A Discrete-event-based Simulator for Distributed Deep Learning

TL;DR

This paper introduces sim4DistrDL, a discrete-event simulator designed to evaluate distributed deep learning systems, addressing the lack of specialized simulation tools for DNN-based distributed applications.

Contribution

The paper presents a novel discrete-event simulation framework that integrates deep learning and network modules for distributed deep learning environments.

Findings

Enables simulation of distributed deep learning configurations

Facilitates early-stage scalability assessment of intelligence applications

Supports analysis of parameter configuration effects

Abstract

New intelligence applications are driving increasing interest in deploying deep neural networks (DNN) in a distributed way. To set up distributed deep learning involves alterations of a great number of the parameter configurations of network/edge devices and DNN models, which are crucial to achieve best performances. Simulations measure scalability of intelligence applications in the early stage, as well as to determine the effects of different configurations, thus highly desired. However, work on simulating the distributed intelligence environment is still in its infancy. The existing simulation frameworks, such as NS-3, etc., cannot extended in a straightforward way to support simulations of distributed learning. In this paper, we propose a novel discrete event simulator, sim4DistrDL, which includes a deep learning module and a network simulation module to facilitate simulation of DNN-based distributed applications. Specifically, we give the design and implementation of the proposed learning simulator and present an illustrative use case.