# Multi-GPU SNN Simulation with Static Load Balancing

arXiv: https://arxiv.org/abs/2102.04681 · HTML: https://mlsanity.com/paper/2102.04681/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 6 of its 51 references that are in the map.
- First posted on arXiv 9 Feb 2021.

## Builds on (most cited first, 6 of 6)

- [BindsNET: A machine learning-oriented spiking neural networks library in Python](https://mlsanity.com/paper/1806.01423/) (2018)
- [SpykeTorch: Efficient Simulation of Convolutional Spiking Neural Networks with at most one Spike per Neuron](https://mlsanity.com/paper/1903.02440/) (2019)
- [BrainFrame: A node-level heterogeneous accelerator platform for neuron simulations](https://mlsanity.com/paper/1612.01501/) (2016)
- [Faster and Simpler SNN Simulation with Work Queues](https://mlsanity.com/paper/1912.07423/) (2019)
- [Fast simulations of highly-connected spiking cortical models using GPUs](https://mlsanity.com/paper/2007.14236/) (2020)
- [Library-based Fast Algorithm for Simulating the Hodgkin-Huxley Neuronal Networks](https://mlsanity.com/paper/2101.07257/) (2021)

## Built on by (most cited first, 1 of 1)

- [Even Faster SNN Simulation with Lazy+Event-driven Plasticity and Shared Atomics](https://mlsanity.com/paper/2107.04092/) (2021)

## Abstract (arXiv)

We present a SNN simulator which scales to millions of neurons, billions of synapses, and 8 GPUs. This is made possible by 1) a novel, cache-aware spike transmission algorithm 2) a model parallel multi-GPU distribution scheme and 3) a static, yet very effective load balancing strategy. The simulator further features an easy to use API and the ability to create custom models. We compare the proposed simulator against two state of the art ones on a series of benchmarks using three well-established models. We find that our simulator is faster, consumes less memory, and scales linearly with the number of GPUs.