CS 181E Resource Site: Fundamentals of Parallel Programming (Fall 2012)
...
This web site contains resources for the Fall 2012 offering of CS 181E at Harvey Mudd College. For general information on this course, please go to see the course home Twiki page and the course syllabus.
Lecture Schedule
| Day | Date (2012) | Topic | Slides | Audio (Panopto) | Code Examples | Assignment | |||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Module 1: Deterministic Shared-Memory Parallelism | ||||||||||||||||||
1 | Wed | Sep 05 | Lecture 1: Introduction, Async-Finish Parallel Programming, Computation Graphs, Abstract Performance Metrics, Amdahl's Law |
|
|
| Parallel Array Sum | lec1-audio | HW0 HW1 (due by 11:59pm on Tuesday, Sep 11th) | |||||||||
2 | Mon | Sep 10 | Lecture 2: Data Races and Determinism, Finish Accumulators, Futures (Tasks with Return Values), Dataflow Programming, Data-Driven Tasks |
|
| Parallel Array Sum (contd), Amdahl's Law, Weak vs. Strong Scaling, Data Races and Determinism | lec2-audio |
| ||||||||||
3 | Wed | Sep 12 | Lecture 3: Abstract vs. Real Performance, Seq clause, Forasync Loops, Loop Chunking, Forall Loops and Barrier Synchronization |
|
|
| Finish Accumulators, Futures (Tasks with Return Values), Dataflow Programming, Data-Driven Tasks |
| HW1 HW2 (due by 11:59pm on Tuesday, Sep 18th) | |||||||||
4 | Mon | Sep 17 | Lecture 4: Parallel Prefix Sum algorithm, Parallel QuickSort, Point-to-point Synchronization, Phasers |
| Programming Patterns, Seq clause, Forall Loops, Barrier Synchronization |
|
| |||||||||||
5 | Wed | Sep 19 | Lecture 5: Parallel MergeSortSystolic Algorithms, Parallel BitonicSorOdd-Even Sort |
|
| HW3 HW2 (due by 11:59pm on Tuesday, Sep 25th) | ||||||||||||
6 | Mon | Sep 24 | Lecture 6: Collective and Point-to-point Synchronization with Phasers, Phased Forasync Loops, Phaser Accumulators, | Bounded Phasers, Memory Consistency Models, Summary of Deterministic Shared-Memory ParallelismLoop Chunking | lec6-audio |
|
| |||||||||||
Module 2: Nondeterministic Shared-Memory Parallelism | ||||||||||||||||||
7 | Wed | Sep 26 | Lecture 7: Actors |
| Critical sections and the Isolated statement, Atomic Variables | HW4 HW3 (due by 11:59pm on Thursday, Oct 2nd4th) | ||||||||||||
8 | Mon | Oct 01 | Lecture 8: Systolic arrays, Systolic algorithms (*Module 3 topic) Observationally Cooperative Scheduling (Guest lecturer: Prof. Melissa O'Neil) | lec8-slides |
|
|
| |||||||||||
9 | Wed | Oct 03 | Lecture 9: Critical sections and the Isolated statement, Monitors, Atomic Variables, Linearizability of Concurrent Objects |
|
| Monitors, Actors | lec9-slides | HW4 HW5 (due by 11:59pm on Friday, Oct 9th12th) | ||||||||||
10 | Mon | Oct 08 | Lecture 10: Linearizability of Concurrent Objects (contd), Safety and Liveness Properties | , Progress Guarantees | lec10-slides |
|
| |||||||||||
Module 3: Distributed-Memory Parallelism | ||||||||||||||||||
11 | Wed | Oct 10 | Lecture 11: Task Affinity with Places, Message Passing Interface (MPI), demonstration of LittleFe cluster system | lec11-slides |
| HW6 HW5 (due by 11:59pm on Wednesday, Oct 16h17th) | ||||||||||||
12 | Mon | Oct 15 | Lecture 12: Cloud Computing, MapReduce, GPU Programming |
|
|
| Module 4: Real-world Parallel Programming Models and Challenges | Message Passing Interface (contd) | lec12-slides |
|
| |||||||
13 | Wed | Oct 17 | Lecture 13: Real-World Parallel Programming Models, Course Review | lec13-slides |
|
| Take-home Final Exam (3-hour duration, due by 5pm on Oct 19th) |