CS735/CS835
Introduction to Parallel and Distributed Programming
Course Outline

1. Introduction [Andrews, Chapter 1]
   • hardware architectures
   • applications and programming styles

2. Programming with Java threads [Andrews, Section 5.4; The Java Language Specification]
   • Thread class
   • monitors
   • numerical integration using Java threads: code and performance

3. Shared-variable programming fundamentals [Andrews, Chapters 2-5]
   • atomic actions and locks
   • barrier synchronization
   • semaphores
   • monitors
   • case study: POSIX threads

4. Programming with Java RMI [Andrews, Section 8.5; The Java Tutorial: RMI; Java RMI documentation]
   • java.rmi package
   • remote method invocation
   • registry service
   • numerical integration using Java RMI: code and performance

5. Message-passing programming [Andrews, Chapter 7]
   • asynchronous message passing
   • synchronous message passing
   • collective communication: hypercube-based algorithms
   • case study: MPI

6. Distributed-programming paradigms [Andrews, Chapter 9]
   • manager/workers
   • heartbeat
   • client/server
   • pipelines

7. Scientific computing applications [Andrews, Chapter 11]
   • grid
   • particle
   • matrix

8. Combinatorial search [Quinn, Chapter 13]
   • parallel branch-and-bound
   • load balancing

9. Combining threads and message passing [PM2 webpage]
   • Java: threads and RMI
   • PM2: threads and RPC

10. Message-passing implementation issues [Active Messages paper; SBP paper; LFC paper;]
    • active messages
    • eliminating O.S. overhead
    • integrating message passing and thread scheduling

11. Data-parallel programming [Andrews, Section 3.5]
    • SIMD hardware
    • parallel prefix
    • data-parallel programming on MIMD hardware

12. Communication via tuple spaces [Andrews, Section 7.7; JavaSpaces spec]]
    • Linda
    • JavaSpaces

13. Distributed shared memory [Andrews, Section 10.4; Pugh's webpage; Hyperion paper]
    • consistency protocols
    • Java memory model
    • UNH Hyperion project

14. Is Java a high-performance programming language? [October 2001 CACM]

Comments and questions should be directed to pjh@cs.unh.edu