Entropy-based profile characterization and classification for automatic profile management

Authors:
Jinpyo Kim;Wei-Chung Hsu;Pen-Chung Yew;Sreekumar R. Nair;Robert Y. Geva
Affiliations:
Department of Computer Science and Engineering, University of Minnesota, Minneapolis, MN;Department of Computer Science and Engineering, University of Minnesota, Minneapolis, MN;Department of Computer Science and Engineering, University of Minnesota, Minneapolis, MN;Intel, Santa Clara, CA;Intel, Santa Clara, CA
Venue:
ACSAC'07 Proceedings of the 12th Asia-Pacific conference on Advances in Computer Systems Architecture
Year:
2007

Citing 6
Cited 0

Elements of information theory

Elements of information theory
System support for automatic profiling and optimization

Proceedings of the sixteenth ACM symposium on Operating systems principles
The Jalapeño dynamic optimizing compiler for Java

JAVA '99 Proceedings of the ACM 1999 conference on Java Grande
Adaptive optimization in the Jalapeño JVM

OOPSLA '00 Proceedings of the 15th ACM SIGPLAN conference on Object-oriented programming, systems, languages, and applications
The Performance of Runtime Data Cache Prefetching in a Dynamic Optimization System

Proceedings of the 36th annual IEEE/ACM International Symposium on Microarchitecture
Ispike: A Post-link Optimizer for the Intel®Itanium®Architecture

Proceedings of the international symposium on Code generation and optimization: feedback-directed and runtime optimization

Quantified Score

Hi-index	0.00

Visualization

Abstract

The recent adoption of pre-JIT compilation for the JVM and .NET platforms allows the exploitation of continuous profile collection and management at user sites. To support efficient pre-JIT type of compilation, this paper proposes and studies an entropy-based profile characterization and classification method. This paper first shows that highly accurate profiles can be obtained by merging a number of profiles collected over repeated executions with relatively low sampling frequency for the SPEC CPU2000 benchmarks. It also shows that simple characterization of the profile with information entropy can be used to guide sampling frequency of the profiler in an autonomous fashion. On the SPECjbb2000 benchmark, our adaptive profiler obtains a very accurate profile (94.5% similar to the baseline profile) with only 8.7% of the samples that would normally be collected using a 1M instructions sampling interval. Furthermore, we show that entropy could also be used for classifying different program behaviors based on different input sets.