[Beowulf] Stroustrup regarding multicore

[Gerry Creager]

Ed Hill wrote:
> On Tue, 26 Aug 2008 14:29:11 -0400 "Perry E. Metzger" wrote:
>> I think part of the issue is that most people doing scientific
>> computing don't have computer science backgrounds, which is a
>> shame.
> 
> 
> Formal CS training can certainly help but I don't view it as a
> requirement.  I've met some remarkably capable programmers who have
> little or no formal CS background.
> 
> And I don't mean to sound like a cheezy motivational speaker but
> if each of us is willing to study the texts a bit, read others' code,
> and learn-by-doing then I don't think there is anything that prevents us
> from becoming capable programmers.  Or engineers.  Or chefs.  Or system
> administrators.  Or..., well, you get the idea.  :-)

I'm loathe to get into this, but:  A lot of computational scientists 
came to the PhD program seeking degrees in numerical modeling in 
meteorology or climatology, or some computational realm: chemistry, 
bioinformatics, petroleum engineering, etc., never dreaming they'd 
become a system administrator and code wrangler, and that their degree 
was granted when an advisor deemed their efforts in keeping the hardware 
and others' software going were deemed sufficient: They'd suffered enough.

They may or may not have sufficient computer science training when they 
started, and they may or may not get formal training in the course of 
their degree process, but don't discount the computer science training 
they acquired along the way.

For the record, I prefer to not sacrifice a new PhD student to the HPC 
gods, but I don't control everyone else.  And, I find that some, when 
they see what goes into HPC, are seduced by the Dark Side.  If they work 
for me, they end up with a somewhat different degree plan, as they have 
to have a couple of C courses, exposure to FORTRAN, and some formal 
exposure to MPI and OpenMP.  Others, however, assume osmosis is 
sufficient for their students, and they don't quite get as good a 
result, in my opinion.

This, however, is the basis for a plan several of us are fostering: 
Cluster condominiums allow for leveraging economies of scale in hardware 
purchase along with professional system administration.  Students 
interested in working on computational science in a pragmatic manner can 
work here and also continue with their nominal studies.  OR change to a 
curriculum in computational science.

gc
-- 
Gerry Creager -- gerry.creager at tamu.edu
Texas Mesonet -- AATLT, Texas A&M University
Cell: 979.229.5301 Office: 979.862.3982 FAX: 979.862.3983
Office: 1700 Research Parkway Ste 160, TAMU, College Station, TX 77843