We searched for the worst cases for correct rounding of the exponential function in the IEEE 754r decimal64 format, and computed all the bad cases whose distance from a breakpoint (for all rounding modes) is less than $10^{-15}$,ulp, and we give the worst ones. In particular, the worst case for $|x| geq 3 imes 10^{-11}$ is $exp(9.407822313572878 imes 10^{-2}) = 1.098645682066338,5,0000000000000000,278ldots$. This work can be extended to other elementary functions in the decimal64 format and allows the design of reasonably fast routines that will evaluate these functions with correct rounding, at least in some domains.
@InProceedings{lefevre_et_al:DagSemProc.06021.11, author = {Lef\`{e}vre, Vincent and Stehl\'{e}, Damien and Zimmermann, Paul}, title = {{Worst Cases for the Exponential Function in the IEEE 754r decimal64 Format}}, booktitle = {Reliable Implementation of Real Number Algorithms: Theory and Practice}, pages = {1--10}, series = {Dagstuhl Seminar Proceedings (DagSemProc)}, ISSN = {1862-4405}, year = {2006}, volume = {6021}, editor = {Peter Hertling and Christoph M. Hoffmann and Wolfram Luther and Nathalie Revol}, publisher = {Schloss Dagstuhl -- Leibniz-Zentrum f{\"u}r Informatik}, address = {Dagstuhl, Germany}, URL = {https://drops.dagstuhl.de/entities/document/10.4230/DagSemProc.06021.11}, URN = {urn:nbn:de:0030-drops-7483}, doi = {10.4230/DagSemProc.06021.11}, annote = {Keywords: Floating-point arithmetic, decimal arithmetic, table maker's dilemma, correct rounding, elementary functions} }
Feedback for Dagstuhl Publishing