ICS'97 - Workshop: Performance Data Mining: Automated Diagnosis, Adaption, and Optimization

Organizers:

Alois Ferscha, University of Vienna, Austria
ferscha@ani.univie.ac.at
Prof. Allen D. Malony,University of Oregon, Eugene, OR, USA
malony@cs.uoregon.edu

Monday, July 7, 9:00 - 16:30

Abstract

The development of performance measurement and analysis techniques and tools for parallel and distributed supercomputers has made it possible to capture a wealth of data about application and system performance behavior. This data embodies the effects of interacting, performance factors found in the program, its algorithms, the architecture and hardware, and the system software, whose interdependent performance relationships grow ever more complex as the supercomputing environment increases in sophistication. Nevertheless, the user is still, for the most part, placed in the central decision-making role in the use of the techniques/tools, the interpretation of the resulting performance information, and the guidance for program or system modification.

Recent work has sought to move human decision-making out of the performance measurement-diagnosis-optimization loop by employing "intelligent" methods based on automated performance measurement, knowledge-based diagnosis frameworks, online, adaptive performance control, and predictive performance models built from detailed empirical analysis. The term "performance data mining" is used to characterize this work.

Workshop Schedule

9:00 - 9:15 Welcome
Alois Ferscha, University of Vienna, Austria
9:15 - 9:45 Introduction
Allen D. Malony, Univ. of Oregon, USA
Slides, 77 K, (gzipped, uuencoded 27 K)
9:45 - 10:30 Performance Optimization of Distributed Applications in an Extensible, Adaptive Environment
Diane Rover, Michigan State University, USA
Paper, 459 K, (gzipped, uuencoded 124 K)
10:30 - 11:00 Coffee Break
11:00 - 11:45 Performance Analysis of HPF+ Kernels
Maria Calzarossa, University of Pavia, Italy
Abstract, 40 K
11:45 - 12:30 Optimistic Network Computing and its Performance Control
Steve Turner, University of Exeter, UK
Paper, 134 K, (gzipped, uuencoded 77 K)
12:30 - 14:00 Lunch Break
14:00 - 14:45 Performance Diagnosis is Dynamic Data Mining
Allen D. Malony, University of Oregon, USA
Slides, 168 K, (gzipped, uuencoded 64 K)
14:45 - 15:30 The Autopilot Performance-Directed Adaptive Control System
Daniel Reed, University of Illinois, Urbana-Champaign, USA
Paper, 194 K, (gzipped, uuencoded 88 K)
15:30 - 16:00 Coffee Break
16:00 - 16:30 wrap-up, Closing
Last Modified: 11:40am MET DST, July 06, 1997