Phase Transition in Multiprocessor Scheduling

Bauke, Heiko; Mertens, Stephan; Engel, Andreas

doi:10.1103/PhysRevLett.90.158701

Full-text links:

Download:

(license)

Current browse context:

cond-mat

< prev | next >

new | recent | 0208

References & Citations

NASA ADS

Bookmark

(what is this?)

Condensed Matter > Disordered Systems and Neural Networks

Title: Phase Transition in Multiprocessor Scheduling

Authors: Heiko Bauke, Stephan Mertens, Andreas Engel

(Submitted on 5 Aug 2002 (v1), last revised 21 Aug 2002 (this version, v2))

Abstract: The problem of distributing the workload on a parallel computer to minimize the overall runtime is known as Multiprocessor Scheduling Problem. It is NP-hard, but like many other NP-hard problems, the average hardness of random instances displays an ``easy-hard'' phase transition. The transition in Multiprocessor Scheduling can be analyzed using elementary notions from crystallography (Bravais lattices) and statistical mechanics (Potts vectors). The analysis reveals the control parameter of the transition and its critical value including finite size corrections. The transition is identified in the performance of practical scheduling algorithms.

Comments:	6 pages, revtex4
Subjects:	Disordered Systems and Neural Networks (cond-mat.dis-nn); Statistical Mechanics (cond-mat.stat-mech)
Journal reference:	Physical Review Letters, vol. 90, 158701 (2003)
DOI:	10.1103/PhysRevLett.90.158701
Cite as:	arXiv:cond-mat/0208081 [cond-mat.dis-nn]
	(or arXiv:cond-mat/0208081v2 [cond-mat.dis-nn] for this version)

Submission history

From: Stephan Mertens [view email]
[v1] Mon, 5 Aug 2002 11:29:01 GMT (35kb)
[v2] Wed, 21 Aug 2002 08:22:53 GMT (35kb)

Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?)

Link back to: arXiv, form interface, contact.