Next: Using the Decision Tree
Up: Introduction
Previous: Why Eigenvalue Templates?
  Contents
  Index
Intended Readership
We hope our templates will address the needs of
three potential user communities for eigenvalue computation tools:
- Students and teachers who want simple but generally
effective algorithms and would like access to the latest
developments in a form that is easy to explain and to understand.
- General engineers and scientists who desire easy-to-use,
reliable software that is also reasonably efficient.
- The high-performance computing
community consisting of those engineers and scientists trying to solve the largest,
hardest applications in their fields. Such users
desire high-speed access to algorithmic details for performance tuning and
to reliability for their problem.
It may seem difficult to address the needs of such diverse
communities with a single document. Nevertheless, we believe
that templates allow us to organize the information in ways
useful to all. For example, we expect students to use the high-level
algorithm descriptions to acquire a basic understanding of
how the methods work. General engineers
and scientists should find the decision tree and pointers to
software adequate for most of their problems. Finally, the high-performance
computing community can use the advice on performance
tuning and assessing the accuracy of the output to help solve their
hardest problems.
Next: Using the Decision Tree
Up: Introduction
Previous: Why Eigenvalue Templates?
  Contents
  Index
Susan Blackford
2000-11-20