Enabling Technologies for High-Performance Computing

10/3/97


Click here to start


Table of Contents

Enabling Technologies for High-Performance Computing

Outline

High-Performance Computing Today

Growth of Microprocessor Performance

Top500 Fastest Installed Computers

PPT Slide

Scalable Multiprocessors

Scientific Computing: 1987 vs 1997

Performance Numbers on RISC Processors

PPT Slide

PPT Slide

PPT Slide

The Maturation of Highly Parallel Technology

Architecture Alternatives

Directions

TeraFlops 1st Achieved

Achieving TeraFlops

3 Major Efforts in the US

Petaflops (1015) Computing Initiative

A Petaflops Computer System

Key Algorithm Research Issues

The Importance of Standards - Software

PVM - Parallel Virtual Machine

Physical and Logical Views of PVM

Programming in PVM

Motivation for a New Design

MPI Standard

New Concepts introduced in MPI

Comparing PVM and MPI

Challenges in Developing Distributed Memory Libraries

Memory Hierarchy

BLAS for Performance

Portability of Numerical Software

Directions - Metacomputing

Problem Solving Environments

PSE’s

Java

Java for Numerical Computations?

Metacomputing Objectives

Motivation for Network Enabled Solvers

Network Computing Classifications

PPT Slide

PPT Slide

PPT Slide

PPT Slide

PPT Slide

Next Step: Hardware & Software Servers

PPT Slide

PPT Slide

NetSolve Available Today

Conclusion

References

Author: Jack Dongarra

Email: dongarra@cs.utk.edu

Home Page: http://www.netlib.org/utk/people/JackDongarra/