Chapter

High Performance Embedded Architectures and Compilers

Volume 5409 of the series Lecture Notes in Computer Science pp 19-33

Predictive Runtime Code Scheduling for Heterogeneous Architectures

  • Víctor J. JiménezAffiliated withBarcelona Supercomputing Center (BSC)
  • , Lluís VilanovaAffiliated withDepartament d’Arquitectura de Computadors (UPC)
  • , Isaac GeladoAffiliated withDepartament d’Arquitectura de Computadors (UPC)
  • , Marisa GilAffiliated withDepartament d’Arquitectura de Computadors (UPC)
  • , Grigori FursinAffiliated withALCHEMY Group, INRIA Futurs and LRI, Paris-Sud University
  • , Nacho NavarroAffiliated withDepartament d’Arquitectura de Computadors (UPC)

* Final gross prices may vary according to local VAT.

Get Access

Abstract

Heterogeneous architectures are currently widespread. With the advent of easy-to-program general purpose GPUs, virtually every recent desktop computer is a heterogeneous system. Combining the CPU and the GPU brings great amounts of processing power. However, such architectures are often used in a restricted way for domain-specific applications like scientific applications and games, and they tend to be used by a single application at a time. We envision future heterogeneous computing systems where all their heterogeneous resources are continuously utilized by different applications with versioned critical parts to be able to better adapt their behavior and improve execution time, power consumption, response time and other constraints at runtime. Under such a model, adaptive scheduling becomes a critical component.

In this paper, we propose a novel predictive user-level scheduler based on past performance history for heterogeneous systems. We developed several scheduling policies and present the study of their impact on system performance. We demonstrate that such scheduler allows multiple applications to fully utilize all available processing resources in CPU/GPU-like systems and consistently achieve speedups ranging from 30% to 40% compared to just using the GPU in a single application mode.