Engineering concurrent software guided by statistical performance analysis

Autoren Bernd Scheuermann
Kevin Hammond
Alex Shaferenko
Chris Jesshope
Heinz Hertlein
Volkmar Wieser
Philip Hölzenspies
Raimund Kirner
Te Boekhorst Iraneus
Editoren K. De Bosschere
E.H. D'Hollander
G.R. Joubert
D. Padua
F. Peters
TitelEngineering concurrent software guided by statistical performance analysis
BuchtitelApplications, Tools and Techniques on the Road to Exascale Computing
Typin Sammelband
VerlagIOS Press
SerieAdvances in Parallel Computing
Band22
AbteilungKVS
ISBN978-1-61499-040-6
MonatMay
Jahr2012
Seiten385-394
SCCH ID#1145
Abstract

This paper introduces the ADVANCE approach to engineering concurrent systems using a new component-based approach. A cost-directed tool-chain maps concurrent programs onto emerging hardware architectures, where costs are expressed in terms of programmer annotations for the throughput, latency and jitter of components. These are then synthesized using advanced statistical analysis techniques to give overall cost information about the concurrent system that can be exploited by the hardware virtualization layer to drive mapping and scheduling decisions. Initial performance results are presented, showing that the ADVANCE technologies provide a promising approach to dealing with near- and future-term complexities of programming heterogeneous multi-core systems.