Re: Optimizations in C/C++ code doing matrix multiplications

lars.gregersen@it.dk (Lars Gregersen)
6 Apr 2002 23:43:03 -0500

          From comp.compilers
Related articles
Optimizations in C/C++ code doing matrix multiplications ele1@gmx.ch (H. Ellenberger) (2002-03-22)
Re: Optimizations in C/C++ code doing matrix multiplications ele1@gmx.ch (H. Ellenberger) (2002-03-31)
Re: Optimizations in C/C++ code doing matrix multiplications marjan.sterk@ijs.si (2002-03-31)
Re: Optimizations in C/C++ code doing matrix multiplications mark@marklacey.com (2002-04-06)
Re: Optimizations in C/C++ code doing matrix multiplications lars.gregersen@it.dk (2002-04-06)
Re: Optimizations in C/C++ code doing matrix multiplications terryg@qwest.net (Terry Greyzck) (2002-04-07)
| List of all articles for this month |
From: lars.gregersen@it.dk (Lars Gregersen)
Newsgroups: comp.compilers
Date: 6 Apr 2002 23:43:03 -0500
Organization: Cybercity
References: 02-03-151
Keywords: optimize
Posted-Date: 06 Apr 2002 23:43:02 EST

On 22 Mar 2002 21:03:51 -0500, "H. Ellenberger" <ele1@gmx.ch> wrote:


>When matrix multiplication is programmed the same way it is
>mathematically defined, then two nested loops running through all
>valid indices are used. ...


>Which compilers are smart enough to automagically optimize this kind
>of data acess by introducing hidden pointer variables (in registers)
>incremented and thus avoiding most multiplications?
>[It's my impression that people who care about fast and, accurate matrix
>multiplication use linpack rather than reinventing it. -John]


Check out ATLAS (Automatically Tuned Linear Algebra Software)


http://www.netlib.org/atlas/


For modern computers (and large matrices) memory access is what counts
when doing matrix operations efficiently.


    Lars


Lars Gregersen
lars.gregersen@it.dk

Post a followup to this message

Return to the comp.compilers page.
Search the comp.compilers archives again.