Newsgroups: comp.parallel.mpi
From: dana@infi.net (Dana Hammond)
Subject: CSC-NASA Job Openings
Organization: InfiNet
Date: 31 Mar 1995 23:10:43 GMT
Message-ID: <3li25j$dgq@lucy.infi.net>




Two high performance computing vacancies exist at the Computer Sciences 
Corporation (CSC) Hampton office in support of NASA Langley Research Center.  
Please forward this to appropriate personnel.

Please submit all resumes/inquires in writing or fax to 

	Computer Sciences Corporation
	c/o Nancy Schott
	3217 North Armistead Avenue
	Hampton, Virginia  23666
        fax:   (804) 766-2571
        phone: (804) 766-8231

Date Originated: 3/27/95

--------------------------------------------------------------------------------
Requisition Number: 497-003
Targeted Hire Date: Immediately

I. Duties and Responsibilities

Develop a prototype library to support programming in a distributed-memory 
computing environment and demonstrate with a medium fidelity Navier-Stokes 
flow solver combining the use of the heterogeneous platforms of the IBM SP-2 
and a set of workstations.  The library is essentially a message passing 
library with some higher-level primitives to reduce the burden of programming 
directly with a low-level message passing library such as the Parallel Virtual 
Machine (PVM) and Message Passing Interface (MPI).  

II. Critical Knowledge

An in-depth knowledge of coarse and fine grain parallelism, involving both
data and task parallelism; an in-depth knowledge of MPI and PVM using the
IBM SP-2, cluster of workstations, and Intel Paragon; and knowledge of the
various issues of parallel and distributed programming including scalability, 
load balancing, data and computational decomposition, and communication 
overhead characteristics.

III. Critical Skills

Significant experience in developing, validating, and optimizing large
parallel, scientific application codes involving computational fluid
dynamics (CFD) or multidisciplinary optimization on the IBM SP-2 and Intel
Paragon, and experience with the networked workstation programming
environment is necessary.  

IV. Educational/Professional Affiliations

A Bachelor of Science Degree in Computer Science, Mathematics, Engineering,
or Physical Science; eight years experience in the development of
scientific applications for high performance computing systems, with at
least three years experience on highly parallel or distributed parallel
systems; and at least two years experience on CFD and multidisciplinary
design and optimization applications.

--------------------------------------------------------------------------------

Requisition Number: 497-002
Targeted Hire Date: June 1, 1995

I. Duties and Responsibilities

Develop and demonstrate a high fidelity viscous flow capability on distributed
workstations to achieve significant jump in compute cycles utilizing coarse-
grain parallelization on existing workstations.  Specifically develop and
document the required software and communications libraries for a block-
structured Navier-Stokes flow solver.  Modify the multiblock version to run on
a heterogeneous workstation cluster demonstrating load-balancing strategies.

II. Critical Knowledge

An in-depth knowledge of CFD, CSM, multidisciplinary design and optimization
techniques; knowledge of MPI and PVM, and various issues of parallel and 
distributed programming including scalability, load balancing, data and 
computational decomposition, and communication overhead characteristics.

III. Critical Skills

Significant experience in developing, validating, and optimizing large
parallel, scientific application codes involving CFD, CSM, and 
multidisciplinary optimization on highly distributed and parallel systems.

IV. Educational/Professional Affiliations

A Bachelor of Science Degree in Computer Science, Mathematics, Engineering,
or Physical Science; eight years experience in the development of scientific 
applications for high performance computing systems, with at least two years 
experience on highly parallel or distributed parallel systems; and at least 
two years experience on CFD, CSM, and multidisciplinary design and 
optimization applications.
--------------------------------------------------------------------------------


