Postgraduate Course: Parallel Programming Languages and Systems (Level 11) (INFR11023)
|School||School of Informatics
||College||College of Science and Engineering
||Availability||Not available to visiting students
|Credit level (Normal year taken)||SCQF Level 11 (Postgraduate)
|Home subject area||Informatics
||Other subject area||None
||Taught in Gaelic?||No
|Course description||The module familiarises students with the issues involved in designing, implementing and applying parallel programming systems. Initial motivation will be provided by consideration of a number of typical high performance applications and parallel architectures. This will highlight the role of parallel software systems as a means of bridging the gap between these and allow abstraction of the issues which must be addressed by any such system (partitioning, communication, agglomeration, scheduling). It will explore the ways in which these challenges have been addressed by a range of systems, including both de facto standards and more adventurous research projects. This course is identical to the level 10 version except for the assessed coursework and additional learning outcome.
Entry Requirements (not applicable to Visiting Students)
|Prohibited Combinations|| Students MUST NOT also be taking
Parallel Programming Languages and Systems (Level 10) (INFR10015)
||Other requirements|| For Informatics PG and final year MInf students only, or by special permission of the School. Students must also be competent programmers in C, C++ or Java and have an appreciation of the connection between log_2 and 2^.
|Additional Costs|| None
Course Delivery Information
|Delivery period: 2011/12 Semester 2, Not available to visiting students (SS1)
||WebCT enabled: No
|No Classes have been defined for this Course|
||Week 1, Tuesday, 09:00 - 09:50, Zone: Central. Bristo Sq 7 LT3 |
|Main Exam Diet S2 (April/May)||2:00|
Summary of Intended Learning Outcomes
|1 - Describe generic issues (as discussed in the syllabus) which must be addressed by any parallel programming system.
2 - Explain, given a description of a previously unseen parallel application, where specific instances of the generic issues will arise.
3 - Explain, in considerable detail, the ways in which the generic issues are addressed by the MPI and Pthreads programming models and their supporting infrastructure.
4 - Apply their practical experience with MPI and Pthreads to write clean, adaptable and scalable parallel programs for simple applications.
5 - Compare the approaches proposed by a range of more speculative programming models.
6 - Review and critically evaluate literature describing new parallel programming models.
|Written Examination 80|
Assessed Assignments 20
Oral Presentations 0
Assessment will be by a combination of
* Short programming exercises, designed to familiarize students with the practicalities of parallel programming.
* pencil-and-paper exercises
* a literature review, investigating a previously unseen parallel programming model and requiring critical analysis of the ways in which this addresses the generic issues identified during the course.
||* Introduction. Conceptual and architectural models of parallelism. Parallel algorithmic paradigms and simple parallel algorithms.Iterative, recursive pipeline and bag-of-task parallelism.
* Generic issues in shared variable programming. Mutual exclusion, condition synchronization. Locks, barriers semaphores and monitors, use and implementation.
* Realisation in Pthreads and Java.
* Issues in message passing programming. Channels and synchronization models.
* Realisation in MPI: communicators, simple and collective operations
* Selected case studies highlighting alternative approaches, for example,
o Associative models. Linda, tuple space and related primitives.
o Shared Virtual Memory with Treadmarks.
Relevant QAA Computing Curriculum Sections: Concurrency and Parallelism
||* G.R. Andrews, Foundations of Multithreaded, Parallel and Distributed Programming, 2000.
* B. Wilkinson, M.Allen, 'Parallel Programming, Techniques and Applications', 1999.
* I Foster, 'Designing and Building Parallel Programs', 1995.
* M.J.Quinn, 'Parallel Programming in C with MPI and OpenMP', 2004.
* Selected research papers.
Timetabled Laboratories 0
Non-timetabled assessed assignments 20
Private Study/Other 60
|Course organiser||Dr Michael Rovatsos
Tel: (0131 6)51 3263
|Course secretary||Miss Kate Weston
Tel: (0131 6)50 2701
© Copyright 2011 The University of Edinburgh - 16 January 2012 6:17 am