EEC 581 Computer Architecture, Fall 2008
Catalog Data: EEC 581 Computer Architecture (4-0-4). Prerequisite: Graduate standing. The design of high-performance computer systems, with emphasis on cost-performance tradeoff, performance evaluation, instruction set design, hardwired control-unit design, micro- and nano-programming, pipelining, memory hierarchy, and I/O interfaces.
Textbook: Computer Architecture: A Quantitative Approach, by John L. Hennessy and David A. Patterson, 4th Edition, Morgan Kaufmann Publishers, Inc., 2006.
Coordinator: Dr. Chansu Yu, Associate Professor of Electrical and Computer Engineering
Room: SH 437, Tel: 2584, Email: firstname.lastname@example.org
TA: Song Cui, Tel: 216-313-3114, Email: email@example.com
Class: TR 6:00-7:50 PM, SH314
Office Hour: TR 2:00-4:00 PM or by appointment
Course objectives: To provide a comprehensive overview of the modern microprocessor organization with an emphasis on various types of parallelism – instruction level parallelism (ILP), thread-level parallelism such as simultaneous multithreading (SMT), and multiprocessors. Topics include including dynamic scheduling, hardware-based speculation, multithreading, shared-memory multiprocessor, distributed shared memory architecture, cache coherency, memory consistency models, memory hierarchy design and storage systems.
Course Requirements: Course requirements include six homework assignments, many pop quizzes and a term project. Students will form a group of three or four for the term project, which consists of four stages and must involve “green” computing. For the project, you need to review about at least fifteen papers to understand what other researchers have already done in the chosen area as well as to appreciate good research in green computing. Two (or three) classes before the midterm exam and another two (or three) before the final exams will be used for your project presentation. Each presentation will be based on power point slides and at most 15 minutes long excluding questions and answers. Each member of a team should participate in one of the two presentations.
Grading Policy: Midterm Exam 20%
Final Exam 30% (Comprehensive)
Quiz 10% (Many pop quizzes, review & preview)
Project 30% (3~4-person project, four sub-parts)
Homework 10% (5 homework assignments)