Interesting. I think I will mention it in my thesis :)
Have we mentioned we made did a similar, but not so comprehensive
comparison of a kernel implemented in
fortran, numpy, pythran and numba? See Figure 9 in:
https://openresearchsoftware.metajnl.com/articles/10.5334/jors.238/
Cheers,
Ashwin
On 7/29/19 12:54 PM, Serge Guelton wrote:
Hi folks,
this article benchmarking various implementation of a scientific kernel is
especially well documented, it's worth having a look!
http://journals.iucr.org/j/issues/2019/04/00/gj5229/gj5229.pdf
++
Serge