Single-Pass Memory System Evaluation for Multiprogramming Workloads

Abstract

Modern memory systems are composed of levels of cache memories, a virtual memory system, and a backing store. Varying more than a few design parameters and measuring the performance of such systems has traditionally be constrained by the high cost of simulation. Models of cache performance recently introduced reduce the cost simulation but at the expense of accuracy of performance prediction. Stack-based methods predict performance accurately using one pass over the trace for all cache sizes, but these techniques have been limited to fully-associative organizations. This paper presents a stack-based method of evaluating the performance of cache memories using a recurrence/ conflict model for the miss ratio. Unlike previous work, the performance of realistic cache designs, such as direct-mapped caches, are predicted by the method. The method also includes a new approach to the problem of the effects of multiprogramming. This new technique separates the characteristics of the individual program from that of the workload. The recurrence/conflict method is shown to be practical, general, and powerful by comparing its performance to that of a popular traditional cache simulator. The authors expect that the availability of such a tool will have a large impact on future architectural studies of memory systems.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
May 01, 1990
Accession Number
ADA222808

Entities

People

  • Thomas M. Conte
  • Wen-mei Hwu

Organizations

  • University of Illinois Urbana–Champaign

Tags

Communities of Interest

  • Space

DTIC Thesaurus Topics

  • Algorithms
  • Classification
  • Compilers
  • Computer Programming
  • Computer Programs
  • Computers
  • Engineering
  • Hierarchies
  • High Performance Computing
  • Illinois
  • Operating Systems
  • Parallel Processing
  • Scheduling (Production)
  • Security
  • Simulations
  • Simulators
  • Universities

Fields of Study

  • Computer science

Readers

  • Computational Modeling and Simulation
  • Parallel and Distributed Computing.