4 ms·
The numbers they were working with were way out past the bounds of doubles, and a significant fraction of the way to the largest quadruple precision number (sec
by GFK_of_xmaspast 12y ago
The numbers they were working with were way out past the bounds of doubles, and a significant fraction of the way to the largest quadruple precision number (second column, bottom of page 1250).
- conistonwater 12y agoLibraries, such as mpfr, that implement multiple-precision floating-point arithmetic are completely standard. At best maybe you can argue that it introduces extra programming work over regular doubles, but in Mathematica it is no work at all as it has an arbitrary-precision floating-point type.