[Beowulf] Heat transfer simulation

Rolando Espinoza La Fuente rhodas at gmail.com
Tue Apr 12 08:23:49 PDT 2005


On Apr 12, 2005 12:52 AM, Mark Hahn <hahn at physics.mcmaster.ca> wrote:
> > The purpose of the simulation is just justify and demostrate the
> > cost-performance and benefits of a beowulf cluster.
> 
> this is a terribly bad idea.  building a cluster MUST be driven
> by real, demonstrable needs, preferably several independent
> researchers who are in physical pain because their simulations
> are too slow ;)

(sorry if my english isn't readable)

Thanks for your comments.

I'm a 20 years old math student (second year) on a public university
(www.umss.edu.bo) of Bolivia (south america),  my career aren't too
apply about optimization and simulation as far i know, actually is
there a work about "routes", finding the short path between two
points.

Of course here isn't "researchers" about parallel programming neither
high performance computing. Because i have some (adv) skills on
"linux" systems and i talk with my academic director about a project
with beowulf clusters, he has interest about computational cost on
heuristics vs exacts alghoritms.

The easy part is making a beowulf cluster (i know), the most dificult
is use the resources, making applications, research, etc.

We haven't "researchers who are in physical pain because their
simulations are too slow ", but we have to. I want to be a researcher
about modeling real problems and using computational resources.

If you have more comments about this, im glad to read.

> but yes, heat-flow simulations are commonly run on clusters;
> as far as I understand, they're not nearly as challenging
> as simulations that involve mass flow or phase change
> (gases, fluids, melting, etc).

The "heat-flow simluation" that i asked is because my academic
director say: "would be interesting seeing a heat transfer simulation
over the cluster and compare the same simulation in a sinlge PC". So
im researching.

PD: If here were researchers about high performance computing and
applied math with computational resources, i'll not asking here.

-- 
~Rolando



More information about the Beowulf mailing list