Compilers for high performance

Topic 4


This topic deals with all subjects concerning the compilation of programs for high performance architectures, from general-purpose platforms to specific hardware designs such as Processing-In-Memory (PIM)-based systems or low-power embedded systems. This includes language aspects, program analysis, program transformation systems for optimization of resource utilization, the interaction between compiler and hardware/operating system, runtime systems, and feedback-oriented and adaptive compilation.
A non-exclusive selection of standard issues is listed below. We specifically welcome submissions that deal with the interface between compiler and runtime system for advanced irregular and adaptive applications and approaches towards dynamic, iterative, and adaptive compilation systems that interact with performance analysis and prediction tools. We include all types of target programming styles, from conventional imperative languages to data parallel, object-oriented, and functional styles. We also invite submissions on practical experiences, in particular case studies to assess the benefits, limitations, and failures of current compiler and runtime systems for high performance computing.

Global Chair:
Prof. Hans P. Zima
Institute for Software Science, University of Vienna, Austria,
and Jet Propulsion Laboratory, California Institute of Technology,
Pasadena, California, USA

Vice Chair:
Prof. Siegfried Benkner
Institute for Software Science
University of Vienna, Austria

Vice Chair:
Dr. Michael O'Boyle
University of Edinburgh, Edinburgh, UK

Local Chair:
Prof. Beniamino Di Martino
Dipartimento di Ingegneria dell'Informazione
Via Roma, 29 - 81031 Aversa (CE) - Italy

Last modified: Fri Nov 28 19:28:34 CET 2003
Next topic Previous topic