CSC 8400 Computer Organization - Syllabus

CSC 8400	Computer Organization	Spring 2007
	SYLLABUS

Meetings Section 1: Mondays 6:15pm-9:00pm, Mendel Science Center G88

Instructor Dr. Thomas Way
160A Mendel Science Center Email: thomas.way@villanova.edu
IM: DrTomWay
Phone: (610) 519-5033

Office hours (See my web site)

Teaching Assistant See instructor for help. General help is available from Programming Assistants.
Office hours: see Programming Assistants' schedule on CS Dept. web site

Textbook Computer Architecture: A Quantitative Approach, John L. Hennessy & David A. Patterson, Morgan Kaufmann Publishers, Fourth Edition, 2006.

Prerequisite CSC 7100 Computer Systems (or the equivalent)

Web site http://www.csc.villanova.edu/~tway and follow the link for CSC 8400

Course
description Evolution of computer architectures; computer performance; instruction set types and addressing modes; advanced pipelining; memory organization; caches, main memory and virtual memory; storage technologies; input/output systems; hardware aspects of interconnection networks; parallel computer architectures. This course is on computer architecture, with the goal being to teach you the basic and advanced concepts and methods used in the organization and design of computer hardware.

Lesson plan

The course closely follows the textbook outline. Some additional material (videos, research papers, etc.) will be included from time to time to supplement what we cover from the text. Some C programming and Unix will be introduced as part of 3 hands-on workshop projects. Advanced computer architecture topics will be studied throughout the semester.

Some students have a strong background in this subject from their undergraduate studies or career experience, while others do not. Every effort will be made to present a balance of advanced material while assuring a thorough understanding of the fundamental concepts of computer architecture.

The advanced nature of this material lends itself to a six-pronged approach to learning about computer architecture. Lectures provide a foundational understanding of the subject matter and an appreciation of the interdependent nature of the subject. Homework assignments involve in-depth analysis, application of concepts and problem solving. Workshop projects involve experimental research and C programming on a Unix system to explore the material. A literature review paper provides experience in graduate level research synthesis and conference paper style writing. A fact-based final exam focuses on developing a broad knowledge and understanding of factual material. One or more videos will be used to provide a broad historical perspective to our topic, learning from the past to guide the future.

Expect to work hard this semester.

Tentative schedule Introduction: Fundamentals of Computer Design (Chapter 1, 1 week)
Instruction Set Principles & Examples (Appendix B, 1 week)
Pipelining and Instruction-Level Parallelism (Appendix A & Chapter 2, 3 weeks)
Limits on Instruction-Level Parallelism (Chapter 3, 1 week)
Multiprocessors & Thread-Level Parallelism (Chapter 4, 2 weeks)
Review of Memory Hierarchy (Appendix C, 1 week)
Memory Hierarchy Design (Chapter 5, 2 weeks)
Storage Systems (Chapter 6, 2 weeks)
Lab time (varies)
Historical context videos (1 week)

Grading policy 30% Homeworks
30% Workshop projects
25% Final exam
15% Literature Review Paper
+/- Participation (attendance, class discussion, intellectual contribution to class)

Final grades

A
A-

93
90

B+
B
B-

86
82
79

C+
C

75
70

Makeup
Policy No missed tests without prior excuse. Each case will be handled separately based on its own merits. Makeup tests might be more difficult than regularly scheduled tests. Each student is responsible for what is covered and assigned in any classes which they miss.

Late Assignment Policy No assignments will be accepted late without the direct consent of the instructor prior to the due date of the assignment. Typical penalty is 10% off for each day an assignment is late. Absolutely no assignments will be accepted beyond the date of the final exam.

Academic Integrity Although collaboration among students is welcome when discussing concepts, ideas and approaches, all graded assignments must represent the student's individual work, as set forth in the University's policy on Academic Integrity. This means that a student who attempts to submit the work of another student, or material copied and pasted from the Internet or other sources, as his or her own will at the minimum receive 0% credit for the assignment, and at the maximum a failing grade for the course, at the discretion of the instructor.

Last updated: 1/16/07