|
New England Database
Society sponsored by Sun Microsystems |
| |
NEDS |
Multithreaded architectures and database query processing
Hank Korth
Lehigh University
Friday, September 8,
2006, 4:00 PM
Volen 101,
Brandeis University
(preceded by a wine and cheese reception at 3:00 pm)
Abstract:
Multicore and multithreaded processors present both
opportunities and challenges in the design of database query-processing
algorithms. Modern database systems gain their parallelism from concurrent
transactions or queries, but parallelism of this sort can result in cache
contention. It is therefore increasingly important to consider parallelism
within individual database queries.
This talk describes experiments done with two of the operations that consume
most of the database CPU cycles: sorting and joining. We revisit the classic
sort benchmark in a main-memory/CPU-cache context and consider the hash-join
algorithm for a single join and a pair of pipelined joins. In each case we
examine performance for a pair of parallel threads and a pair of parallel
processors sharing a cache. Our results indicate issues that will rise to
greater importance with the emergence of new processors with an increased number
of cores and increased number of potential parallel threads.
This work is joint with Philip Garcia, formerly of Lehigh and now a Ph.D.
student at the University of Wisconsin, Madison.
Speaker Bio:
Henry F. Korth is Weiseman Professor and Chair of the Department of Computer
Science and Engineering at Lehigh University. He is director of the Systems,
Experimentation, and Analysis Laboratory for Databases (SEAL DB). His
publications include three books, one of which, Database Systems Concepts, is
now in its fifth edition; over 100 journal articles, conference publications and
other technical papers; and eight book chapters. Korth also holds eight patents.
Before his arrival at Lehigh, Korth held positions of leadership with Lucent
Technology's Bell Laboratories in Murray Hill, N.J. As Director of Database
Principles Research, he led teams of Ph.D. researchers in investigating XML data
management, web-based data, main-memory database systems, real-time systems,
parallel systems and other topics. At Lucent, Korth was a founding member of
the Joint Development and Delivery Center, which managed the rapid transition of
Bell Labs prototypes into marketplace products. He was a member of two teams
that received the Bell Labs President's Silver Award, in 1998 for the QTMT
aggregation engine and in 1999 for the DataBlitzT main-memory storage manager.
Korth also led the creation of two other Lucent products: the InterprenetT IP
Usage Suite, and the WebLightningT Web Accelerator.
Before joining Lucent, Korth served as vice president of Panasonic Technologies
and director of the Matsushita Information Technology Laboratory, the leading
computer systems research laboratory for Matsushita Electric Industrial Co. In
this position, Korth conducted research in multidatabase systems, mobile
computing, database user interfaces and multimedia information systems. Korth
is a Fellow of the Association for Computing Machinery, (ACM) and a Fellow of
the Institute of Electrical and Electronics Engineers (IEEE,). At the 1995
International Conference on Very Large Data Bases, a paper he co-authored titled
"A Model of CAD Transactions" was chosen as "Most Influential Paper from the
Proceedings of Ten Years Ago." Korth has served on the editorial boards of
several computing journals, and currently serves on the board of ACM
Transactions on Database Systems. From 1983 to 1992, Korth served on the
faculty of the Department of Computer Sciences at the University of Texas at
Austin, where he attained the rank of associate professor with tenure. Prior to
that, he was a research staff member at the IBM T. J. Watson Research Center.
Hank holds Ph.D., M.A., and M.S.E degrees from Princeton University, Princeton,
New Jersey, in computer science, and an A.B degree in mathematics from Williams
College, Williamstown, Massachusetts.
Maintained by Dina Goldin dqg AT cs.brown.edu