NSF Program Director Andrey Kanaev said in an October 2024 press release that the agency’s award for the THOR Project was crucial for advancing the NSF’s mission to drive innovation and broaden access to vital research resources.
“By making bio-inspired computing resources available to a wider community of researchers in computer science, neuroscience and computational physics, this project will contribute to democratizing access to advanced tools and fostering breakthroughs in energy-efficient, resilient AI through neuromorphic computing,” Kanaev said.
In addition to bolstering access, contributors to the THOR Project hope to spark a new wave of algorithm design, neuromorphic computing applications and hardware/software design that would be similar in scale to the impact that high performance computing (HPC) systems had on the research community when they became more accessible. HPCs allowed researchers to “scale up” so they could work with more complex computational methods and larger data sets.
In turn, advances in neuromorphic computing could boost several technology fields because it’s a method that requires less energy than conventional computing and has fewer latency issues. Neuromorphic computing shows particular promise in improving AI research because of its capability to enhance machine learning algorithms with more efficiency and flexibility.
Efficiency and flexibility are the core tenets of neuromorphic computing, which is inspired by the unique complexity of the brain. Despite being a compact organ, the brain’s billions of neurons retain knowledge, interpret senses, initiate movements and control behaviors — all while expending very little energy. The brain also adapts over time, responding to survival instincts and managing the demands of its surroundings while still fostering intellectual curiosity.
Similarly, neuromorphic computing systems use intricate neural networks of chips that deploy artificial neurons and synapses to process information and solve problems. The networks simulate how neurons in the body transmit information through discrete spikes over time, allowing the systems to adapt and more efficiently process patterns in data.
The NSF grant will further UTSA’s vision to create a national hub for open access to large-scale neuromorphic platforms through industry partnerships. To complement these platforms, the THOR Project team plans to develop training and education materials to cover the fundamentals of neuromorphic learning algorithms and systems. Open access to these resources will be available to researchers as well as K-12 students.
NSF Program Director Andrey Kanaev said in an October 2024 press release that the agency’s award for the THOR Project was crucial for advancing the NSF’s mission to drive innovation and broaden access to vital research resources.
“By making bio-inspired computing resources available to a wider community of researchers in computer science, neuroscience and computational physics, this project will contribute to democratizing access to advanced tools and fostering breakthroughs in energy-efficient, resilient AI through neuromorphic computing,” Kanaev said.
In addition to bolstering access, contributors to the THOR Project hope to spark a new wave of algorithm design, neuromorphic computing applications and hardware/software design that would be similar in scale to the impact that high performance computing (HPC) systems had on the research community when they became more accessible. HPCs allowed researchers to “scale up” so they could work with more complex computational methods and larger data sets.
In turn, advances in neuromorphic computing could boost several technology fields because it’s a method that requires less energy than conventional computing and has fewer latency issues. Neuromorphic computing shows particular promise in improving AI research because of its capability to enhance machine learning algorithms with more efficiency and flexibility.
Efficiency and flexibility are the core tenets of neuromorphic computing, which is inspired by the unique complexity of the brain. Despite being a compact organ, the brain’s billions of neurons retain knowledge, interpret senses, initiate movements and control behaviors — all while expending very little energy. The brain also adapts over time, responding to survival instincts and managing the demands of its surroundings while still fostering intellectual curiosity.
Similarly, neuromorphic computing systems use intricate neural networks of chips that deploy artificial neurons and synapses to process information and solve problems. The networks simulate how neurons in the body transmit information through discrete spikes over time, allowing the systems to adapt and more efficiently process patterns in data.
Several large manufacturers have recognized neuromorphic computing’s promise in recent years. In 2024, Intel announced the Hala Point system in collaboration with Sandia National Laboratories, which is powered by 1,152 of its Loihi 2 neuromorphic processors.
As industry giants have invested more in the commercialization of neuromorphic computing elements, UTSA has emerged as one of the nation’s leading institutions in neuromorphic computing and neuro-inspired AI research. The university even hosted the Neuro-Inspired Computational Elements (NICE) conference in 2023, bringing together neurocomputing researchers from across the globe alongside leaders from Intel, IBM and the National Institute of Standards and Technology to learn about advances in the field.
In many ways, the MATRIX AI Consortium at UTSA has served a fulcrum between industry and academia — getting familiar with the increasing commercial appeal of neuromorphic computing while also recognizing how neuromorphic computing tools can supercharge important research. Kudithipudi says that’s why it’s important that researchers from UTSA have a hand in bringing neuromorphic computing to a wider audience.
“The field is at a pivotal moment,” Kudithipudi says, “and ensuring access to a broader group of researchers is critical at this stage.”
