On 06/25/2012 03:51 PM, Akshay Srinivasan wrote:
> This is an automated email from the git hooks/post-receive script. It was
> generated because a ref change was pushed to the repository containing
> the project "matlisp".
>
> The branch, tensor has been updated
> via 0f1b57f2c90f00aac4aa5ea6e7240ae69690409f (commit)
> from 8bb55ab5b53aa70785619511fcd6457b3bb79401 (commit)
>
> Those revisions listed above that are new to this repository have
> not appeared on any other notification email; so we list those
> revisions in full, below.
>
> - Log -----------------------------------------------------------------
> commit 0f1b57f2c90f00aac4aa5ea6e7240ae69690409f
> Author: Akshay Srinivasan <aks...@gm...>
> Date: Mon Jun 25 15:46:15 2012 +0530
>
> Added mod-loop for doing multi-index loops seemlessly. Must add more checks.
> Added tensor-copy; can run things at about 3x BLAS (dcopy) speed (!) :)
Sorry the point about 3x BLAS speed is plainly wrong. It is about 3-4
times faster running on SBCL compared to fortran/C, when working with
*strided* tensors. Basically, I mean that multi-index loops in C are
much slower than on SBCL.
BLAS-dcopy is only useful for copying linearly ordered tensors, and is
about 10 times faster than using a multi-index loop in lisp; which is
hardly surprising.
Akshay
|
