Exploring the Benefits of Multiple Hardware Contexts in a Multiprocessor Architecture: Preliminary Results

Abstract

A fundamental problem that any scalable multiprocessor must address is the ability to tolerate high latency memory operations. This paper explores the extent to which multiple hardware contexts per processor can help to mitigate the negative effects of high latency. In particular, we evaluate the performance of a directory-based cache coherent multiprocessor using memory reference traces obtained from three parallel applications. We explore the case where there are a small fixed number (2-4) of hardware contexts per processor and the context switch overhead is low. In contrast to previously proposed approaches, we also use a simple context-switch criterion, namely a cache miss or a write-hit so shared data. Our results show that the effectiveness of multiple contexts depends on the nature of the applications, the context switch overhead, and the inherent latency of the machine architecture. Given reasonably low overhead hardware context switches, we show that two of four contexts can achieve substantial performance gains over a single context. For one application, the processor utilization increased by about 65% with two contexts and by about 100% with four contexts.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Nov 16, 1988
Accession Number
ADA210349

Entities

People

  • Anoop Gupta
  • Wolf-detrich Weber

Organizations

  • Stanford University

Tags

Communities of Interest

  • Materials and Manufacturing Processes

DTIC Thesaurus Topics

  • Computer Programming
  • Computers
  • Computing System Architectures
  • Crossbar Switches
  • Directories
  • Efficiency
  • Instruction Set Architecture
  • Instructions
  • Models
  • Multiprocessors
  • Personal Computers
  • Pipelines
  • Simulations
  • Simulators
  • Switches
  • Teleoperation
  • Three Dimensional

Fields of Study

  • Computer science

Readers

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