View on GitHub

BrainGrid

A project to facilitate construction of high-performance neural simulations

1. Introduction

1.1 What is BrainGrid?

BrainGrid is an open-source spiking neural network simulator that is intended to aid scientists and researchers by providing pre-built code that can be easily modified to fit different models. The program also offers the ability to easily transition to a GPU-centered model with little additional work, which can provide a potential speedup of up to a twentieth of the original runtime.

Our current focus is on single-threaded simulators and simulators running on GPUs using NVIDIA’s CUDA libraries. We expect to shortly thereafter roll out support for multi-cluster, and potentially multi-threading using OpenMP in the future.

1.2 What is BrainGrid for?

BrainGrid is a framework for quickly implementing GPU-speed-up spiking neural networks. If you have a mathematical model for a spiking neural network, or if you have an idea for one, BrainGrid will provide you with a quick way of getting that network implemented in code and running on CPUs or GPUs.

In sum, BrainGrid is designed to do the following:

  1. Simulate spiking neural networks with built-in models (such as leaky-integrate-and-fire neuron, izhikevich neuron)
  2. Provide a framework to make it easier to code models of your own
  3. Make these models FAST - provide code to make it easy to migrate models to single or multiple GPUs, providing a huge speedup from single or multi-threaded host-side models.

1.3 Why do we need BrainGrid?

The initial principles that we are basing BrainGrid on are as follows:


» Next: 2. Installation


« Go back to BrainGrid Home page