In this course we will study various parallel computing architectures (shared memory vs. distributed memory), programming mechanisms for using these architectures, topologies of parallel computing platforms (mesh, hypercube, etc.), and parallel algorithms.

Online resources.

CPSC 250, or permission of the instructor. Familiarity with Unix is assumed.

At the end of this course successful students will be able to:

1. design, implement, and test parallel algorithms, using the C++ programming language, using shared memory and distributed memory architectures,
2. analyze the efficiency of parallel algorithms and their implementations,
3. informally prove the correctness and efficiency of various parallel algorithms on a variety of architectures, and
4. determine appropriate architectures and algorithms for use in solving problems.

Regular attendance in class is highly recommended. Regardless of attendance, students are responsible for all material covered or assigned in class.

The course will meet in class for 3 hours during the week. The concepts studied in class will be complemented by several programming assignments. There will be two tests in class (on October 1, and November 12) during the semester. A final project will serve as the final exam.

Make-up tests will be available by pre-arrangement only in case of scheduling conflicts. After the test, make-ups will be available only in case of documented medical emergencies.

Besides the exams, there will be quizzes in class, regular homework assignments, programming projects, and a co-curricular requirement.

Quizzes: Quizzes will be in class and will be announced one class period before the quiz.

Programming projects: There will be several programming projects assigned during the semester. Programs will be graded on correctness, style and documentation. Programs are due by midnight on the assigned date. All programs are to be turned in electronically; instructions for submission will be given in the assignment handout.

No late work (programs, homeworks, quizzes, etc.) will be accepted.

Co-curricular Requirement: The Mathematics, Computer Science and Physics department offers a series of discussions that appeal to a broad range of interests related to these fields of study. These co-curricular sessions will engage the community to think about ongoing research, novel applications and other issues that face these disciplines. Each student is required to attend at least three of these sessions, and turn in a short paper describing the contents of the session, and his/her critical reflections about the topic and content. These papers are due in class within a week of the session. A paper submitted beyond a week from the event being discussed in the paper will not be accepted. The MCSP Conversation Series website has the schedule of talks in the series.

The final grade will be computed based on the grades in the quizzes, tests, the final exam, home works and programming projects according to the following weights:

The final course grade will be calculated as follows:

Students are expected to adhere to the Academic Integrity policies of Roanoke College. All work submitted for a grade is to be strictly the work of the student unless otherwise specified by the instructor. The policies as outlined in the Academic Integrity handbook will be enforced in the course.

Graded programs are subject to the Roanoke College Academic Integrity policies. Copying a program or a portion of a program (even a single line) or reading another person's program to obtain ideas for solving a problem is plagiarism. Other examples of integrity violation include writing code for some else, using code written by someone else, telling someone else how to solve a problem or having someone tell you how to solve a problem (and using their method). These cases apply to any work that is handed in for a grade under the instructor's assumption that the work is your own. Unless specified otherwise by the instructor, discussion among students should be limited to general discussion of concepts and language details, not specific aspects of a solution to the assigned problem.