Paul Lu Associate Professor Department of Computing Science

BIOGRAPHICAL DATA

• Ph.D. Computer Science, 2000, University of Toronto
  Title: Scoped Behaviour for Optimized Distributed Data Sharing

• M.Sc. Computing Science, 1993, University of Alberta;
  Title: Parallel Search of Narrow Game Trees
• B.Sc. Honours Computing Science, 1991, University of Alberta;

PUBLICATIONS

• Selected publications, by date.
• Selected publications, by type.
• CiteSeer (not all hits are my papers)
• DBLP Bibliography Server

CURRENT GRADUATE STUDENTS

1. Cam Macdonell (Ph.D., File Systems and Virtual Machines)
2. Yang Wang (Ph.D., Scheduling and File Systems in Trellis)
3. Jordan Patterson (M.Sc., TBA)
4. Erkan Unal (M.Sc., TBA)

TEACHING

January to April 2009:

• [ NEW COURSE ]: Science 100: Integrated Science
• CMPUT 201: Practical Programming Methodology
• [ NEW COURSE ]: CMPUT 497: Cluster Challenge and Computational Science
• MINT 706: Internet Applications and Programming
• MINT 714: High-Performance Server Programming

FORMER GRADUATE STUDENTS

1. Calvin Chan (M.Sc., Defended December 2004)
   • Thesis: Language Extensions for Multiple Code Inheritance in Java (740 KB, Adobe Acrobat PDF)
   • Co-supervised with Duane Szafron
   • Current location: Random Knowledge, Inc., Edmonton, Alberta
2. Mike Closson (M.Sc., Defended September 2004)
• Thesis: The Trellis Network File System (500 KB, Adobe Acrobat PDF)
• Current location: Developer, Platform Computing, Inc., Markham, Ontario, Canada.
3. Maria Cutumisu (M.Sc., Defended December 2002)
   • Thesis: Multiple Code Inheritance in Java (1.1 MB, Adobe Acrobat PDF)
   • Co-supervised with Duane Szafron
   • Current location: Ph.D. Student, Dept. of Computing Science, University of Alberta, Edmonton, Alberta
4. Meng Ding (M.Sc., Defended May 2005)
   • Thesis: Trellis-SDP: File-Level Data Parallelism (880 KB, Adobe Acrobat PDF)
   • Current location: Developer, Platform Computing, Inc., Markham, Ontario, Canada.
5. Adrian Driga (M.Sc., Defended December 2001)
   • Thesis: Parallel FastLSA: A Parallel Algorithm for Pairwise Sequence Alignment (880 KB, Adobe Acrobat PDF)
   • Current location: Bioinformatician, Polyomx Project, Cross Cancer Institute, Edmonton, Alberta
6. Chris Dutchyn (M.Sc., Defended March 2002)
   • Thesis: Multi-Dispatch in the Java Virtual Machine: Design, Implementation, and Evaluation (1.6 MB, Adobe Acrobat PDF)
   • Co-supervised with Duane Szafron
   • Current location: Assistant Professor, Dept. of Computer Science , University of Saskatchewan, Saskatoon, SK
7. Roman Eisner (M.Sc., Defended September 2005)
   • Thesis: Predicting Protein Function Using Machine-Learned Hierarchical Classifiers (not yet on-line; x.x MB, Adobe Acrobat PDF)
   • Co-supervised with Duane Szafron
   • Current location: The Human Metabolome Project, Dept. of Computing Science, University of Alberta, Edmonton, Alberta
8. Mark Goldenberg (M.Sc., Defended April 2003)
   • Thesis: TrellisDAG: A System for Structured DAG Scheduling (1 MB, Adobe Acrobat PDF)
   • Co-supervised with Jonathan Schaeffer
   • Current location: Living in Israel.
9. Nicholas Lamb (M.Sc., Defended August 2005)
• Thesis: Data-Conscious Scheduling of Workflows in Metacomputers (590 KB, Adobe Acrobat PDF)
• Current location: Agilent Technologies, Vancouver, BC.
10. Paul Nalos (M.Sc., Defended August 2006)
• Thesis: Fault Tolerance through Redundancy for Batch Workloads (1.5 MB, Adobe Acrobat PDF)
• Current location: Developer, Safe Software , Surrey, BC.
11. Cam Macdonell (M.Sc., Defended September 2002)
• Thesis: Trigger Scripts for Extensible File Systems (390 KB, Adobe Acrobat PDF)
• Current location: Ph.D. student, Dept. of Computing Science, University of Alberta, Edmonton, Alberta
12. Ernie Novillo (M.Sc., Defended October 2001)
    • Thesis: On-Line Performance Monitoring of Shared Memory Parallel Programs Using Barriers (990 KB, Adobe Acrobat PDF)
    • Current location: Developer, Digital Extremes, Inc., London, Ontario.
13. Chris Pinchak (M.Sc., Defended December 2002)
    • Thesis: Placeholder Scheduling for Overlay Metacomputing (1.1 MB, Adobe Acrobat PDF)
    • Current location: Ph.D. Student, Dept. of Computing Science, University of Alberta, Edmonton, Alberta
14. Limin Zhang (M.Sc., Defended September 2003)
    • Thesis: A User-Level Library for Parallel Disk Input-Output (780 KB, Adobe Acrobat PDF)
    • Current location (as of 2004): Software Developer, SAFLINK Corporation

SOFTWARE

1. Stupid Barrier Tricks (SBT) : A user-level library to conveniently interface with low-level hardware performance counters. Supports easy on-line monitoring and debugging of single-program multiple-data (SPMD) applications with barriers. (Category: Parallel performance tool.)
2. PBSWeb : A Web-based interface to the Portable Batch System (PBS). Supports submitting and monitoring batch jobs. Will support the automatic control and submission of jobs to multiple different queues at different high-performance resource centers. Supported by a grant from C3. (Category: Metacomputing/grid computing system)

TEACHING

Previous (1999 to 2008):

• Graduate Course: CMPUT 605 Software Systems Research Seminar. Starts in First Term; for credit in Second Term. (January 2001-2005)
• Graduate Course: CMPUT 681 Parallel and Distributed Systems (January 2000, September 2000, January 2002, September 2002, January 2005, September 2007)
• Graduate Course: MINT 706: Internet Applications and Programming (2005, 2006, 2007, 2008)
• Graduate Course: MINT 714: High-Performance Server Programming (2008)
• CMPUT 201: Practical Programming Methodology (September 2005, January 2007, September 2007, January 2009)
• CMPUT 229: Computer Organization and Architecture I. (September 2000, September 2001)
• CMPUT 285: Computer Organization and Architecture II. (January 1999) (Now revised and known as CMPUT 229 ).
• CMPUT 379: Operating System Concepts. (January 1999, September 1999, January 2002, September 2002, September 2003, January 2004, January 2005)
• CMPUT 399: Computational Science and Clusters (January 2008)
• CMPUT 497: Cluster Challenge and Computational Science (January 2009)
• CMPUT 498: Concurrency, Performance, and Architectures in Software Systems (January 2004)

RESEARCH INTERESTS

High-Performance Computing and Metacomputing

I am interested in all aspects of high-performance computing, but especially parallel and distributed systems. My research program centers on systems software and, most recently, software infrastructure for wide-area overlay metacomputing. I am the PI of the Trellis Project.

The generous financial support of the the Natural Sciences and Engineering Research Council of Canada (NSERC), Alberta Science and Research Authority, the University of Alberta, the Canada Foundation for Innovation (CFI), SGI, Sun Microsystems, and C3.ca is greatly appreciated.

Currently, I am a member of the Software Systems Research Group.

Bioinformatics: Proteome Analyst and PA-GOSUB

The Protoeme Analyst (PA) project is developing a web-based system for high-throughput protein annotation. Using machine-learned classifiers, PA can predict several properties of proteins (from their primary sequence information), including GO molecular function and subcellular localization. In fact, PA is the current, most-accurate predictor of subcellular localization.

Proteome Analyst has been generously supported by the Canadian Protein Engineering Network of Centres of Excellence (PENCE) and Sun Microsystems.

See the Bioinformatics Research Group web page.

Other

In 1996, I co-edited (with Greg Wilson) a Parallel Programming Using C++, a book describing many of the key C++-based parallel programming systems. That book is available from MIT Press. The foreword is by Bjarne Stroustrup. The table of contents is available here Gzip'ed Postscript (67 kbytes). The slides from a survey talk at the POOMA'96 conference is available here Gzip'ed Postscript (20 kbytes).

In the past, I developed Aurora, a distributed shared data (DSD) system based on a standard C++ class library and run-time system. The system exploits language mechanisms for creating abstract data types and does not contain any language extensions.

Also, I've worked on the parallel search of game trees as part of the Chinook checker playing program project. Chinook is the World Man-Machine Champion, the first computer program to win a human world championship. This feat is recognized by the Guinness Book of World Records.