Title: Can we couple Mesh Generation  and Partitioning Problems?

Speaker: Nikos P Chrisochoides, Notre Dame


In this talk, we are making an attempt to formulate a new approach for
the parallel generation and  partitioning of unstructured  3D Delaunay
meshes.  The new approach couples the mesh generation and partitioning
problems into a single optimization problem.  Traditionally, these two
problems are solved  separately,  first generating the  mesh  (usually
sequentially)  and then partitioning the  mesh, either sequentially or
in parallel.  In the traditional  approach,  the overheads due to  I/O
and data movement  exceed  90% of the total  execution  time.  Even if
parallel   partitioning   schemes   are     employed, data   movement,
synchronization, and data  structure  translation overheads are  high;
for applications which require frequent  remeshing (e.g.  crack growth
simulations),  these overheads are  prohibitive.   We present a method
for solving the mesh partitioning and placement problem simultaneously
with the mesh   generation  problem.  By  eliminating unnecessary  and
redundant cache, local, and remote memory accesses, we can speedup the
generation   and redistribution  process, for  very   large meshes, by
almost an order of magnitude  compared to traditional approaches.  Our
results show  that we can  achieve nearly perfect equi-distribution of
mesh elements  over the processors,  while maintaining reasonably good
separator  size,  all while   improving the quality    of the  mesh by
eliminating many of   the  problems inherent  in traditional  parallel
constrained mesh generation.


URLs: 1) www.cse.nd.edu/~nikos
----  2) http://www.medusa.nd.edu (under construction, will be completed
                                   by the end of the 1999)

Bio: 
----

Dr. Chrisochoides' work focuses on  the integration of algorithmic and
applications-driven research in parallel   computing with research  in
parallel runtime software  systems for high  end architectures.  He is
working on  the development  of guaranteed-quality industrial strength
parallel unstructured mesh generation.

Nikos Chrisochoides received his Ph.D in Computer  Science and M.Sc in
Applied Mathematics from Purdue University. He is the first Alex Nason
Postdoctoral Fellow  in  Computational Sciences at Northeast  Parallel
Architectures Center,  Syracuse   University;  there  he worked   with
Geoffrey C.  Fox on parallel  runtime support systems for adaptive and
irregular computations.  From  1995  to 1997 he  worked as  a Research
Associate  in  the Computer   Science and Advanced  Computing Research
Institute, Cornell University.  At   Cornell he continued his work  on
parallel runtime systems and  investigated the integration of parallel
compiler and problem  specific environment technologies; He joined the
faculty   of  the University  of    Notre  Dame (Computer Science  and
Engineering) as an Assistant Professor, in  1997.  In 1998 he received
the NSF Career  Award. Also, from 1994  to  1996 he has been  Visiting
Scientist    at Institute for  Computer    Applications in Science and
Engineering at NASA/Langley Research Center, Virginia.

-------------------------------------------------------------------------
Nikos P. Chrisochoides             Phone: (219) 631-8811
Assistant Professor                FAX:   (219) 631-9260
Computer Science and Engineering   URL: http://www.cse.nd.edu/~nikos
University of Notre Dame           Email: nikos@cse.nd.edu
Notre Dame, IN 46556
-------------------------------------------------------------------------