Blind Justice Defeats the Multi-Armed Locality Bandit

Tuesday, November 4, 2008 - 12:04am

3:30 – 4:30
Computer Science Conference Room, Harold Frank Hall Rm. 1132


SPEAKER: Martin Hirzel (Part 1)
IBM Research

Title: Blind Justice Defeats the Multi-Armed Locality Bandit

Abstract: Good data layouts improve cache and TLB locality of
object-oriented software, but unfortunately, selecting an optimal data
layout a priori is NP-hard. This talk presents layout auditing, a just
technique for selecting the best among a set of layouts online (while
the program is running). Layout auditing blindly applies different
layouts over time and observes their performance. As it becomes
confident about which layout performs best, it selects that layout with
higher probability. But if a phase shift causes a different layout to
perform better, layout auditing learns the new best layout. We
implemented our technique in a product Java virtual machine, using
copying generational garbage collection to produce different layouts,
and tested it on 20 long-running benchmarks and 4 hardware platforms.
Given any combination of benchmark and platform, layout auditing
consistently performs close to the best layout for that
combination,without requiring offline training.

Martin Hirzel works at IBM Research in Hawthorne, NY, on improving the
performance and usability of high-level programming languages. He also
teaches classes at New York University. Martin received his Ph.D. from
the University of Colorado at Boulder in 2004. In his spare time, he
invents cheesy titles for papers and talks.

SPEAKER: Priya Nagpurkar (Part 2)
IBM TJ Watson Research Center

Title: Characterizing Emerging Web 2.0 Workloads


Understanding the demands of emerging workloads on the systems that host
them is important in designing future systems. Web 2.0 is one such
emerging class of workloads. Web 2.0 represents the evolution of the web
from a source of information to a platform. The technologies behind Web
2.0 enable users to more easily participate, collaborate and share in
web-based communities. The server-side workloads generated by Web 2.0
applications are different from traditional web workloads and present
new challenges to underlying systems. This talk presents a detailed
characterization of several applications that exploit Web 2.0
technologies at different layers in the system stack.

Priya is a Research Staff Member at IBM’s TJ Watson Research Center in
Yorktown.She joined IBM in October 2007, after completing her M.S and
Ph.D. in Computer Science at the University of California, Santa
Barbara. Her research interests include runtime analysis and
feedback-directed optimization to improve performance. Her doctoral
research focused on runtime techniques to improve the performance of
Java programs. At IBM, she has been working on characterizing emerging