[Beowulf] Teaching Scientific Computation (looking for the perfect text)

[Tim Cutts]

On 21 Nov 2007, at 12:09 am, Joe Landman wrote:

> Jim Lux wrote:
>
>>> Octave: http://www.gnu.org/software/octave/
>> Octave is nice, but.... the graphics are MUCH better in Matlab, and
>
> Agreed.  Octave uses Gnuplot which is OK.
>
>> there's all those toolboxes full of cool stuff (signal processing,  
>> control systems, maps, etc.)
>
> Octave has quite a few as well, though they are not identical to the  
> Matlab versions.
>
>> And, an academic license for Matlab is only $100.  That's less than  
>> the
>
> Anyone need an adjunct ... :) I was under the impression that the  
> license fees were much stiffer than that.  For a cluster, $100*N for  
> N = 16 .. 32 is not bad at all.  Or am I missing something.

If you don't meet their very narrow definition of what constitutes  
"academic" it gets very expensive indeed.  And don't forget you have  
to pay extra for every toolbox you want.

> Give someone a good, powerful, flexible tool, and get out of their  
> way.  Lower barriers to use.  This has become somewhat of a mantra  
> with me as of late.

Well, it makes the user happy.  For a while.  Until a year or two down  
the line with a deadline imminent, they try to run R/Matlab/whatever  
on a dataset that's four times the size and realise they need to start  
from scratch because these programs don't handle large datasets that  
well, and they have to go and write it in C/FORTRAN anyway.

Tim


-- 
 The Wellcome Trust Sanger Institute is operated by Genome Research 
 Limited, a charity registered in England with number 1021457 and a 
 company registered in England with number 2742969, whose registered 
 office is 215 Euston Road, London, NW1 2BE.