CS3220 Computer Architecture

Lecturers

John McCallum,
Office: S15 #03-13, Phone: 874-2990, Email: johnm

Yuen Chung Kwong,
Office: S16 #04-04, Phone 874-2831, Email: yuenck

Course Outline

• Execution control and microprogramming;
• I/O interfaces;
• Cache;
• Virtual memory;
• Instruction pipelining;
• Superscaler processors;
• Interconnections;
• Multiprocessor systems;
• Vector processors;
• Cache coherence;
• Parallel programming;
• Stack, functional programming and dataflow architectures
• Various Processor examples;

Textbooks and Course Materials

John L. Hennessy and David A. Patterson
Computer Architecture: A Quantitative Approach, 2nd Edition
Morgan Kaufmann, 1996 [likely to be 6th printing, 2001]
about $35 for the international student edition.

Course Assessment

Exam 70%; Assignments 30%

Overview

Computer architecture: Study of the relation between hardware and software: how hardware can support the needs of programs, or, how hardware structure can best match structure of algorithms, languages and operating systems. Much of subject is devoted to parallel systems, because they give rise to most problems of hardware-software relation, and because higher speed is a constant need of all software. Larger size is another need, but that has been basically solved with relatively simple arrangements (virtual memory; cache)

The 80s and 90s were exciting times for computer architecture because we had

1. Phenomenal increase in hardware density and CPU complexity & power
2. Practical techniques to harness parallel processing within single CPUs
3. Standard operating systems supporting widely use applications readily booted onto new CPUs
4. Networks and client server operating systems making processing power and software suites readily accessible

Why students should study computer architecture:

• learn the secrets that companies are spending big $$$ to develope
• learn what to look for when buying computers
• you understand the buzzwords bandied about with new computers
• you have pride of understanding what others do not
• in short, you will be a more knowledgeable IT professional

References:

Hennessy and Patterson
Computer Architecture: A Quantitative Approach, 2nd Edition
Morgan Kaufmann, 1996 

resources for H&P


IA-64 Application Developer's Architecture Guide [PDF 476 pages]