Background Memo to Lab Computational Leads

Background Memo to Lab Computational Leads

Dear XYZ,

The subject concerning which I write was introduced in an e-mail from

Walt Polansky on April 2, and has both year-long and very immediate

implications.

I was charged by the Office of Science of the DOE with helping to

"identify rich and fruitful directions for the computational sciences from

the perspective of scientific and engineering applications." This is an

effort that will require lots of time and energy to do well, and I have no

intention of wasting my time, yours, or our colleagues on anything less

than a job with which the Office is satisfied, and which gives Ray Orbach

the strong scientific case he wants and needs to continue to promote

computational science as a prominent part of the DOE research portfolio.

The first deliverable of this effort is a report desired by July 30 - in

time to be useful in setting FY 2005 budget priorities, no doubt, but to

be useful beyond. This report will not be the first of its kind, and it

certainly will have no pretense to be the last. My view of a successful

outcome is that it should help to expand the folds of the scientific tent

that is SciDAC, rounding out this initiative in terms of disciplinary

areas addressed (both fundamental science and enabling technology), and

motivating the acquisition of hardware to support its growth. It may or

may not be packaged as SciDAC, but it seems foolish to ignore the initial

success and popularity of this initiative, or to let be forgotten that the

original SciDAC vision included some program components that are not yet

implemented.

The larger goals of our effort will require much more than a report with

outstanding content and presentation. The construction of the report must

itself be a mechanism for creating scientific vision and consensus and for

uniting the DOE SC community. The latter is perceived by some of

influence to be fragmented and unable to play under a single baton. This

is inaccurate, in my view, but the same perception is also obtained by

catching fine and dedicated musicians showing off in the warm-up hall. We

need only to find some music we can all play that contains our solos, to

be sure, but sounds great out in the public hall. Our product must be

approachable on many levels, by connoisseurs and scientific amateurs.

I will need a core group to help me plan the major workshop described by

Walt for our effort. The workshop will most likely be in June. To plan

for it, we need a pre-workshop, and the latter has just been scheduled.

Dr. Orbach would like to give the charge to as many members of the core

group as we can assemble on Wednesday, April 23, in Washington, DC. I

recognize that this is very short notice, so we may not have a full house.

I have made up a "wish list" of 16 people, which has the properties of

spanning the five program offices of the Office of Science fairly well, of

including someone from each of the labs where major computing is done, and of consisting primarily of practicing scientists, who can spin the

technical rolodex and lead writing teams in their disciplines when

assignments are made. Not every invitee for April 23 need be part of the

ultimate core group, but we should try to identify suitable participants

as early as possible.

Ideally, I would like to approach invitees right away, and no later than

by Tuesday evening, April 15, due to the tightness of the calendar.

Thanks for your prompt attention and very best regards,

David