Dhireesha Kudithipudi is spearheading the first open-access neuromorphic computing hub in the U.S. at the University of Texas at San Antonio, aiming to democratize artificial intelligence research.
Indian American computer scientist Dhireesha Kudithipudi is transforming the landscape of artificial intelligence (AI) in the United States. As the founding director of the MATRIX AI Consortium at the University of Texas at San Antonio (UTSA), she is at the forefront of launching THOR: The Neuromorphic Commons, the first open-access hub of its kind in the country.
Funded by the National Science Foundation, the THOR project seeks to democratize access to neuromorphic computing, a field that emulates the architecture of the human brain to process information. Unlike traditional silicon chips, which consume significant amounts of electricity regardless of the task, neuromorphic systems operate on an “event-based” model, activating only when new data is detected.
“THOR is the U.S. national hub for neuromorphic computing,” Kudithipudi stated. She also holds the Robert F. McDermott Chair in Engineering at UTSA. “We are democratizing the technology, expanding industry-academia partnerships, and serving as a catalyst for bringing neuromorphic computing closer to real-world applications.”
Historically, access to such advanced hardware has been limited to elite corporate laboratories or well-funded academic institutions. In contrast, UTSA’s new initiative functions similarly to a public library, allowing researchers and students nationwide to apply for free access to run experiments. This approach significantly lowers the barrier to entry for the next generation of engineers.
At the core of the hub is the SpiNNaker2 system, a substantial platform featuring approximately 400,000 processing elements. Developed in collaboration with SpiNNcloud, this hardware utilizes energy-efficient ARM-based cores, akin to those found in smartphones, to simulate the pulsing signals of biological neurons and synapses.
The practical implications of this energy efficiency are profound. According to the research team, neuromorphic chips have the potential to revolutionize medical devices. For instance, they could enable pacemakers to adapt in real-time to a patient’s physical distress or allow hearing aids to intelligently filter background noise without quickly draining their batteries.
In addition to energy savings, Kudithipudi and her colleagues are addressing the issue of “catastrophic forgetting,” a common flaw in AI systems where machines lose previously acquired knowledge when learning new information. By mimicking the brain’s “lifelong learning” capabilities, THOR could facilitate the development of AI that evolves continuously.
This initiative involves a nationwide collaboration, with contributions from experts at UT Knoxville, UC San Diego, and Harvard University. The official launch of THOR is scheduled for February 23, marking a significant milestone for UTSA’s newly established College of AI, Cyber and Computing.
For Kudithipudi, the overarching goal is to ensure that the future of computing is not only more powerful but also more accessible and sustainable for all.
The information for this article was sourced from The American Bazaar.