Course Description
Advanced and high-performance computer architecture. Topics covered include hierarchical memory design, cache memory design, pipelining, vector processing and parallel processing. Case studies of vector processors and multi-processor architectures: shared memory, distributed memory, data parallel (SIMD) and control parallel (MIMD) architectures, and heterogeneous parallel computing on a network of workstations.
Athena Title
ADV COMP ARCH
Prerequisite
CSCI 6720
Semester Course Offered
Not offered on a regular basis.
Grading System
A - F (Traditional)
Course Objectives
The student is expected to develop a clear understanding of the cost-performance tradeoffs involved in the design of high-performance and parallel computer architectures. The impact of architectural decisions on algorithm design, compiler design and operating systems design are emphasized.
Topical Outline
Review of Conventional Architecture Design and Analysis, Shortcomings of Conventional Architectures, Cache and Virtual Memory Design, Pipelining, Vector Processors, Superscalar Architectures, Multiprocessing Techniques: Shared Memory, Distributed Memory, Data Parallelism and Control Parallelism, Interconnection Networks, Heterogeneous Parallel Computing on a Cluster of Workstations, Reconfigurable Architectures. In addition to the material from the textbook, papers from the current research literature will be discussed. Students are also expected to work on a self-directed project either individually or in a small group. The students are expected to present their project towards the end of the semester and also submit a technical report/term paper.