Bio-Inspired Programming Can Lessen AI Computing Needs

Machine learning methods such as neural networks have been successfully used in real-time computer vision and signal processing areas. But as demand for AI grows, the computing and energy resources needed to run this complicated processing can’t keep up. Anup Das, PhD, assistant professor of electrical and computer engineering is working with researchers, including students in the Vertically Integrated Projects program, on neuromorphic systems, which mimic biological neurons and synapses and can lessen the computing load.

Anup Das, PhD

Das and his research team are working on several federal grants to develop compiler tool chains to translate a user’s machine learning program to low-level languages that can be interpreted by neuromorphic systems. A key initiative is to develop a common representation across different platforms so that computers across a network can work together to take on smaller chunks of processing. Resource optimization strategies are being developed to improve program performance, as well as an Operating System- like framework that will allow programmers to easily deploy their machine learning programs on neuromorphic systems. The open-sourced programming tools will enable faster development and commercialization of neuromorphic systems in the U.S. and facilitate collaboration with other such communities worldwide.