(j3.2006) What do typical processors do?

Steve Lionel steve
Tue Jul 18 19:12:52 EDT 2017


I would absolutely use *GEMM from an optimized library and not assume the
compiler will recognize that particular paradigm. A good library, such as
Intel MKL, can make use of parallelism as well as processor-specific
instruction sets. *GEMM, in particular, gets LOTS of attention from library
developers.

Steve

On Tue, Jul 18, 2017 at 7:00 PM, Van Snyder <Van.Snyder at jpl.nasa.gov> wrote:

> Should I expect a processor to optimize
>
>   C = matmul ( A, conjg(transpose(B)) )
>
> without making two or three temps, or should I write a matmul that has
> options to do that, or use *GEMM?
>


-- 
.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.j3-fortran.org/pipermail/j3/attachments/20170718/346c1279/attachment.html 



More information about the J3 mailing list