On Saturday, 24 September 2016 at 17:46:07 UTC, Ilya Yaroshenko
wrote:
On Saturday, 24 September 2016 at 13:49:35 UTC, Andrei
Alexandrescu wrote:
On 9/24/16 9:18 AM, John Colvin wrote:
On Saturday, 24 September 2016 at 12:52:09 UTC, Andrei
Alexandrescu wrote:
Could you also add a comparison with SciPy? People often say
it's just
fine for scientific computing.
That's just BLAS (so could be mkl, could be openBLAS, could
be netlib,
etc. just depends on the system and compilation choices)
under the hood,
you'd just see a small overhead from the python wrapping.
Basically,
everyone uses a BLAS or Eigen.
I see, thanks. To the extent the Python-specific overheads are
measurable, it might make sense to include the benchmark.
An Eigen comparison would be interesting.
That'd be awesome especially since the article text refers to
it.
Andrei
Eigen was added (but only data, still need to write text).
Relative charts was added. You was added "Acknowledgements"
section --Ilya
It would also be interesting to compare the results to Blaze [1].
According to https://www.youtube.com/watch?v=hfn0BVOegac it is
faster than Eigen and on some instances faster than even Intel
MKL.
[1]: https://bitbucket.org/blaze-lib/blaze