When
Where
Speaker: Felix Wang (Sandia National Laboratories)
Title: Scaling Neuromorphic Simulation for Biological Connectomes
Abstract: We provide an overview of neuromorphic computing, their potential advantages, and some associated spiking neural network algorithms. Neuromorphic computing is a class of technologies that are designed according to brain-inspired principles such as massively parallel processing and asynchronous event-based communication. We also discuss some of our recent work in mapping and simulating biologically realistic connectomes on scalable digital neuromorphic hardware platforms. In particular, we implement the whole-brain connectome of the adult Drosophila melanogaster (fruit fly) from the FlyWire Consortium containing 140K neurons and 50M synapses on the Intel Loihi 2 neuromorphic platform. We describe the challenges and solutions of this task that come from the sparse, recurrent, and irregular connectivity structure of biological networks as well as neuromorphic hardware constraints, such as fan-in and fan-out memory limitations. Significantly, we achieve a neuromorphic implementation that is orders of magnitude faster than numerical simulations on conventional hardware, and we also find that performance advantages increase with sparser activity. These results affirm that today's scalable neuromorphic platforms are capable of implementing and accelerating biologically realistic models -- a key enabling technology for advancing neuro-inspired AI and computational neuroscience.