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

[Michael Jinks]

On Tue, Nov 20, 2007 at 12:33:43PM -0600, Nathan Moore wrote:
> I regularly teach a college course in a physics department that deals with
> scientific computation.  After students take the course, I expect that
> they'll be able to write simple "c-tran" style programs for data analysis,
> write basic MD or MC simulations, and be fairly fluent in Mathematica.
<snip>

I have a related (I hope) question.

I don't teach a course, but some colleagues and I are in the position
of "computation ambassador" to a range of researchers (faculty, grad,
and undergrad).  Our semi-imaginary "classic case" is a professor in the
humanities who might have an abstract sense that large-scale computation
could aid their research, but no idea how to actually apply the
available technology.  In the worst case, this person regards anything
with a command line as terra incognita, and will require some serious
hand holding early on.

We can provide the hand holding, and we know a fair amount about
computers, clusters, what's (im)possible, but none of us have a deep
technical background in the "research" part of research computing, so
we've spent a lot of time lately asking ourselves how to bridge the gap.

So I hope I'm not hijacking the thread, but beyond issuing a "me too" to
Nathan's textbook question I'd also like to know if anybody has some
general tips on how to get started, a sort of "adult primer" on research
computing which we could benefit from ourselves as well as passing along
to the curious.

Thanks,
-j