Thread scheduling for cache locality

James Philbin, Jan Edler, Otto J. Anshus, Craig C. Douglas, Kai Li

Research output: Contribution to journalArticle

Abstract

This paper describes a method to improve the cache locality of sequential programs by scheduling fine-grained threads. The algorithm relies upon hints provided at the time of thread creation to determine a thread execution order likely to reduce cache misses. This technique may be particularly valuable when compiler-directed tiling is not feasible. Experiments with several application programs, on two systems with different cache structures, show that our thread scheduling method can improve program performance by reducing second-level cache misses.

Original languageEnglish (US)
Pages (from-to)60-71
Number of pages12
JournalSIGPLAN Notices (ACM Special Interest Group on Programming Languages)
Volume31
Issue number9
StatePublished - Sep 1996
Externally publishedYes

Fingerprint

Scheduling
Application programs
Experiments

ASJC Scopus subject areas

  • Computer Graphics and Computer-Aided Design
  • Software

Cite this

Philbin, J., Edler, J., Anshus, O. J., Douglas, C. C., & Li, K. (1996). Thread scheduling for cache locality. SIGPLAN Notices (ACM Special Interest Group on Programming Languages), 31(9), 60-71.

Thread scheduling for cache locality. / Philbin, James; Edler, Jan; Anshus, Otto J.; Douglas, Craig C.; Li, Kai.

In: SIGPLAN Notices (ACM Special Interest Group on Programming Languages), Vol. 31, No. 9, 09.1996, p. 60-71.

Research output: Contribution to journalArticle

Philbin, J, Edler, J, Anshus, OJ, Douglas, CC & Li, K 1996, 'Thread scheduling for cache locality', SIGPLAN Notices (ACM Special Interest Group on Programming Languages), vol. 31, no. 9, pp. 60-71.
Philbin J, Edler J, Anshus OJ, Douglas CC, Li K. Thread scheduling for cache locality. SIGPLAN Notices (ACM Special Interest Group on Programming Languages). 1996 Sep;31(9):60-71.
Philbin, James ; Edler, Jan ; Anshus, Otto J. ; Douglas, Craig C. ; Li, Kai. / Thread scheduling for cache locality. In: SIGPLAN Notices (ACM Special Interest Group on Programming Languages). 1996 ; Vol. 31, No. 9. pp. 60-71.
@article{442e80c977b54572a16c989b0c8c7edd,
title = "Thread scheduling for cache locality",
abstract = "This paper describes a method to improve the cache locality of sequential programs by scheduling fine-grained threads. The algorithm relies upon hints provided at the time of thread creation to determine a thread execution order likely to reduce cache misses. This technique may be particularly valuable when compiler-directed tiling is not feasible. Experiments with several application programs, on two systems with different cache structures, show that our thread scheduling method can improve program performance by reducing second-level cache misses.",
author = "James Philbin and Jan Edler and Anshus, {Otto J.} and Douglas, {Craig C.} and Kai Li",
year = "1996",
month = "9",
language = "English (US)",
volume = "31",
pages = "60--71",
journal = "ACM SIGPLAN Notices",
issn = "1523-2867",
publisher = "Association for Computing Machinery (ACM)",
number = "9",

}

TY - JOUR

T1 - Thread scheduling for cache locality

AU - Philbin, James

AU - Edler, Jan

AU - Anshus, Otto J.

AU - Douglas, Craig C.

AU - Li, Kai

PY - 1996/9

Y1 - 1996/9

N2 - This paper describes a method to improve the cache locality of sequential programs by scheduling fine-grained threads. The algorithm relies upon hints provided at the time of thread creation to determine a thread execution order likely to reduce cache misses. This technique may be particularly valuable when compiler-directed tiling is not feasible. Experiments with several application programs, on two systems with different cache structures, show that our thread scheduling method can improve program performance by reducing second-level cache misses.

AB - This paper describes a method to improve the cache locality of sequential programs by scheduling fine-grained threads. The algorithm relies upon hints provided at the time of thread creation to determine a thread execution order likely to reduce cache misses. This technique may be particularly valuable when compiler-directed tiling is not feasible. Experiments with several application programs, on two systems with different cache structures, show that our thread scheduling method can improve program performance by reducing second-level cache misses.

UR - http://www.scopus.com/inward/record.url?scp=17044404702&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=17044404702&partnerID=8YFLogxK

M3 - Article

AN - SCOPUS:17044404702

VL - 31

SP - 60

EP - 71

JO - ACM SIGPLAN Notices

JF - ACM SIGPLAN Notices

SN - 1523-2867

IS - 9

ER -