Scalable Construction of Spiking Neural Networks using up to thousands of GPUs

TL;DR

This paper introduces a scalable method for constructing large spiking neural networks on multi-GPU clusters, enabling efficient simulation of brain-like models at unprecedented scale.

Contribution

It presents a novel network construction approach tailored for MPI-based multi-GPU systems, facilitating large-scale neural simulations.

Findings

Achieved scalable performance on cortical models using point-to-point communication.

Demonstrated effective scaling with collective communication methods.

Enabled simulation of networks with billions of synapses.

Abstract

Diverse scientific and engineering research areas deal with discrete, time-stamped changes in large systems of interacting delay differential equations. Simulating such complex systems at scale on high-performance computing clusters demands efficient management of communication and memory. Inspired by the human cerebral cortex -- a sparsely connected network of $\mathcal{O}(10^{10})$ neurons, each forming $\mathcal{O}(10^{3})$--$\mathcal{O}(10^{4})$ synapses and communicating via short electrical pulses called spikes -- we study the simulation of large-scale spiking neural networks for computational neuroscience research. This work presents a novel network construction method for multi-GPU clusters and upcoming exascale supercomputers using the Message Passing Interface (MPI), where each process builds its local connectivity and prepares the data structures for efficient spike exchange across the cluster during state propagation. We demonstrate scaling performance of two cortical models using point-to-point and collective communication, respectively.