1
Efficient Program Scheduling for Heterogeneous
Multi-core Processors
Jian Chen and Lizy K. John
ECE Department, University of Texas at Austin
Austin, TX 78712, USA
chenjian@mail.utexas.edu, ljohn@ece.utexas.edu
ABSTRACT
Heterogeneous multicore processors promise high execution
efficiency under diverse workloads, and program scheduling is
critical in exploiting this efficiency. This paper presents a novel
method to leverage the inherent characteristics of a program for
scheduling decisions in heterogeneous multicore processors. The
proposed method projects the core’s configuration and the program’s
resource demand to a unified multi-dimensional space, and uses
weighted Euclidean distance between these two to guide the
program scheduling. The experimental results show that on average,
this distance based scheduling heuristic achieves 24.5% reduction in
energy delay product, 6.1% reduction in energy, and 9.1%
improvement in throughput when compared with traditional
hardware oblivious scheduling algorithm.
Categories and Subject Descriptors: C.1.3 [Processor
Architectures]: Other Architecture Styles – Hybrid Systems
General Terms: Design, Performance
Key Words: Heterogeneous Multi-core, Energy-Delay Product,
Program Scheduling
1. INTRODUCTION
Heterogeneous multicore processors (HMP) have been demonstrated
to be an attractive design alternative to its homogeneous counterpart,
as it has a unique advantage in improving both system throughput
and execution efficiency. Although most of the existing multicore
processors are homogeneous, this design paradigm leads to an
inevitable dilemma. That is, replicating smaller cores compromises
the throughput of the high-complexity single-threaded applications;
whereas replicating larger cores sacrifices the execution efficiency
of the low-complexity low-priority threads. The HMP, however,
integrates cores of different types or complexities in a single chip,
and hence is able to address both throughput and efficiency for
various workloads