IPME and External Clients: Enhancing Performance by Offloading Simulation Workload to External Clients; Explaining and Simplifying the Process

Abstract

The speed of simulation execution as performed by the Integrated Performance Modelling Environment (IPME) degrades significantly when computationally demanding functions, such as iterative mathematical calculations, are included in a model. Much of this degradation can be mitigated by transferring those functions that require significant computational resources (cpu processing) to an external client using the external client architecture that accompanies IPME. This client communicates with IPME using TCP/IP network protocol, exchanging values of common variables over the network. Using an external client allows more processing power to be dedicated to computationally expensive tasks and are generally more robust to increases in computational demand. A computationally demanding sample client is used to show the execution performance differences when the procedure resides within the IPME task network and when it is offloaded to an external client. This report also outlines how to use the sample client source code to build a client program, extending the developer's approach for client program development to create a more flexible interface. This will provide another option for client developers who may prefer a more intuitive template for developing their customized client than the sample client provided with IPME.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Dec 01, 2007
Accession Number
ADA479120

Entities

People

  • Brad Cain
  • Ter Haar Phil

Organizations

  • Defence Research and Development Canada

Tags

Communities of Interest

  • Human Systems
  • Materials and Manufacturing Processes

DTIC Thesaurus Topics

  • C Programming Language
  • Classification
  • Computer Programming
  • Computer Programs
  • Computing System Architectures
  • Degradation
  • Engineering
  • Environment
  • Language
  • Load Monitoring
  • Models
  • National Security
  • Network Protocols
  • Programming Languages
  • Security
  • Simulations
  • Workload

Fields of Study

  • Computer science

Readers

  • Parallel and Distributed Computing.
  • Systems Analysis and Design
  • Team-Based Human-Centered Cognitive Task Decision Making and Information Performance.