[Numpy-discussion] performance matrix multiplication vs. matlab

Gabriel Beckers beckers@orn.mpg...
Sun Jun 7 07:43:44 CDT 2009


On Sun, 2009-06-07 at 18:37 +0900, David Cournapeau wrote:
> That's why compiling atlas by yourself is hard, and I generally advise
> against it: there is nothing intrinsically hard about it, but you need
> to know a lot of small details and platform oddities to get it right
> every time. That's just a waste of time in most cases IMHO, unless all
> you do with numpy is inverting big matrices,

I have been trying intel mkl and icc compiler instead, with no luck. I
run into the same problem during setup as reported here:

http://www.mail-archive.com/numpy-discussion@scipy.org/msg16595.html


Sigh. I guess I should not get into these matters anyway; I am just a
simple and humble user...

As far as I understand the Ubuntu atlas problems have been found for
complex types, which I don't use except for fft. I guess I'll continue
to use the ubuntu libraries then and hope for better days in the future.

Best, Gabriel



More information about the Numpy-discussion mailing list