--- Lots of good feedback on the request for comments
--- Workshop announcement for Dec 11 at NCSA attached
--- Short questions (please send quick answers to summers@astro.princeton.edu)
- are you attending the Dec 11 workshop?
- do you have suggestions for workshop topics & speakers?
- will you be at Supercomputing 95?
************************************************************************
FEEDBACK
--- Cosmology is not the only/primary objective.
Cosmology was used as an example in the RFC. It is, in fact, the
hardest application for GRAPE to justify (and many respondents pointed
this out). A couple opinions stated that GRAPE would offer no advantage
for cosmology (essentially arguing that the calculations are RAM
limited, so GRAPE's speed is no advantage).
--- Cosmology applications need hydrodynamics.
The pure gravity calculations in cosmology will start to run out of
interesting problems in the next five years or so. Hydro will be where
the action is. Most are optimistic that relatively straightforward ways
of incorporating SPH without loading the front end can be found.
--- Galaxy simulations will find big benefit.
Simulations on the galaxy scale (mergers and such) have similar
structure as cosmology simulations, but fewer particles and thus much
more CPU limited. Should see strong benefit from GRAPE. Worries are
about hydro and radiative and other processes calculated on the front
end and balancing this load with GRAPE speed.
--- It will be important to have "highlight" projects.
To get big funding, there will have to be some target problems
which the new machine will solve that no other method can get. For GRAPE
4, it was the core oscillations in a globular cluster. Similar targets
in all regimes will have to be defined. The computational goal of a
petaflops machine is flashy, but not enough.
--- Funding through the NSF centers re-competition?
The NSF is expected to announce a re-competition for the NSF
supercomputing centers (perhaps already?). Our project can serve as a
high profile adjunct to one of the centers. Both Larry Smarr of NCSA and
Ralph Roskies of PSC have expressed interest in the project. Details
will await the structure of the re-competition.
************************************************************************
* *
* Workshop on GRAPE:TNG Design, Applications, and Funding *
* *
* National Center for Supercomputing Applications *
* Champaign/Urbana, Illinois *
* *
* December 11, 1995 *
* *
************************************************************************
LOCATION
--------
The National Center for Supercomputing Applications is located in the
Beckman Institute on the campus of the University of Illinois at
Urbana/Champaign. The address is 6th and University, however the main
entrance is on the side of the building facing away from University Ave.
TRANSPORTATION
--------------
Champaign/Urbana airport is serviced by several major airlines. There is
a shuttle service at the airport which meets incoming flights. Amtrak
rail service is also available to Champaign. Champaign/Urbana is about
a three hour drive from Chicago.
HOTEL
-----
Hotel accommodations will be at the Hampton Inn located on University
Ave, one half block from NCSA. Please contact Jean Soliday
(jsoliday@newton.ncsa.uiuc.edu) about reservations.
COSTS
-----
There is no fee for the workshop. There is also no funding to provide
support for travel.
CONTACTS
--------
Piet is out of town, but in email contact until the meeting. Jun, Mike,
and Frank will be at Supercomputing 95 the week before the meeting.
Piet Hut piet@sns.ias.edu 609-734-8075
Jun Makino makino@kyohou.c.u-tokyo.ac.jp
Steve McMillan steve@zonker.drexel.edu 215-895-2723
Mike Norman norman@ncsa.uiuc.edu 217-244-6099
Frank Summers summers@astro.princeton.edu 609-258-3810
DISCUSSION GUIDELINES
---------------------
We hope to keep this workshop somewhat informal, with lots of
opportunity to discuss topics to the depth necessary for a clear
proposal picture. Participants are expected to read up a bit on GRAPE
background (see references below) and have ideas in mind on how the
hardware can be used/adapted to their purposes as well as ideas and
estimates of their technical requirements (CPU speed, memory, I/O
bandwidth, storage). If we come prepared, a good overview of the project
will emerge, leading to a very strong proposal.
The people listed in the following schedule will give introductory talks
about the subject and then lead discussions as appropriate. This
schedule is a target structure, while the true course may deviate. The
day is divided into three parts: (1) define what we've got to work with,
(2) figure out how it can best be adapted and used, and (3) combine
these ideas into a coherent proposal outline.
DRAFT SCHEDULE
--------------
8:30 Coffee & muffins
9:00 Overview of Grape history -- Makino
9:45 Overview of supercomputing history and future -- Smarr
10:30 Current design ideas for GRAPE:TNG -- ???
11:15 GRAPE:TNG applications, stellar dynamics -- Hut
11:45 GRAPE:TNG applications, galaxy dynamics -- ???
12:15 Lunch
1:00 GRAPE:TNG applications, cosmology -- Summers
1:30 Discussion on optimization of design and applications
2:15 NSF Centers re-competition -- Smarr/Norman
3:00 Discussion on funding proposal
5:30 Dinner and discussion until we drop
REFERENCES
----------
***** Web pages *****
- GRAPE-4 system:
http://butterfly.c.u-tokyo.ac.jp:8080/pub/people/makino/grape4.html
- Grand Challenge Cosmology Consortium:
http://zeus.ncsa.uiuc.edu:8080/GC3_Home_Page.html
- HPCC group at University of Washington:
http://www-hpcc.astro.washington.edu/
- The NASA HPCC Project at Los Alamos National Lab
***** Papers *****
- GRAPE 4 - Makino & Taiji (SC 95)
http://butterfly.c.u-tokyo.ac.jp:8080/pub/people/makino/papers/sc95.ps
- GRAPE 4 - Makino et al. (SIAM PPSC 95)
http://butterfly.c.u-tokyo.ac.jp:8080/pub/people/makino/papers/siampp95.ps
- GRAPESPH code - Steinmetz
http://xxx.lanl.gov/abs/astro-ph/9504050/
- P3MG3A cosmology code - Brieu, Summers, & Ostriker
ApJ, 453, 566 (1995)
http://xxx.lanl.gov/abs/astro-ph/9411001/
- PASJ SPECIAL FEATURE ON GRAPE - Vol 45, No 3 (1993)
- GRAPE Overview - Ebisuzaki et al
PASJ 45, 269 (1993)
- SPH and GRAPE 1 - Umemura et al.
PASJ 45, 311 (1993)
- GRAPE 3 - Okumura et al.
PASJ 45, 329 (1993)
- HARP - Makino, Kokubo, & Taiji
PASJ 45, 349 (1993)
- Treecode and GRAPE - Makino
PASJ 43, 621 (1991)
- GRAPE 2 - Ito et al.
PASJ 43, 547 (1991)
- GRAPE 1 - Sugimoto et al.
Nature 345, 33 (1990)
LIST OF INTERESTED PARTIES
--------------------------
Attending
Josh Barnes barnes@zeno.ifa.hawaii.edu
Philippe Brieu philippe@umich.edu
David Chernoff chernoff@spacenet.tn.cornell.edu
Bruce Elmegreen bge@watson.ibm.com
Gus Evrard evrard@pablo.physics.lsa.umich.edu ?
Carlos Frenk C.S.Frenk@durham.ac.uk ?
Douglas Heggie heggie@castle.ed.ac.uk
Lars Hernquist lars@corona.ucsc.edu
Piet Hut piet@sns.ias.edu X
George Lake lake@astro.washington.edu
Doug Lin lin@lick.UCSC.EDU ?
Jun Makino makino@kyohou.c.u-tokyo.ac.jp X
Steve McMillan steve@zonker.drexel.edu X
Jordi Miralde-Escude jordi@sns.ias.edu
Michael Norman norman@ncsa.uiuc.edu X
Jerry Ostriker jpo@astro.princeton.edu
Stan Peale peale@io.ucsb.edu
Frazer Pearce frazerp@central.sussex.ac.uk
Gerry Quinlan gquinlan@physics.rutgers.edu ?
Ralph Roskies roskies@a.psc.edu
Steinn Sigurdsson steinn@ast.cam.ac.uk
Larry Smarr smarr@ncsa.uiuc.edu X
Matthias Steinmetz mhs@mpa-garching.mpg.de
Thomas Sterling tron@cesdis1.gsfc.nasa.gov X
Tatsushi Suginohara tatsushi@yayoi.phys.s.u-tokyo.ac.jp
Frank Summers summers@astro.princeton.edu X
Yasushi Suto suto@phys.s.u-tokyo.ac.jp
Makoto Taiji taiji@chianti.c.u-tokyo.ac.jp X
Peter Teuben teuben@astro.umd.edu X
David Weinberg dhw@picasso.mps.OHIO-STATE.EDU