High-Precision Floating-Point Arithmetic in Scientific Computation

  • Authors:

  • David H. Bailey

  • Affiliations:

  • Lawrence Berkeley National Laboratory

  • Venue:
  • Computing in Science and Engineering
  • Year:
  • 2005

Quantified Score

Hi-index 0.00

Visualization

Abstract

At the present time, IEEE 64-bit floating-point arithmetic is sufficiently accurate for most scientific applications. However, for a rapidly growing body of important scientific computing applications, a higher level of numeric precision is required: some of these applications require roughly twice this level; others require four times; while still others require hundreds or more digits to obtain numerically meaningful results. Such calculations have been facilitated by new high-precision software packages that include high-level language translation modules to minimize the conversion effort. These activities have yielded a number of interesting new scientific results in fields as diverse as quantum theory, climate modeling and experimental mathematics, a few of which are described in this article. Such developments suggest that in the future, the numeric precision used for a scientific computation may be as important to the program design as are the algorithms and data structures.