NA Digest Monday, January 24, 1994 Volume 94 : Issue 4

Today's Editor:
Cleve Moler
The MathWorks, Inc.
moler@mathworks.com

Submissions for NA Digest:

Mail to na.digest@na-net.ornl.gov.

Information about NA-NET:

Mail to na.help@na-net.ornl.gov.

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


From: Bengt Littmann <littmann@tetibm3.ee.tu-berlin.de>
Date: Wed, 19 Jan 1994 10:38:23 +0100
Subject: Mode Matching Code

I have written a mode matching code that computes the scattering matrices
(S11, S12, S21, S22) of a _rectangular_ _waveguide_ _cross-section_ _jump_.
No assumptions as to symmetry have been made, provided that the cross-section
of the smaller guide lies within the boundaries of the bigger one.

The program is written in f77 and assumes that one has the LAPACK
compiled as library, as well as the BLAS (preferrably machine-BLAS). I have
used the cpp (C preprocessor) to make automatic conversion of declarations,
generic functions and LAPACK/BLAS-names to double precision. e.g. CGETRS ->
ZGETRS. It has been fairly well tested so far; we've calculated filters and
accelerator sections with it.

Anyone interrested send a mail to the address below and I'll be happy to
send the code. BTW, it comes as UNIX shell archive (shar).

Cheers, Bengt (email: littmann@tetibm3.ee.TU-Berlin.DE)


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

From: Cheryl Carey <cmmc@ama.caltech.edu>
Date: Wed, 19 Jan 94 16:25:23 PST
Subject: Low Rank Up/Down Dates

Does anyone know of any applications which involve multiple low rank
up/down dates to a nonsymmetric system of equations?.

Many thanks in advance

Cheryl Carey cmmc@ama.caltech.edu
Applied Math. 217-50, Caltech, Pasadena, CA 91125.


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

From: Matthias Mann <MATTHIAS@dosuni1.rz.Uni-Osnabrueck.DE>
Date: Thu, 20 Jan 94 14:06:55 +0000 (MEZ)
Subject: Query about Fortran 90 to C Translation

Dear Members of Na-net,
I would like to know if there is a Fortran -> C Translator, which is suitable
for use with Fortran90. I know that also the F90 compilers can do this.
The reason why I would like to know this is because we are using Linux
for the program development. Up to now netlibs f2c worked very well,
(as a preprocessor for the Gnu C Compiler) but it does not support F90.
Unfortunately we do not have a F90 compiler running under Linux, but we would
like to write programs in F90.
Any hints and suggestions are very welcome.

Dr. Matthias Mann
University of Osnabrueck
Physics Department
D-49069 Osnabrueck
Germany
Phone : +49-541-969-2684
Fax : +49-541-969-2670
Internet: MATTHIAS@TITAN.RZ.UNI-OSNABRUECK.DE


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

From: Sue J. Worden <sjw@bongo.cc.utexas.edu>
Date: Thu, 20 Jan 1994 08:59:47 -0600
Subject: SPEC Benchmarks

Hello,

I would appreciate pointers to materials that describe the
SPEC benchmarks, available to me via anonymous ftp, gopher,
or www. Something along the lines of Jack Dongarra's nice
document on the LINPACK benchmarks would be most helpful.

Thanks,

Sue Worden
Mathematical Services
Computation Center, Mail Code 12700
University of Texas at Austin
Austin, Texas USA 78712-1110
voice: (512) 471-3359
fax: (512) 471-1582
email: sjw@bongo.cc.utexas.edu


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

From: Robert W. Leland <rwlelan@cs.sandia.gov>
Date: Thu, 20 Jan 94 11:34:08 MST
Subject: Software for Partitioning Graphs

NEW SOFTWARE FOR PARTITIONING GRAPHS

Many problems which arise in the course of scientific computing can be
conveniently described in terms of graph partitioning. A prominent
example is the problem of decomposing a large, unstructured grid across
the processors of a parallel computer. Other applications include generating
nested dissection orderings for sparse matrix factorizations and devising
efficient circuit layouts.

We have developed a software package called "Chaco" which implements a
variety of graph partitioning algorithms. These include the spectral
bisection, quadrisection and octasection algorithms, the inertial method,
the Kernighan-Lin method, and a multilevel partitioner. Techniques can
mixed and matched as well, for example Kernighan-Lin can be applied to
refine a spectral partitioning.

A user's guide and papers describing the algorithms are available by
anonymous ftp to cs.sandia.gov in the directory pub/tech_reports/bahendr.
The software can be obtained at no charge under a research license agreement
and may also be licensed for commercial application. Interested parties
should contact one of the authors at the addresses given below.

Bruce Hendrickson (bahendr@cs.sandia.gov)
Rob Leland (rwlelan@cs.sandia.gov)
Mail Stop 1109
Sandia National Laboratories
Albuquerque, NM 87185


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

From: Joseph Traub <traub@cs.columbia.edu>
Date: Sat, 22 Jan 94 19:50:50 EST
Subject: A Matrix Eigenvalue Problem

The following eigenvalue problem was sent to me by Professor J. Luttinger,
a theoretical physicist at Columbia. Please send your responses to Professor
Luttinger at the e-mail address given at the end of his msg.

The problem is concerned with the representation of the largest
eigenvalue of a class of matrices. They are all symmetric and positive in the
sense of Perron. The largest eigenvalue is represented as an integral (from
zero to $2\pi$) of the log of a trigonometric polynomial. Here are a couple
that are fairly easy to work out:

1) The matrix is 2x2,

$$M_2 = \pmatrix{e^k&e^{-k}\cr e^{-k}&e^k \cr}\ , {\rm k}\ real}$$

The result is $2 log {\lambda_m\over 2} = {1\over 2\pi}
\int_0^{2\pi} d\theta log (cosh 2k + sinh 2k\ cos\ \theta).\ \ \lambda_m$ is
the largest eigenvalue of $M_2.$

2) This matrix is 4x4:

$$M_4 = \pmatrix{e^{2(k_1+k_2)}&e^{-k_1}&e^{-2k_2}&e^{k_1}\cr
e^{-k_1}&e^{2k_2}&e^{k_1}&e^{-2(k_1+k_2)}\cr
e^{-2k_2}&e^{k_1}&e^{2(k_1+k_2)}&e^{-k_1}\cr
e^{k_1}&e^{-2(k_1+k_2)}&e^{-k_1}&e^{2k_2}\cr}, k_i\ real}.$$

The result is

$$log {\lambda_m\over 4} = {1\over 2\pi} \int_0^{2\pi} d\theta log(cosh(2k_1)
cosh(2k_2) + sinh(2k_1) cos\theta + sinh(2k_2) cos2\theta).

(The largest eigenvalue of $m_4$ is again given by $\lambda_m.)$

There are other problems of the same general class, but with matrices
that get larger and larger. If done directly they involve seemingly hopeless
amounts of algebra, and yet I have reason to believe that a fairly simple
closed form exists.

J. Luttinger
arb@cuphyf@phys.columbia.edu


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

From: Georg Hebermehl <HEBERMEHL@iaas-berlin.d400.de>
Date: Mon, 24 Jan 1994 14:05:40 +0100
Subject: BLAS for the PowerPC

Hallo,
Does anyone know whether manufacturers' BLAS Level 1,2 and 3 are implemented
for the MPC 601 (Motorola PowerPC601 processor).
Thanks in advance
Georg Hebermehl (email: hebermehl@iaas-berlin.d400.de).


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

From: William Schiesser <wes1@Lehigh.edu>
Date: Sat, 22 Jan 1994 12:01:40 EST
Subject: New Book on Computational Mathematics

Now available:

Computational Mathematics
in Engineering and Applied Science:
ODEs, DAEs, and PDEs

by

W. E. Schiesser
Lehigh University

CRC Press, Inc.
2000 Corporate Blvd., N.W.
Boca Raton, FL 33431 USA

1-800-272-7737

ISBN: 0-8493-7373-5

W. E. Schiesser wes1@lehigh.edu (Internet) .
Iacocca Hall, Rm D307 (WES) (610) 758-4264 (office) .
Lehigh University (610) 758-4260 (secretary) .
111 Research Drive (610) 758-5057 (fax) .
Bethlehem, PA 18015 USA (please note the change in the .
area code from 215 to 610) .

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

From: Matthias Hocks <Matthias.Hocks@iampc4.mathematik.uni-karlsruhe.de>
Date: Mon, 24 Jan 94 11:51:26 +0100
Subject: New Book on Numerical Toolbox for Verified Computing

The book
NUMERICAL TOOLBOX FOR VERIFIED COMPUTING I - Basic Numerical Problems
(Springer Series in Computational Mathematics. Eds.: R. L. Graham, J. Stoer,
R. Varga. Vol. 21)
by R. Hammer, M. Hocks, U. Kulisch, D. Ratz
is now available. It is published by Springer-Verlag, Heidelberg, New York
and has 360 pages.

Summary:
This book presents an extensive set of tools for solving basic numerical
problems with verification of the results using the scientific computer
language PASCAL-XSC. It contains implementations of algorithms and many
examples and exercises (available via FTP). Some topics covered are usually
not found in standard numerical analysis texts. It is written for engineers,
mathematicians and scientists working in scientific computing.

Table of Contents:
Introduction * Preliminaries - The Features of PASCAL-XSC - Mathematical
Preliminaries * One-Dimensional Problems - Evaluation of Polynomials -
Automatic Differentiation - Nonlinear Equations in One Variable -
Global Optimization - Evaluation of Arithmetic Expressions - Zeros of
Complex Polynomials * Multi-Dimensional Problems - Linear Systems of
Equations - Linear Optimization - Automatic Differentiation for Gradients,
Jacobians, and Hessians - Nonlinear Systems of Equations -
Global Optimization * Utility Modules


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

From: Marc Moonen <Marc.Moonen@esat.kuleuven.ac.be>
Date: Tue, 18 Jan 94 11:30:56 +0100
Subject: SVD and Signal Processing Workshop in Belgium

3rd International Workshop on
SVD and Signal Processing

August 22--25, 1994
Leuven, Belgium

This Workshop on Singular Value Decomposition and Signal Processing
is a continuation of two previous workshops of the same name which were
held in Les Houches, France (1987), and Kingston, Rhode Island,
U.S.A. (1990).

Papers are solicited for technical sessions on SVD algorithms,
architectures and applications.
Submission of summary: March 15, 1994

Conference Committee : B. De Moor, M. Moonen, E. Deprettere, G.H. Golub,
S. Hammarling, F.T. Luk, P. Van Dooren

For further information contact :
Mr. L. De Lathauwer,
ESAT-K.U.Leuven, K. Mercierlaan 94, 3001 Heverlee, Belgium
lieven.delathauwer@esat.kuleuven.ac.be


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

From: Marc Moonen <Marc.Moonen@esat.kuleuven.ac.be>
Date: Tue, 18 Jan 94 11:30:56 +0100
Subject: Algorithms and Parallel VLSI Workshop in Belgium

3rd International Workshop on
Algorithms and Parallel VLSI Architectures

August 29--31, 1994
Leuven, Belgium

Preliminary Announcement and Call for Papers

This workshop is a continuation of two previous workshops of the same name
which were held in Pont-a-Mousson, France (1990), and Bonas, France (1991),
although the scope has been partly shifted and expanded.

Papers are solicited for technical sessions on parallel algorithms,
parallel architectures and parallel compilation.

Submission of summary: March 15, 1994

Conference Committee : M. Moonen, F. Catthoor, E. Deprettere, P. Dewilde,
P. Quinton, Y. Robert, J. Vandewalle

For further information contact :
Mr. F. Vanpoucke
ESAT-K.U.Leuven, K. Mercierlaan 94, 3001 Heverlee, Belgium
filiep.vanpoucke@esat.kuleuven.ac.be


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

From: Issman Emmanuel <issman@alpha1.vki.ac.be>
Date: Tue, 18 Jan 1994 16:40:36 +0100
Subject: Computational Fluid Dynamics Lectures in Belgium

25th ANNUAL COMPUTATIONAL FLUID DYNAMICS LECTURE SERIES
VON KARMAN INSTITUTE FOR FLUID DYNAMICS (BELGIUM)
21-25 MARCH, 1994

The aim of this annual lecture series is to provide in-depth presentations
on well-established methods and on recent advances in the field of numerical
flow simulation. The program is at a specialist level complementary to the
Introduction to Computational Fluid Dynamics (Lecture Series No. 2).

This year the course will recognize the 25th anniversary of the VKI lecture
series in CFD by focusing on the key algorithmic achievements of the past 25
years which constitute the basis of modern CFD for aerodynamic flows today :
shock capturing TVD high resolution schemes for compressible flows, multigrid
acceleration techniques, advances in linear algebra (Generalized Minimum
Residual and Related Methods), unstructured grid methods and solution
adaptivity. Key scientists at the origin of some of the most successful methods
have been asked to review the lessons learned form the past, in order to
determine the important area's where future developments will take place.

With respect to high resolution shock capturing space discretization, both
finite volume methods, cell vertex residual distribution schemes and SUPG
finite element discretizations will be reviewed, emphasizing the different
nonlinear mechanisms in the schemes which allow for monotonic discontinuity
capturing. For application to systems of hyperbolic equations, both the
classical Riemann solver flux difference splitting approach and the more recent
multidimensional flux balance splitting methods will be discussed.

The second subject will be on fast iterative solvers for efficient steady
state computation. Both Krylov subspace methods and multigrid methods will
receive considerable attention, with application to flow problems on structured
as well as unstructured grids.

In addition to these two major subjects, attention will be given to a new
development of non-dissipative numerical schemes suitable for aeroacoustics.

Lectures will be given in English and printed notes will be distributed
to participants at registration on the first day.

The speakers include:

N. Kroll, DLR, Germany.
T.J. Barth, NASA Ames Research Center, California, USA.
P.L. Roe, U. Michigan, Ann Arbor, USA
R. Radespiel, DLR, Germany.
Y. Saad, U. of Minnesota, Minnesota, USA.
G. Degrez and Mr. E. Issman, von Karman Institute.
H. Deconinck, von Karman Institute.
C. Rossow, DLR, Germany.
H. Paillere and Mr. J-C. Carette, von Karman Institute.

The Director of this lecture series is Prof. H. Deconinck of the von Karman
Institute (email: deconinck@vki.ac.be).

For registration contact :

Lecture Series Secretary
von Karman Institute
Waterloose steenweg - Chaussee de Waterloo, 72
B-1640 Sint-Genesius-Rode Rhode-Saint-Genese (Belgium)
FAX 32 2 3582885
TEL 32 2 3581901

This Lecture Series is supported by the COMETT programme trough
the association CSAMI.


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

From: Karen Hahn <khahn@cs.rutgers.edu>
Date: Tue, 18 Jan 94 13:48:49 EST
Subject: IMACS Turbulence Workshop

IMACS WORKSHOP ON
(theoretical and computational aspects of)
TURBULENCE
Rutgers University, New Brunswick, NJ, USA
February 10-11, 1994

PROGRAM:

R. Agarwal (McDonnell Douglas Aerospace) USA
"Turbulence Modeling in Aerospace CFD Applications - Current State of the Art"

G. Berkooz (Cornell Univ.) USA
"Dynamics and control of coherent structures in the wall layer of the
turbulent boundary layer"

M. Deville ((EPFL) Switzerland
"Finite Element Computation of Two-Equation Turbulence Models"

D. Fishelov (The Hebrew Univ.) Israel
"Three-dimensional Computations of Turbulent Flow Over Bodies with
Non-Cartesian Geometry"

Y. Kimura (Univ. of Colorado) USA
"Turbulence Simulations in Complex Time"

R. Krasny (Univ. of Michigan) USA
"Vortex Sheet Computations"

J. Lee (Wright-Patterson AFB) USA
"Chaos and Direct Numerical Simulation in Turbulence"

A. Majda (Princeton Univ.) USA, (with E. Tabak, Princeton Univ.)
"Singular Front Formation in a Model for Quasi-Geostrophic Flow"

J. Morrison (Nasa Langley Research Center) USA
"Three-Dimensional Solution Algorithm for Arbitrary Configurations with
Two-Equation and Reynolds Stress Turbulence Models"

R. Pelz (Rutgers Univ.) USA
no title yet

V. Russkikh (Inst. of Atomic Energetics) Russia
"Estimate of Laminar flow's stability"

B. Turkington (Univ. of Massachusetts) USA
"Statistical Equilibrium Computations of Coherent Structures"

E. Tziperman (Geophysical Fluid Dynamics Lab., Princeton Univ.) USA
(Weizman Institute), Israel
"Multilevel Turbulence Simulations"

E. Weinan (Inst. for Advanced Study-Princeton) USA
"Boussinesq Turbulence"

N. Zabusky (Rutgers Univ.) USA
"Collapse, intensification and reconnection in vortex dominated flows at high
Reynolds number"

M. Zelman (Rutgers Univ.) USA
"Nonlinear Mechanisms of Boundary Layer Turbulisation"

Dr. N. Zabusky will give a demonstration at the Vizlab on Friday, Feb. 11.


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

From: Sven Hammarling <sven@num-alg-grp.co.uk>
Date: Wed, 19 Jan 94 17:04:19 GMT
Subject: NAG Users Association Conference 1994

NAG Users Association Conference 1994 - Invitation and Programme

University of Oxford, Wednesday 13 - Friday 15 April 1994
Featuring a Special Session on Fortran 90 and an Open Day at NAG Ltd

What is NAGUA?

The NAG Users Association is a self-financing non-profitmaking body which
exists to promote communications between the Numerical Algorithms Group (NAG)
and users of the services and products NAG develops, maintains and markets. It
provides information to NAG on the requirements of users, who are in turn kept
informed of developments in services.

NAGUA Conferences consist of both business and technical sessions providing an
opportunity for users to meet each other and members of NAG, to hear about
others' experiences and to find out about new products and services.

Attendance at the conference is open to any user of NAG's products and
services.

The thirteenth conference of the NAG Users Association will be held in the
historic City of Oxford.

Speakers include:

Brian Ford, Director, NAG Limited
Ken Brodlie, University of Leeds
Victoria Pennington, University of Leeds
Marco Vaccari, ENEA
Sven Hammarling, NAG Limited
John Reid, Rutherford Appleton Laboratory
John Steel, London Parallel Applications Centre & QMW Computing
Christian de Polignac, Institut Laue Langevin
Yann-Herve De Roeck, IFREMER
Jerzy Wasniewski, Danish Computer Centre for Research and Education
Herve Butault, INRIA
Iain Fleming and Richard Hann, NAG Limited
Jeremy Du Croz, NAG Limited
Marco Codutti, Brussels Free University
Benno Sueselbeck, Westfaelische Wilhelms-Universitaet Muenster
Hagen Scherb, GSF/Medis
Norbert Koeckler, Universitaet Gesamthochschule Paderborn

For an invitation and booking form contact

NAG Users Association
PO Box 426
Oxford
OX2 8SD
United Kingdom

Telephone +44(0)865 311102
Fax +44(0)865 310139
email nagua@nag.co.uk


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

From: Keyes David E <keyes@icase.edu>
Date: Fri, 21 Jan 1994 16:56:12 -0500
Subject: NASA Workshop on Parallel Numerical Algorithms

ICASE/NASA Langley Workshop on Parallel Numerical Algorithms, 23-25 May 94

Purpose:
To acquaint developers of high-performance applications codes based on
partial differential equations or integral equations with the state of
the art in core areas of algorithms and the parallel computational
environment, and to engage representative experts from the algorithms
community in discussions about their applicability.

Rationale:
A principal parallel challenge in large computational fluid dynamics,
computational structural mechanics, and computational electricity and
magnetism problems is the implicit solution process. Hence, these
applications areas meet in a common algorithmic core. However, results
on model problems are often not transferable. The workshop will promote
more useful development and more rapid integration of parallel
algorithms in high-performance NASA-related applications.

TECHNICAL PROGRAM
Keynote Address:
Algorithmic Requirements in Aerospace Computations, Tom Zang, NASA Langley

Linear System Solvers:
dense methods -- James Demmel, Univ. of California-Berkeley
sparse direct methods -- Mike Heath, Univ. of Illinois-Urbana
sparse iterative methods -- Henk Van der Vorst, Univ. of Utrecht
eigensolvers -- Dan Sorensen, Rice University

Preconditioning Methods:
approximate and incomplete factorizations -- Tony Chan, UCLA
multilevel methods -- Steve McCormick, Univ. of Colorado-Boulder
domain decomposition methods -- Barry Smith, UCLA

Fast Transforms, Fast Summation, and Fast Operator Application:
FFTs and generalized transforms -- Richard Pelz, Rutgers University
fast multipole methods -- Jacob White, MIT
multiresolution application of operators -- Ami Harten, Tel-Aviv University

Tools and Environments:
decomposition and mapping methods -- Alex Pothen, Old Dominion Univ./ICASE
compiler and run-time environments -- Dennis Gannon, UIPUI

Performance:
performance metrics and benchmarking -- David Bailey, NASA Ames
performance ``debugging'' -- Bill Gropp, Argonne National Laboratory

Venue:
Hampton Radisson, Hampton, Virginia

For further workshop and hotel information:
Contact Emily Todd, Conference Coordinator, ICASE (emily@icase.edu,
804-864-2174)


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

From: William Behrman <behrman@sccm.Stanford.EDU>
Date: Fri, 21 Jan 94 21:50:47 -0800
Subject: High-Precision Unconstrained Minimization

Can you recommend a means to solve an unconstrained minimization
problem to high precision?

For our application, in order to achieve adequate precision in the
final solution we need the calculation to be performed in at least
32 decimal digits of precision.

We have double precision Fortran routines for the objective
function, gradient, and Hessian. We have access to several brands
of workstation: Sun, DEC, and SGI. We also have access to
Mathematica and Maple.

Any suggestions would be greatly appreciated.

Thank you very much,
Sincerely,

Bill Behrman
Scientific Computing
Stanford


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

From: Francois Cellier <CELLIER@cadmus.ece.arizona.edu>
Date: Fri, 21 Jan 1994 19:57:08 -0700 (MST)
Subject: Conference on Bond Graph Modeling and Simulation

ANNOUNCEMENT AND CALL FOR PAPERS
The Society for Computer Simulation International (SCS) presents:

1995 INTERNATIONAL CONFERENCE ON BOND GRAPH MODELING AND SIMULATION
January 15-18, 1995
Tropicana, Las Vegas, Nevada

Part of the 1995 SCS Western Multiconference on Computer Simulation


The 1995 International Conference on Bond Graph Modeling and Simulation
brings together research paper presentations, panel sessions, tutorials,
workshops, seminars, industrial applications, and software demonstrations that
use Bond Graph modeling methods. There will be also featured plenary speakers
and some special invited presentations.

This conference will focus on Bond Graph modeling techniques for dynamic
systems. Theoretical principles for electrical, mechanical, hydraulic,
pneumatic, and control system applications will be presented. Leading
industrial users of the method in automotive, aircraft, fluid power, kinematics,
multibody systems, and social and biological systems have been invited.
Research papers are welcome in the following categories of presentation
sessions.

. Tutorials . Mechanical Systems and Robotics
. Panel Discussions . Electrical and Power Systems
. Software and Tools . Control Systems
. Bond Graph Theory . Thermal and Chemical Systems
. Advanced Bond Graph Methodology . Biomechanics and Prosthetics
. Bond Graphs and Block Diagrams . Ecological Systems
. Computer Graphics and B.G. Modeling . Social and Economic Systems

DEADLINES:
. April 30, 1994 Short abstracts due (200 words/2 copies).
. May 15, 1994 Invitations to submit full papers will be extended.
. June 30, 1994 Full papers due (five copies).
. August 15, 1994 Notification of paper acceptance to authors.
. September 30, 1994 Camera Ready Copies due.

Manuscripts can be submitted either to the General Chairman or the
Program Chairman. Please add a cover letter stating your name, affiliation,
mail address, telephone number, FAX number, and EMail address. Electronic
submission of the draft manuscripts in the form of a compressed and unencoded
postscript file is encouraged. Abstracts will be reviewed by the two Chairmen
for suitability to the conference. Each full paper will be reviewed by at least
two members of the International Program Committee.

General Chairman Program Chairman

Francois E. Cellier,Ph.D. Jose J. Granda, Ph.D.
Associate Professor Professor
Dept. of Electr. and Computer Engr. Department of Mechanical Engineering
University of Arizona California State University, Sacramento
Tucson, Ariz. 85721 Sacramento, Calif. 95819
U.S.A. U.S.A.

Phone: +1(602)621-6192 Phone: +1(916)278-5711
FAX: +1(602)621-8076 FAX: +1(916)278-5949
EMail: Cellier@ECE.Arizona.Edu EMail: GrandaJJ@ECS.CSUS.Edu


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

From: Jerzy Wasniewski <unijw@fysik.uni-c.dk>
Date: Sun, 23 Jan 1994 10:51:23 +0100 (MET)
Subject: Workshop on Parallel Scientific Computing in Denmark

To the Danish Computing Community:

Jack Dongarra is coming to Denmark. He will be a visiting
professor at UNI*C for two months, June and July, 1994.
Professor Jack Dongarra is famous for his work on supercomputing
and numerical analysis. He introduced the concept of BLAS (Basic
Linear Algebra Subprograms) for the Vector and Parallel Computers
and he is the co-author of several numerical books and numerical
libraries, like LINPACK and LAPACK. He is also deeply involved with
PVM (Parallel Virtual Machine) and development of scalable
libraries.
First of all we are together organising the Workshop on Parallel
Scietific Computing and Tutorial on heterogeneous network computing
(PVM and Hence). To get information about the Workshop and Tutorial,
subscribe to the mailing list. Just submit a message to
listserv@vm.uni-c.dk
containing the command
sub para94-l your-name
To register for the Workshop and Tutorial and to obtain more
information about the Workshop and Tutorial, send a message to
workshop@uni-c.dk
or send a fax: (+45) 45 93 02 20 addressing to the "Workshop on
Parallel Scientific Computing" or to Jerzy Wasniewski at UNI*C.
If another School or Department would like to organise either a
lecture or a seminar with Jack Dongarra, please contact Jerzy
Wasniewski at UNI*C.
Jerzy Wasniewski
UNI*C
DTH, Bldg 304
2800 Lyngby, Denmark
Tel: (+45) 42 88 39 99 + 2426
Fax: (+45) 45 93 02 20
E-mail: jerzy.wasniewski@uni-c.dk


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

From: Herman.te.Riele@cwi.nl
Date: Mon, 24 Jan 1994 17:46:53 +0100
Subject: Symposium on Parallel Computing in the Netherlands

CWI - RUU SYMPOSIA "MASSIVELY PARALLEL COMPUTING AND APPLICATIONS"

In 1993-1994, the Centre for Mathematics and Computer Science Amsterdam (CWI)
and the University of Utrecht (RUU) are organising a series of symposia
on massively parallel computing and applications.

This is to announce the fourth meeting which centres around the theme:

PARALLEL NUMERICAL ALGORITHMS AND SOFTWARE

Date: Friday, February 4, 1994
Location: CWI, Kruislaan 413, Amsterdam, The Netherlands
Room: Z011
Speakers:
Eric ten Cate and Edwin Vollebregt (Delft Univ. of Technology)
Maya Neytcheva (Catholic University Nijmegen)
Martin van Gijzen (Utrecht University)
Johan De Keyser (Catholic University Leuven, Belgium)

For further information, e.g., about how to reach CWI, contact H.J.J. te Riele
(CWI, tel. +31 20 5924106)
If you wish to receive a LaTeX-file of the abstracts of the lectures,
please send a request to herman@cwi.nl .

Herman te Riele (CWI)
Henk van der Vorst (RUU en CWI)


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

From: Kearfott Ralph B <rbk5287@usl.edu>
Date: Tue, 18 Jan 1994 14:01:59 -0600
Subject: Position at University of Southwestern Louisiana

UNIVERSITY OF SOUTHWESTERN LOUISIANA
Department of Mathematics

Applications are invited for a tenure-track assistant or
associate professorship appointment in numerical analysis or
computational mathematics. Duties involve teaching, research
and other related activities associated with an academic
position. Junior applicants must show strong research promise;
senior applicants should have demonstrated leadership in
research and willingness to supervise masters or doctoral
students. Success in attracting outside funding is preferred.
Applicants must have a Ph.D. degree by the time of appointment,
and are expected to excel in teaching. The appointment will
begin with the academic year 1994-1995. Salary is negotiable,
depending on qualifications. A current resume', three letters
of recommendation and any other supporting materials should be
sent to

C. Y. Chan, Head
Department of Mathematics
University of Southwestern Louisiana
Lafayette, LA 70504-1010

The screening of applicants will begin on March 15, 1994.
EEO/AA.


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

From: J. M. Aitchison <aitchison@rmcs.cranfield.ac.uk>
Date: Fri, 21 JAN 94 15:00:31 +0100 (BST)
Subject: Position at Royal Military College of Science

Royal Military College of Science Cranfield University
Shrivenham.

Applied Mathematics & Operational Research Group
LECTURESHIP

A vacancy exits for a lecturer to join the Applied Mathematics and
Operational Research Group at RMCS, Shrivenham, which is a faculty of
Cranfield University but situated in south west Oxfordshire.

The research interests of the group include mathematical modelling,
numerical methods, scientific software and simulation. Research is funded
by contracts with both defence and non-defence industries and by SERC
awards.

The teaching committments of the group include the modular MSc programs in
Scientific Applications Software, Numerical Methods & Software Systems,
Mathematical Modelling and in Military Operational Research. These courses
are taken by both full and part time students and the first two have
accepted by the SERC as suitable for the award of Advanced Course
Studentships. Other teaching committments include a wide range of service
courses to engineering students.

Applications are invited from candidates with compatible research and
teaching interests for a post of lecturer to start on 1st April 1994
or as soon as possible thereafter.

Further particulars and application forms may be obtained from the
Personnel Office, RMCS, Shrivenham Wilts SN6 8LA. Tel 0793 785758 quoting
reference SDM5.

Informal enquiries may be made to Dr J M Aitchison, tel : 0793 785276,
email : aitchison@rmcs.cranfield.ac.uk or na.aitchison@na-net.ornl.gov.


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

End of NA Digest

**************************
-------