Granular layEr Simulator: Design and Multi-GPU Simulation of the Cerebellar Granular Layer.

Giordana Florimbi,Stefano Masoli,Emanuele Torti,Francesco Leporati,Egidio D'Angelo

doi:10.3389/fncom.2021.630795

Abstract

In modern computational modeling, neuroscientists need to reproduce long-lasting activity of large-scale networks, where neurons are described by highly complex mathematical models. These aspects strongly increase the computational load of the simulations, which can be efficiently performed by exploiting parallel systems to reduce the processing times. Graphics Processing Unit (GPU) devices meet this need providing on desktop High Performance Computing. In this work, authors describe a novel Granular layEr Simulator development implemented on a multi-GPU system capable of reconstructing the cerebellar granular layer in a 3D space and reproducing its neuronal activity. The reconstruction is characterized by a high level of novelty and realism considering axonal/dendritic field geometries, oriented in the 3D space, and following convergence/divergence rates provided in literature. Neurons are modeled using Hodgkin and Huxley representations. The network is validated by reproducing typical behaviors which are well-documented in the literature, such as the center-surround organization. The reconstruction of a network, whose volume is 600 × 150 × 1,200 μm3 with 432,000 granules, 972 Golgi cells, 32,399 glomeruli, and 4,051 mossy fibers, takes 235 s on an Intel i9 processor. The 10 s activity reproduction takes only 4.34 and 3.37 h exploiting a single and multi-GPU desktop system (with one or two NVIDIA RTX 2080 GPU, respectively). Moreover, the code takes only 3.52 and 2.44 h if run on one or two NVIDIA V100 GPU, respectively. The relevant speedups reached (up to ~38× in the single-GPU version, and ~55× in the multi-GPU) clearly demonstrate that the GPU technology is highly suitable for realistic large network simulations.

Highlights

The challenge to understand, reproduce and simulate the human brain activities needs more and more High-Performance Computing (HPC) support, in particular, where heterogeneous elements, described by complex mathematical models, have to be simulated as fast as possible (Bouchard et al, 2016)
Starting from the simple Leaky Integrate and Fire (LIF) model up to the more complex Hodgkin Huxley (HH) one, all the mathematical representations are characterized by a variable number of differential equations, which strongly increases the computational load of the simulations (Izhikevich, 2004)
The system is equipped with two NVIDIA RTX 2080 Graphics Processing Unit (GPU) (Turing architecture), each one with 2944 CUDA cores, 8 GB of DDR6 memory and working at 1.8 GHz

Summary

Introduction

The challenge to understand, reproduce and simulate the human brain activities needs more and more High-Performance Computing (HPC) support, in particular, where heterogeneous elements, described by complex mathematical models, have to be simulated as fast as possible (Bouchard et al, 2016). Several research groups work on reproducing the functionalities of very large areas of the brain (Beyeler et al, 2014; Cremonesi and Schürmann, 2020) To this aim, they need multicore and/or manycore technologies capable of reducing the processing time and of ensuring the power, memory, and storage capabilities offered by HPC solutions (Fidjeland et al, 2013). They need multicore and/or manycore technologies capable of reducing the processing time and of ensuring the power, memory, and storage capabilities offered by HPC solutions (Fidjeland et al, 2013) Another aspect to consider is the model to use for the neuron representation and the detailed morphologies introduced in the network. Particular attention should be given to the time integration step that directly determines the number of times that the differential equations have to be solved

Objectives

Methods

Results

Conclusion