ECE 511 - Computer Architecture

Official Description

Advanced concepts in computer architecture: design, management, and modeling of memory hierarchies; stack-oriented processors; associative processors; pipelined computers; and multiple processor systems. Emphasis on hardware alternatives in detail and their relation to system performance and cost. Course Information: Same as CSE 521. Prerequisite: ECE 411 or CS 433.

Prerequisites

Credit in CS 433 or ECE 411

Subject Area

Hardware Systems

Course Directors

Description

Advanced concepts in computer architecture; design, management, and modeling of memory hierarchies, stack-oriented processors, associative processors, pipelined computers, and multiple processor systems; and focuses on hardware alternatives in detail and their relation to system performance/cost.

Notes

Same as CSE 521.

Topics

  • Processor design: memory hierarchy; bus, cache, and shared memory; disk arrays, pipelining and superscalar techniques; reducing branch penalties; multivector computers
  • Parallel array processors: SIMD computers; interconnection networks; systolic arrays
  • Multiprocessors: program and network properties; scalability issues; performance evaluation; caches and consistency issues; parallel programming and compilers; scheduling algorithms
  • Distributed-memory multicomputers: architecture; distributed shared memory; routing and network design issues; load balancing/data mapping algorithms; performance evaluation and visualization; design of parallel algorithms; scalable, multithreaded, and dataflow architectures; applications
  • Distributed processing: distributed operating system; load balancing algorithms

Detailed Description and Outline

Topics:

  • Processor design: memory hierarchy; bus, cache, and shared memory; disk arrays, pipelining and superscalar techniques; reducing branch penalties; multivector computers
  • Parallel array processors: SIMD computers; interconnection networks; systolic arrays
  • Multiprocessors: program and network properties; scalability issues; performance evaluation; caches and consistency issues; parallel programming and compilers; scheduling algorithms
  • Distributed-memory multicomputers: architecture; distributed shared memory; routing and network design issues; load balancing/data mapping algorithms; performance evaluation and visualization; design of parallel algorithms; scalable, multithreaded, and dataflow architectures; applications
  • Distributed processing: distributed operating system; load balancing algorithms

Same as CSE 521.

Texts

K. Hwang, Advanced Computer Architecture: Parallelism, Scalability, Programmability, McGraw-Hill, 1993.

Journal and conference articles from current literature.

Last updated

2/13/2013