• Corpus ID: 1624143

Scaling to 1024 software processes and hardware cores of the distributed simulation of a spiking neural network including up to 20G synapses

@article{Pastorelli2015ScalingT1,
  title={Scaling to 1024 software processes and hardware cores of the distributed simulation of a spiking neural network including up to 20G synapses},
  author={Elena Pastorelli and Pier Stanislao Paolucci and Roberto Ammendola and Andrea Biagioni and Ottorino Frezza and Francesca Lo Cicero and Alessandro Lonardo and Michele Martinelli and Francesco Simula and Piero Vicini},
  journal={ArXiv},
  year={2015},
  volume={abs/1511.09325}
}
This short report describes the scaling, up to 1024 software processes and hardware cores, of a distributed simulator of plastic spiking neural networks. A previous report demonstrated good scalability of the simulator up to 128 processes. Herein we extend the speed-up measurements and strong and weak scaling analysis of the simulator to the range between 1 and 1024 software processes and hardware cores. We simulated two-dimensional grids of cortical columns including up to ~20G synapses… 

Figures and Tables from this paper

Impact of exponential long range and Gaussian short range lateral connectivity on the distributed simulation of neural networks including up to 30 billion synapses

A preliminary measure of the impact of exponentially decaying lateral connectivity compared to that of shorter-range Gaussian decays on the scaling behaviour and memory occupation of a distributed spiking neural network simulator (DPSNN) is provided.

Deliverable D 2 . 1 Census of the Applications version 1 . 2 – 16 June 2016

This document is owned and copyrighted as specified here. Permission to make digital or hard copies, or to post on web servers this document in its entirety without fee is granted provided that: (i)

References

SHOWING 1-10 OF 15 REFERENCES

Spiking network simulation code for petascale computers

It is shown that the novel architecture presented here scales to the largest petascale supercomputers available today, and quantitatively discusses the performance on two supercomputing systems.

Power, Energy and Speed of Embedded and Server Multi-Cores applied to Distributed Simulation of Spiking Neural Networks: ARM in NVIDIA Tegra vs Intel Xeon quad-cores

This short note regards a comparison of instantaneous power, total energy consumption, execution time and energetic cost per synaptic event of a spiking neural network simulator (DPSNN-STDP)

Efficient Event-Driven Simulation of Large Networks of Spiking Neurons and Dynamical Synapses

The main impact of the new approach is a drastic reduction of the computational load incurred upon introduction of dynamic synaptic efficacies, which vary organically as a function of the activities of the pre- and postsynaptic neurons.

Diverse population-bursting modes of adapting spiking neurons.

Two different population-bursting regimes emerge, depending on the level of SFA in networks with noisy emission rate, due to the finite number of neurons, in networks as an effective nonlinear two-dimensional system.

A novel slow (< 1 Hz) oscillation of neocortical neurons in vivo: depolarizing and hyperpolarizing components

A novel slow oscillation in intracellular recordings from cortical association areas 5 and 7, motor areas 4 and 6, and visual areas 17 and 18 of cats under various anesthetics is described and synchronous inhibitory periods in both neurons are demonstrated.

Slow and fast rhythms generated in the cerebral cortex of the anesthetized mouse.

This study allows us to carry out interareal comparisons and provides a baseline to compare against cortical emerging activity from genetically altered animals.

Making Waves: Initiation and Propagation of Corticothalamic Ca2+ Waves In Vivo

The Evolution of the Brain, the Human Nature of Cortical Circuits, and Intellectual Creativity

The issue of the authors' humanity is discussed from a neurobiological and historical perspective because of the increase in size and complexity of their brains and the addition of microcircuits with a similar basic structure.

Distributed simulation of polychronous and plastic spiking neural networks: strong and weak scaling of a representative mini-application benchmark executed on a small-scale commodity cluster

A natively distributed mini-application benchmark representative of plastic spiking neural network simulators, designed to generate spiking behaviors and synaptic connectivity that do not change when the number of hardware processing nodes is varied, simplifying the quantitative study of scalability on commodity and custom architectures.