CS 575 Supercomputing for the Sciences
Text: High Performance Computing 2nd Edition, Oreilly
Dowd and Severence

Dr. Kris Stewart (stewart@rohan.sdsu.edu)
San Diego State University

This URL is http://www.stewart.cs.sdsu.edu/cs575/HPC2text_toc.html

High Performance Computing, 2nd Edition

Table of Contents

Preface

PART 1: Modern Computer Architectures

Chapter 1. What Is High Performance Computing?

   Why Worry About Performance?

   Scope of High Performance Computing

   Studying High Performance Computing

   Measuring Performance

   The Next Step
Chapter 2. High Performance Microprocessors

   Why CISC?

   Fundamentals of RISC

   Second-Generation RISC Processors

   RISC Means Fast

   Out-of-Order Execution: The Post-RISC Architecture

   Future Trends: Intel IA-64 and EPIC

   Closing Notes
Chapter 3. Memory

   Memory Technology

   Registers

   Caches

   Cache Organization

   Virtual Memory

   Improving Memory Performance

   Closing Notes
Chapter 4. Floating-Point Numbers

   Reality

   Representation

   Effects of Floating-Point Representation

   Improving Accuracy Using Guard Digits

   History of IEEE Floating-Point Format

   IEEE Floating-Point Standard

   IEEE Storage Format

   IEEE Operations

   Special Values

   Exceptions and Traps

   Compiler Issues

   Closing Notes

PART 2: Programming and Tuning Software

Chapter 5. What a Compiler Does

   History of Compilers

   Which Language to Optimize?

   Optimizing Compiler Tour

   Optimization Levels

   Classical Optimizations

   Closing Notes
Chapter 6. Timing and Profiling

   Timing

   Subroutine Profiling

   Basic Block Profilers

   Virtual Memory

   Closing Notes
Chapter 7. Eliminating Clutter

   Subroutine Calls

   Branches

   Branches Within Loops

   Other Clutter

   Closing Notes
Chapter 8. Loop Optimizations

   Operation Counting

   Basic Loop Unrolling

   Qualifying Candidates for Loop Unrolling

   Nested Loops

   Loop Interchange

   Memory Access Patterns

   When Interchange Won't Work

   Blocking to Ease Memory Access Patterns

   Programs That Require More Memory Than You Have

   Closing Notes

PART 3: Shared-Memory Parallel Processors

Chapter 9. Understanding Parallelism

   Dependencies

   Loops

   Loop-Carried Dependencies

   Ambiguous References

   Closing Notes
Chapter 10. Shared-Memory Multiprocessors

   Symmetric Multiprocessing Hardware

   Multiprocessor Software Concepts

   Techniques for Multithreaded Programs

   A Real Example

   Closing Notes
Chapter 11. Programming Shared-Memory Multiprocessors

   Automatic Parallelization

   Assisting the Compiler

   Closing Notes

PART 4: Scalable Parallel Processing

Chapter 12. Large-Scale Parallel Computing

   Amdahl's Law

   Interconnect Technology

   A Taxonomy of Parallel Architectures

   A Survey of Parallel Architectures

   The Top 500 Report

   Shared Uniform Memory MIMD

   Shared Non-Uniform Memory MIMD Systems

   Distributed-Memory MIMD Architecture

   Single Instruction, Multiple Data

   Closing Notes
Chapter 13. Language Support for Performance

   Data-Parallel Problem: Heat Flow

   Explicitly Parallel Languages

   FORTRAN 90

   Problem Decomposition

   High Performance FORTRAN (HPF)

   Closing Notes
Chapter 14. Message-Passing Environments

   Parallel Virtual Machine (PVM)

   Message-Passing Interface (MPI)

   Closing Notes

PART 5: Benchmarking

Chapter 15. Using Published Benchmarks

   User Benchmarks

   Industry Benchmarks

   Closing Notes
Chapter 16. Running Your Own Benchmarks

   Choosing What to Benchmark

   Types of Benchmarks

   Preparing the Code

   Closing Notes

PART 6: Appendixes

Appendix A. Processor Architectures
Appendix B. Looking at Assembly Language
Appendix C. Future Trends: Intel IA-64
Appendix D. How FORTRAN Manages Threads at Runtime
Appendix E. Memory Performance
Index

Return to CS 575 Home Page [Fall2003]