Andrea Arpaci-Dusseau's Papers
Research by Andrea Arpaci-Dusseau
My papers are roughly divided into four main areas:
- Scheduling and Resource Allocation in Clusters
- General Issues in Clusters and NOW
- Parallel Sorting (Modeling with LogP and Disk-to-disk NOW-Sort)
- Split-C Language
Scheduling and Resource Allocation in Clusters
See the Implicit
Coscheduling page for related information.
- Implicit Coscheduling: Coordinated Scheduling with Implicit
Information in Distributed Systems
Andrea C. Arpaci-Dusseau
Ph.D. Dissertation, University of California, Berkeley, December 1998.
Also available as a U.C. Berkeley Technical Report (UCB//CSD-99-1052)
Abstract, Compressed Postscript,
and Postscript
- Building MPI for Multi-Programming Systems using Implicit Information
Frederick C. Wong,
Andrea C. Arpaci-Dusseau,
David E. Culler
In The
6th European PVM/MPI User's Group Meeting
Available as: Abstract,
PDF,
Compressed Postscript,
and Postscript
- A Cooperative Approach to Two-Phase Waiting
Andrea C. Arpaci-Dusseau,
David E. Culler
A open problem for the more theoretically inclined...
Available as: Abstract,
Compressed Postscript,
and Postscript
- Scheduling with Implicit Information in Distributed Systems
Andrea C. Arpaci-Dusseau,
David E. Culler,
Alan Mainwaring.
In
Sigmetrics'98 Conference on the Measurement and Modeling of Computer
Systems,
Available as:
Abstract,
Compressed
Postscript,
Postscript
and Slides
- Extending Proportional-Share Scheduling to a Network of
Workstations
Andrea C. Arpaci-Dusseau,
David E. Culler.
In International Conference on Parallel and Distributed Processing
Techniques and Applications (PDPTA'97)
, Las Vegas, Nevada, June 1997.
Available as:
Abstract
and PostScript
- Effective Distributed Scheduling of Parallel Workloads
Andrea C. Dusseau,
Remzi H. Arpaci,
David E. Culler.
In Sigmetrics'96
Conference on the Measurement and Modeling of Computer Systems,
Philadelphia, PA, May 1996.
Available as: Abstract, Compressed Postscript, Postscript, Slides, and
Simulation Data
- Re-examining Scheduling and Communication in Parallel Programs
Andrea C. Dusseau, Remzi H. Arpaci, David E. Culler.
University of California, Berkeley, Computer Science Technical Report:
UCB//CSD-95-881
Available as: Postscript
- The Interaction of Parallel and Sequential Workloads on a Network of
Workstations
Remzi H. Arpaci,
Andrea C. Dusseau,
Amin M. Vahdat,
Lok T. Liu,
Thomas E. Anderson,
David
A. Patterson.
In SIGMETRICS '95 , 1995.
Available as:
Abstract, PostScript,
and Slides
- Towards Global Process Management in a Network of Workstations
Remzi H. Arpaci,
Andrea C. Dusseau,
Amin M. Vahdat.
Study for Parallel Computer Architecture (CS258), Spring 1995
Available as: HTML
and Slides
Overview of Clusters and NOW
See the NOW web page for more
information and papers about the U.C. Berkeley NOW project.
- Parallel Computing on the Berkeley NOW
David Culler, Andrea Arpaci-Dusseau, Remzi Arpaci-Dusseau, Brent Chun,
Steve Lumetta, Alan Mainwaring, Richard Martin, Chad Yoshikawa, Frederick
Wong
In JSPP'97 9th Joint Symposium on Parallel Processing, Kobe,
Japan, 1997
Available as: Abstract
and Postscript
Parallel Sorting
NOW-Sort describes our
work on disk-to-disk sorting in a Network of Workstations.
- High-Performance Sorting on Networks of Workstations.
Andrea C. Arpaci-Dusseau,
Remzi H. Arpaci-Dusseau,
David E. Culler,
Joseph M. Hellerstein,
David A. Patterson.
In
ACM SIGMOD Conference on the Management of Data, Tucson, Arizona, May, 1997.
Available as:
Abstract and
PostScript
- Searching for the Sorting Record: Experience with NOW-Sort
Remzi Arpaci-Dusseau,
Andrea Arpaci-Dusseau,
David E. Culler,
Joseph M. Hellerstein,
David A. Patterson.
In 2nd SIGMETRICS Symposium on Parallel and Distributed Tools,
August 1998.
Available as:
Abstract and Postscript
- The Architectural Costs of Streaming I/O: A Comparison of
Workstations, Clusters, and SMPs.
Remzi Arpaci-Dusseau,
Andrea Arpaci-Dusseau,
David E. Culler,
Joseph M. Hellerstein,
David A. Patterson.
HPCA 4, Las Vegas, February, 1998.
Available as:
Abstract
and PostScript
In the following papers, LogP, a model for communication performance, is used to predict the
execution time on the CM-5 of four parallel sorting algorithms written in
Split-C.
- Modeling Parallel Sorts with LogP on the CM-5
Andrea C. Dusseau
Masters Report, May 1994 (Technical Report: UCB/CSD-94-829)
Available as: Abstract,
Postscript,
Slides,
and Split-C code
Note: You probably don't want to print out the whole document; after
page 32 the Postscript contains just Split-C source code.
- Fast Parallel Sorting Under LogP: Experience with
the CM-5
Andrea C. Dusseau, David E. Culler, Klaus Erik Schauser, and Richard P. Martin
In IEEE Transactions on Parallel and Distributed Systems,
August 1996. Volume 7, Number 8, pp. 791-805.
- Fast Parallel Sorting under LogP: from Theory to Practice
David E. Culler, Andrea C. Dusseau, Richard P. Martin, and Klaus Erik
Schauser
In Portability and Performance for Parallel Processing, Chapter 4,
pp.71-98, 1993, John Wiley & Sons.
Available as: Postscript
Note: This paper is a subset of the above Journal version.
Split-C Language
Split-C is a parallel extension of the C programming language that supports
efficient access to a global address space on current distributed memory
multiprocessors.
- Parallel Programming in Split-C
David E. Culler, Andrea C. Dusseau, Seth Copen Goldstein,
Arvind Krishnamurthy, Steven Lumetta, Thorsten von Eicken,
and Katherine Yelick
In Supercomputing93, pp. 262-273, Nov. 1993
Available as: Abstract and Postscript
- Introduction to Split-C
David E. Culler, Andrea Dusseau, Seth Copen Goldstein,
Arvind Krishnamurthy, Steven Lumetta, Steve Luna, Thorsten von Eicken,
and Katherine Yelick
March 6, 1994
Availabe as: Abstract and
Postscript
Back to Andrea's Home Page
Last modified June 10, 1999 by dusseau@cs.Berkeley.EDU