Computing Integer Powers in Floating-Point Arithmetic

Peter Kornerup, Vincent Lefevre, Jean-Michel Muller
2007 Asilomar Conference on Signals, Systems and Computers. Conference Record  
We introduce two algorithms for accurately evaluating powers to a positive integer in floating-point arithmetic, assuming a fused multiply-add (fma) instruction is available. We show that our log-time algorithm always produce faithfully-rounded results, discuss the possibility of getting correctly rounded results, and show that results correctly rounded in double precision can be obtained if extended-precision is available with the possibility to round into double precision (with a single
more » ... with a single rounding). * Peter Kornerup is with SDU, Odense, Denmark; Vincent Lefèvre and Jean-Michel Muller are with
doi:10.1109/acssc.2007.4487226 fatcat:lcgdxdht4ndopp5vowuaczadou