[Beowulf] Re: Linux laptops, and M$ advertisement

[David Mathog]

From: "Peter St. John" <peter.st.john at gmail.com> wrote

> The ad is http://www.microsoft.com/windowsserver/facts/default.mspx?R=cf
> 
> The testamonials all seem to be equivalent to "I"m a PhD in Rocket Science
> but setting up a Linux Cluster is too complex, so using MS saves me
time to
> do Rocket Science".

That's the way it sort of looks on the surface, but read a little deeper
and it comes out differently.  For instance, follow the link to
the Virginia Polytechnic case study 

  http://www.microsoft.com/windowsserver/facts/casestudies/vti.mspx

and it says "Using desktop PCs, graduate students at the Computational
Bioinformatics and Bioimaging Laboratory (CBIL) didn’t have enough
computing power to run complex data and statistical analysis programs
quickly."   Which to me means " they are probably running matlab on
Windows and may not know anything else".  Sure enough, dig around
with google a bit and one finds this:

  http://www.ari.vt.edu/about_ARIHPCC.html

where Matlab is the first application listed.  Matlab is
available for linux too, but ESP is not required to guess what 
happened:

1.  Matlab too slow for some analysis on desktop Windows.
2.  Users beg for more power.
3.  Wow, we can do this on a windows cluster, then our Windows 
administrator won't need to learn anything new.
4.  Buy a windows cluster.

and possibly...

5. Oooh, did Microsoft just offer to subsidize this, even better!

Can't say if the last one happened or not.  Also couldn't find anything
on their site indicating they ever tried a linux cluster instead.
Note also the meeting where they presented their results...

There's nothing wrong with using a Microsoft cluster to expand
the power of an existing Windows application if that's a lot less work
than porting it to linux.  However, the case study would have
been a lot more convincing had there been any info that showed
a valid comparison for cost or speed between the Microsoft cluster
and a linux cluster running on the same hardware. 

Regards,

David Mathog
mathog at caltech.edu
Manager, Sequence Analysis Facility, Biology Division, Caltech