Simple and Effective Distributed Computing with a Scheduling Service

TL;DR

This paper presents a simple, low-cost method for setting up distributed computing using standard networked PCs, significantly increasing throughput without special hardware or software modifications.

Contribution

It introduces an easy-to-implement scheduling service that leverages existing PCs for high-performance computing, requiring minimal setup and no specialized hardware or software.

Findings

Performance can rival supercomputers with enough PCs

Setup involves little or no re-coding of applications

Operates almost invisibly to PC owners

Abstract

High-throughput computing projects require the solution of large numbers of problems. In many cases, these problems can be solved on desktop PCs, or can be broken down into independent "PC-solvable" sub-problems. In such cases, the projects are high-performance computing projects, but only because of the sheer number of the needed calculations. We briefly describe our efforts to increase the throughput of one such project. We then explain how to easily set up a distributed computing facility composed of standard networked PCs running Windows 95, 98, 2000, or NT. The facility requires no special software or hardware, involves little or no re-coding of application software, and operates almost invisibly to the owners of the PCs. Depending on the number and quality of PCs recruited, performance can rival that of supercomputers.