Neuromorphic computing is increasingly being explored for use in artificial intelligence and energy-efficient computing, but its potential relationship to neuroscience research has been limited. Today’s neuromorphic systems are reaching brain-like scales and can emulate more of the brain’s complex dynamics than ever before. In my talk, I will describe how considering the theoretical benefits of neuromorphic computing can help us understand both how do design more effective brain-inspired algorithms as well as give us a foothold in understanding how information may be encoded within the brain. I will explain three specific examples. First, I will describe our novel neuromorphic algorithm for iteratively solving large-scale sparse linear systems for finite elements (NeuroFEM). The NeuroFEM approach shows attractive energy and speed compared to conventional methods such as conjugate gradients on standard hardware. Second, I will present graphical neural activity threads, or GNATs, which present a new perspective for spatio-temporal trajectories that we see as potentially relevant for both describing cortical dynamics and neuromorphic algorithm design. Finally, I will show our results in modeling emerging brain connectomes onto neuromorphic hardware, showing how brain-inspired hardware is uniquely powerful for neural simulations. Together, these examples begin to reveal a path by which we may be able to truly bring neuroscience and neuromorphic computing together.

Dr. Brad Aimone is a Distinguished Member of Technical Staff in the Center for Computing Research at Sandia National Laboratories, where he is a lead researcher in leveraging computational neuroscience to advance artificial intelligence and in using neuromorphic computing platforms for future scientific computing applications. Brad leads several research efforts on designing neural algorithms for NeuroAI, scientific computing applications, and neuromorphic machine learning implementations.

Brad has published over 100 peer-reviewed journal and conference articles in venues such as Advanced Materials, Neuron, Nature Neuroscience, Nature Electronics, Communications of the ACM, and PNAS and he is one of the co-founders of the Neuro-Inspired Computational Elements, or NICE, conference. Brad led the team that was awarded the 2023 Misha Mahowald Prize in Neuromorphic Engineering.

Prior to joining the technical staff at Sandia in 2011, Dr. Aimone was a postdoctoral research associate at the Salk Institute for Biological Studies, with a Ph.D. in computational neuroscience from the University of California, San Diego and Bachelor’s and Master’s degrees in chemical engineering from Rice University.

The hippocampus is a core region of the brain with a major role in learning and memory and it features a diversity of neuronal types with various connectivity patterns. For example, its cornu ammonis (CA3) subregion offers pattern completion capabilities supported by the resting-state behavior to optimize memory storage and retrieval. These insights could lead to novel efficient computing approaches, particularly when co-designed with emerging hardware technologies. Emerging analog hardware technologies, like memristors, show promise for dense energy efficient systems given their ultra-scalable footprint and better energy/bit consumption. In this talk, I will describe our efforts to draw inspiration from the CA3 subregion of the rodent hippocampus to design efficient and robust memristor-based neuromorphic computing. I will summarize our interdisciplinary efforts across the innovation stack, from new types of materials and synaptic devices to new types of prototyping systems and hippocampal-inspired algorithms. I will present our memristor synaptic devices, the integration with CMOS neuronal circuitry as well as our modular mixed-signal prototyping platform which was collaboratively designed for benchmarking memristive neural networks of up to 20,000 memristor devices. I will then describe a small-scale CA3-inspired network with neuronal diversity and biologically-realistic resting state dynamics co-designed on memristor hardware. Benefiting from memristor noise, the hardware implementation shows continuous periodic behavior, outperforming simulated hardware.

Gina Adam is an associate professor with the Electrical and Computer Engineering department at George Washington University. Her group works on the development of emerging non-volatile memory devices and novel hardware foundations that will enable new ways of neuro-inspired computing. She received her Ph.D. in electrical and computer engineering from the University of California Santa Barbara in 2015 and was a research scientist at the Romanian National Institute for Research and Development in Microtechnologies and a visiting scholar at École Polytechnique Fédérale de Lausanne before joining GWU. She was the recipient of an International Fulbright Science and Technology award in 2010, a Mirzayan fellowship at the National Academy of Engineering in 2012, a H2020 Marie Sklodowska-Curie grant from the European Commission in 2016, a NSF CRII award in 2020, and AFOSR YIP and NSF CAREER awards in 2023 and DOE Early Career Research Program Award in 2024.