Floating point optimizations for pow() (note: all the histograms could
   totally be integers):
   - [1]https://stackoverflow.com/questions/6475373/optimizations-for-pow-
   with-const-non-integer-exponent
   - [2]https://martin.ankerl.com/2007/10/04/optimized-pow-approximation-f
   or-java-and-c-c/
   I think it was the second answer to the stackoverflow link that
   contained a pow() implementation that could be made arbitrarily
   accurate and took just a few instructions to execute.
   Integer optimizations for pow():
   - [3]https://gist.github.com/orlp/3551590
   - [4]https://github.com/ampl/gsl/blob/master/sys/pow_int.c
   I'm not yet finding a library that contains these from google.  The
   optimizations are almost a decade old.

References

   1. https://stackoverflow.com/questions/6475373/optimizations-for-pow-with-const-non-integer-exponent
   2. https://martin.ankerl.com/2007/10/04/optimized-pow-approximation-for-java-and-c-c/
   3. https://gist.github.com/orlp/3551590
   4. https://github.com/ampl/gsl/blob/master/sys/pow_int.c