with SNICs that can dynamically modify their algorithm in a real-time learning process to control various multifunctional intelligent systems (such as self-piloted unmanned aerial vehicle, SPUAV), enable us to circumvent the fundamental limitations of existing AI systems based on computers, thus establish a cutting-edge advanced education platform for new multifunctional intelligent systems with unique self-programming functionality and high efficiency in complex, dynamic environments.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Feb 29, 2024
- Source ID
- FA95502310086
Entities
People
- Yong Chen
Organizations
- Air Force Office of Scientific Research
- United States Air Force
- University of California, Los Angeles