Is that available for non-Intel platforms? Given it is from Intel, I would be surprised... More importantly, it doesn't seem to support arbitrary precision decimal floating point, which the decNumber package does, just the 3 new IEEE formats (32-bit, 64-bit, 128-bit), which the decNumber package has. If it performs much better, it might be interesting to make a package that used the Intel library optionally on Intel platforms (if it is restricted to Intel) for the 3 IEEE formats, and use the decNumber package (which was written originally by Mike Cowlishaw, whose decimal arithmetic specification is at the heart of the new IEEE decimal floating point standard)
Scott On Friday, April 17, 2015 at 12:45:14 PM UTC-4, Steven G. Johnson wrote: > > > > On Friday, April 17, 2015 at 12:03:43 PM UTC-4, Scott Jones wrote: >> >> (and also to help adding a decimal float package using decNumber, which >> is under the ICU license >> > > Why not the Intel library > > > https://software.intel.com/en-us/articles/intel-decimal-floating-point-math-library > > > It's under 3-clause BSD. >
