NA Digest Thursday, June 18, 1992 Volume 92 : Issue 24

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: Kamalini Ramdas Jain <RAMDAS61@WILMA.WHARTON.UPENN.EDU>
Date: Tue, 9 Jun 92 16:27 EDT
Subject: Where are NATO Summer Conferences?

I am trying to find out the topics and locations for any NATO sponsored
student summer camps this summer. If you have information about these camps,
I would appreciate it greatly if you could write to me at:
ramdas61@wharton.upenn.edu
Thank you,
Kamalini Ramdas Jain
RAMDAS61@WILMA.WHARTON.UPENN.EDU


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

From: John Prentice <john@aquarius.unm.edu>
Date: Tue, 9 Jun 92 15:03:57 MDT
Subject: Looking for Access to an Ada Compiler

One of the projects I had hoped to undertake this summer was to learn
Ada programming. However, somewhat to my amazement, I have been unable
to find an Ada compiler on any of the computers I have access to, which
includes two DoE labs and two DoD labs (so much for Ada being the
workhorse of the DoD). There are also no Ada compilers available at
this university, at least on any systems I have been able to find. So,
I would like to pulse the community to see if anyone has a computer with
Ada mounted on it and who would be willing to give me a guest account for
a few months. I realize that na.net is a odd place for such a request,
but it seemed like a reasonable place to start.

John

John K. Prentice
Quetzal Computational Associates
3200 Carlisle N.E., Albuquerque, NM 87110-1664; 505-889-4543
john@aquarius.unm.edu -or- jkprent@cs.sandia.gov


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

From: Nick Higham <higham@vortex.mathematics.manchester.ac.uk>
Date: Fri, 12 Jun 92 12:24:52 BST
Subject: New Book on Parallel Numerical Algorithms

A book on parallel numerical algorithms that has just been
published may be of interest to NA-Digest readers. It is by Len
Freeman (Department of Mathematics, University of Manchester) and
Chris Phillips (Computing Laboratory, University of Newcastle).
In the preface the authors say

``The material covered by this book will be of relevance to final
year undergraduate and postgraduate courses on parallel numerical
algorithms given in departments of computer science, mathematics
and engineering''.

The details are, in the form of a BibTeX entry:

@book{frph92,
author = "T. L. Freeman and C. Phillips",
title = "Parallel Numerical Algorithms",
publisher = "Prentice Hall",
address = "Hemel Hempstead, Hampshire",
year = 1992,
mynote = "ISBN 0-13-651597-5.
Prentice Hall International Series in Computer Science,
C. A. R. Hoare, Series Editor.
Softback: \pounds 19.95 (UK), USA price to be announced."
}

--Nick Higham


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

From: Murli Gupta <MMG@gwuvm.gwu.edu>
Date: Wed, 10 Jun 92 14:43:24 EDT
Subject: GAO Report-"Patriot Missile Defence"

A couple of weeks ago, David Keaton gave a brief description
of the GAO report "Patriot Missile Defence: Software Problem
Led to System Failure at Dhahran, Saudi Arabia".

I acquired a copy of the report and found it to be very interesting
from a numerical analysis perspective; many others seem to be
interested in further details contained in the report.

The report number is GAO/IMTEC-92-26 and is available
from GAO at (202) 275-6241.

Further details of the GAO Report

The Patriot missile defence battery uses a 24 bit arithmatic
which causes the representation of real time and velocities to
incur roundoff errors; these errors became substantial when the
patriot battery ran for 8 or more consecutive hours.

As part of the search and targeting procedure, the Patriot
radar system computes a "Range Gate" that is used to track and
attack the target. As the calculations of real time and velocities
incur roundoff errors, the range gate shifts by substantial margins,
especially after 8 or more hours of continuous run.

The following data on the effect of extended run time on patriot
operatios from Appendix II of the report would be of interest to
numerical analysts everywhere.

Hours Real time Calculated Time Inaccuracy Approximate shift
(seconds) (seconds) (seconds) in range gate
(meters)
0 0 0 0 0
1 3600 3599.9966 .0034 7
8 28800 28799.9725 .0275 55
20a 72000 71999.9313 .0687 137
48 172800 172799.8352 .1648 330
72 259200 259199.7528 .2472 494
100b 360000 359999.6667 .3333* 687

a: continuous operation exceeding 20 hours-target outside range gate
b: Alpha battery [at Dhahran] ran continuously for about 100 hours

* corrected value [GAO report lists .3433]

On February 21, 1991 the Patriot Project Office sent a message
to all patriot sites stating that very long run times "could cause
a shift in the range gate, resulting in the target being offset".
However, the message did not specify "what constitutes very long run
times. According to the Army officials, they presumed that the users
would not run the batteries for such extended periods of time that
the Patriot would fail to track targets. Therefore, they did not
think that more detailed guidance was required".

The air fields and seaports of Dhahran were protected by six Patriot
batteries. Alpha battery was to protect the Dhahran air base.
On February 25, 1991, Alpha battery had been in operation for over 100
consecutive hours. That's the day an incoming Scud struck an Army
barracks and killed 28 American soldiers.

On February 26, the next day, the modified software, which
compensated for the inaccurated time calculation, arrived
in Dhahran.


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

From: Eberhard Kranich <puett@wmwap1.math.uni-wuppertal.de>
Date: Fri, 12 Jun 92 14:47:54 +0200
Subject: Bibliography on Interior Point Methods

In these days I'm rewriting and updating my bibliography on interior point
methods for mathematical programming which can be accessed via NETLIB :

send index from bib

For the updating process I need some help : Please check the bibliography, only
on your papers and talks, and if you miss reports or if citations are incorrect
send the corrections via e-mail to me. According to the talks please send a
crossreference list "talk -> report", because I believe that there exists an
underlying report to each talk, differing only in title.

In case you can help me I'm very much obliged to you.

-- Dr. Eberhard Kranich
FB Mathematics 7
Uni Wuppertal
Gauss-str. 20
Room G14.13
D-W-5600 Wuppertal 1
Germany
e-mail : puett@wmwap1.math.uni-wuppertal.de or na.kranich


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

From: W. E. Schiesser <WES1@SSCVX1.SSC.GOV>
Date: Wed, 17 Jun 1992 12:12:56 -0500 (CDT)
Subject: Programs from Schiesser's Book Available

The programs in the book:

"The Numerical Method of Lines Integration of Partial
Differential Equations", Academic Press, San Diego,
CA, ISBN: 0-12-624130-9

are available on a 3.5 inch, DOS-formatted, 1.44 mb diskette
(Fortran source code). Send requests to:

W. E. Schiesser

Until August 15, 1992:

SSC Laboratory, MS 4003
2550 Beckleymeade Avenue
Dallas, TX 75237 USA

(214) 708-4804 (fax)
wes1@sscvx1.ssc.gov

After August 15, 1992:

Iacocca Hall, D307
Lehigh University
111 Research Drive
Bethlehem, PA 18015 USA

(215) 758-5297 (fax)
wes1@lehigh.edu


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

From: Haiyan Liu <haliu@slate.Mines.Colorado.EDU>
Date: Tue, 9 Jun 92 9:26:15 MDT
Subject: Multigrid Short Course at University of Colorado at Denver

ANNOUNCEMENT

The University of Colorado at Denver is sponsoring the Third
Multigrid Short Course on campus from March 31 to April 4, 1993.
The basic components of the course include
. fundamental principles and algorithms
. practical applications
. software demonstration and distribution

Suggestions about contents, arrangements, and other aspects of the
course are encouraged. For example, people interested in having other
topics included, or those interested in only attending part of the course
(possibly for a reduced price), please indicate so. Such suggestions and
comments should be sent to Chaoqun Liu either by e-mail at

cliu@copper.colorado.edu

or by regular mail at

Chaoqun Liu
Computational Mathematics Group
Campus Box 170
University of Colorado at Denver
P.O. Box 173364
Denver, CO 80217-3364

Following is the tentative program:

MULTIGRID SHORT COURSE
March 31 - April 4, 1993

Chairman : Chaoqun Liu

Lecturers : Achi Brandt
Williams Briggs
Chaoqun Liu
Steve McCormick
John Ruge

Monday - Tuesday
Elementary Multigrid Tutorial, Basic Approach, Elliptic Equations,
Adaptive Methods

Wednesday - Thursday
Computational Fluid Dynamics, Incompressible Navier-Stokes Equations,
Compressible Flows, Grid Generation, Two-Phase Porous Media Flows,
Flow Transition, Combustion

Friday
Multiscale Methods in Particle Simulation

Monday-Friday
Software Demonstration and Distribution

The registration fee will be $475 ($175 for students), which will include
refreshments, computer laboratory access, and materials. Also, a block of
reduced-rate rooms will be arranged in the neighboring Executive Tower Inn.

Please Contact Chaoqun Liu (phone (303)556-4811, fax (303)556-3377,
e-mail cliu@copper.denver.colorado.edu).


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

From: Hans-Hermann Frese <frese@sc.ZIB-Berlin.DE>
Date: Thu, 11 Jun 92 11:41:26 +0200
Subject: NAGUA Workshop on Fortran 90

NAGUA Workshop on
Fortran 90
The future of Fortran here today
Monday 22 June 1992
Hannover, Germany

The Workshop is to be held on Monday 22 June 1992 at the
Interconti Hotel in Hannover.

"Fortran 90 through the eyes of a Fortran 77 Programmer"
Wilhelm Gehrke, RRZN, University of Hannover

"How NAG and its users will benefit from Fortran 90"
John Reid, Rutherford Appleton Laboratory

"Fortran 90 Compiler and Tools"
Malcolm Cohen, NAG Limited

"Designing a Fortran 90 Numerical Library"
Jeremy Du Croz, NAG Limited

For more information, contact:
NAGUA Administrator, PO Box 426, OXFORD, United Kingdom OX2 8SD
Fax: +44 865 310139.
------------------------------

From: Paul Van Dooren <dooren@imafs.ima.umn.edu>
Date: Tue, 9 Jun 92 16:06:52 CDT
Subject: Mathematical Theory of Networks and Systems

First Announcement
and
Call for Papers

INTERNATIONAL SYMPOSIUM ON THE
MATHEMATICAL THEORY OF NETWORKS AND SYSTEMS
1993

August 2 - 6, 1993, Regensburg, Germany

The International Symposium on the Mathematical Theory of Networks and Systems
(MTNS) is held biannually since 1973. The symposium offers theoretical engineers, researchers and mathematicians in systems, control and circuit theory a
platform to discuss recent developments, to exchange new ideas and to analyse
trends for future research. It also provides them with an opportunity to
establish and maintain contacts with colleagues. Topics of interest include
mathematics for control, multivariable control, systems modelling, optimal
control, filtering, robotics, VLSI, neural networks and control theoretic
aspects of artificial intelligence.

The scientific program of the symposium consists of plenary lectures by invited
speakers, parallel lectures by invited speakers, mini courses, invited sessions
and contributed lectures.

Prospective authors should submit three copies of an unpublished paper in the
English language by November 1, 1992, The length of the paper should not exceed
10 pages. The cover page of the paper should contain: title, author^s name,
affiliation and address, fax-number, e-mail address, telephone number, an
abstract and keywords. In the case of joint authorship, the first name
mentioned will be used for all correspondence, unless otherwise requested.

Proposals for invited sessions should also be submitted by November 1, 1992.
An invited session normally comprises 3 - 5 lectures of 25 minutes duration,
but other arrangements are possible. The proposal should contain the title
of each lecture as well as the name, affiliation, address, fax-number,
e-mail address and telephone number of each proposed speaker and of the
organizer(s). In addition, a one page abstract of each lecture is required.

Papers and proposals for invited sessions should be sent to MTNS 93, U. Helmke,
Department of Mathematics, Regensburg University, 8400 Regensburg, Germany,
Fax-No.: +49 941 943-2305, e-mail: Helmke@vax1.rz.uni-regensburg.dbp.de.
Requests for information should be sent to one of the chairmen (U. Helmke or
R. Mennicken).

Dates:

November 1, 1992 Deadline for submission of papers and proposals for special
sessions.
March 1, 1993 Notification of acceptance of contributed papers and special
sessions.
April 1, 1993 One page abstract is due.

August 2 - 6, 1993 Symposium

Chairmen:

Uwe Helmke
Reinhard Mennicken
Department of Mathematics, Regensburg University, 8400 Regensburg, Germany.

Fax: +49 941 943-2305
Telex: 65658 unire d
E-Mail: Helmke@vax1.rz.uni-regensburg.dbp.de
Mennicken@vax1.rz.uni-regensburg.dbp.de


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

From: HJJ te Riele <Herman.te.Riele@cwi.nl>
Date: Thu, 11 Jun 1992 10:45:06 GMT
Subject: ERCIM Course on Large-Scale Scientific Parallel Computing

ERCIM advanced 4-day course on
Large-Scale Scientific Parallel Computing

October 20-23, 1992, RAL, Abingdon, UK

ERCIM

ERCIM stands for European Research Consortium for Informatics and Mathematics
and is comprised of INRIA (France), GMD (Germany), RAL (UK), CWI (The
Netherlands), INESC (Portugal), IEI/CNR (Italy) and FORTH (GREECE):
all nationally funded research institutes in the field of informatics and
applied mathematics.
One of ERCIM's aims is to further the applications of informatics and
mathematics through an active policy of technology transfer.

Purpose of the course

The course is designed to teach students, researchers, scientists, and
engineers how to exploit parallel and vector computers, both present and
future, in order to produce appropriate and efficient codes on these
machines. Experience has shown that gaining know-how in this field
can be a time-consuming activity. Participants can benefit from insight,
knowledge and experience of the course teachers in this area, and gain
a comprehensive introduction to the field.

Contents of the course

The initial part of the course deals with basic principles of
vector/parallel architectures and their software and programming
methodologies. Some emphasis is placed on approaches and techniques
that facilitate the achievement of parallel software portability.
The treatment of linear algebra on vector and parallel machines will be
a central theme of the third day, including direct and iterative solution
methods, sparse matrix techniques, and considerations of shared versus
distributed memory algorithms. The final section of the course will cover
finite element methods, adaptive and multi- grid stategies, and
applications in the fields of computational physics, computational fluid
dynamics, and the factorisation of large numbers.

Location

The course is fully residential in Cosener's House (Abingdon, UK; near the
Rutherford Appleton Laboratory and close to Oxford) on an attractive location
beside the River Thames, allowing good opportunity for participants to discuss
topics with the lecturers.


Course secretariat (for more information and subscription)

Mr. Frans Snijders
CWI
Kruislaan 413; P.O. Box 4079
1009 AB Amsterdam
The Netherlands
tel. +31-20-5924171
fax +31-20-5924199
email: franss@cwi.nl

Course Directors

HJJ te Riele, CWI, email: herman@cwi.nl
CP Wadsworth, RAL, email: cpw@inf.rl.ac.uk


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

From: Mike Osborne <mike@thrain.anu.edu.au>
Date: Thu, 18 Jun 92 11:41:20 +1000
Subject: Positions at Australian National University

Positions available at The Australian National University

School of Mathematical Sciences
Centre for Mathematics and its Applications
Program in Advanced Computation and Modelling

Applications are invited for two positions in the research program in
Advanced Computation and Modelling. The successful candidates will be
expected to support and enhance current research activity in numerical
analysis, the computational solution of partial differential and integral
equations, ill-posed problems, optimization, statistical computing, and
symbolic and exact computation. Emphasis is being placed on the development
both of algorithms exploiting advanced computer technology and concepts, and
of data visualisation techniques. Opportunities for industrial collaboration
exist through close collaboration with the CSIRO Division of Mathematics and
Statistics, and links between the University and the Fujitsu Corporation.
Computing facilities include a Fujitsu VP2200 supercomputer, a 16000
processor CM2, and a 32 processor CM5, in addition to medium scale machines
and workstations. The University through its Centre for Information Sciences
Research is an affiliate of the Argonne ACRF and has received a prototype
Fujitsu AP1000 (a 128 processor distributed memory parallel computer) for
joint development work.

Two positions are available: one in the Institute of Advanced Studies
component of the School, the other in the Faculties component supported by an
special ARC grant to the Centre for Mathematics and its Applications. These
appointments will be for periods of one, two or three years and may be
extendible. The initial level of appointment, depending on the candidates
experience and qualifications, will be in the following salary ranges:

Research Fellow; $A41,000-$A48,688 p.a.
Postdoctoral Fellow $A36,285-$A38,950 p.a.

In addition, part-time appointments may be available as well as extended
visiting appointments during leave of absence from a home institution.
Assistance with relocation expenses will also be provided.

Enquiries and requests for further information can be addressed to Professor
M.R.Osborne email mike@thrain.anu.edu.au or na.osborne@na-net.ornl.gov. Fax
(06) 247 2347.

Applications: These should be sumitted in duplicate to The Secretary, The
Australian National University, GPO Box 4, Canberra, ACT 2601, Australia,
with whom applications close on 24 July, 1992. They should include a
curriculum vitae, list of publications, and the names of at least three
academic referees. Email addresses and/or fax numbers should be provided for
referees so that appointment procedures can be expedited. The University
reserves the right not to make an appointment or to appoint by invitation at
any time.


THE UNIVERSITY IS AN EQUAL OPPORTUNITY EMPLOYER


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

From: Graham de Vahl Davis <MBEHNIA@ibm.gwdg.de>
Date: Tue, 09 Jun 92 18:59:58 MSZ
Subject: Position at University of New South Wales

UNIVERSITY OF NEW SOUTH WALES
Sydney AUSTRALIA
RESEARCH ASSOCIATE

A new position of Research Associate is available in the School of
Mechanical and Manufacturing Engineering as part of a project to
develop design software for power utility boilers. The R & D
program is being undertaken jointly by UNSW with Sydney University,
Pacific Power Corporation of NSW and Aust. Supercomputing Technology.

The program involves the development of innovative robust computational
algorithms able to predict complex turbulent, high temperature two-
phase flow through power utility boilers and the conversion of the
research code into operational design software to be used by non-
specialist engineers.

An applicant is sought with a PhD, research experience in CFD,
heat transfer and turbulence modelling and a willingness to
contribute to the production of commercially viable software.

Experience in developing and using industrial computer codes is
essential. Experience on vector supercomputers or massively
parallel computers would be an advantage.

The person appointed will work on the boiler heat transfer aspect
of the project, developing a finite-volume computer code tp
predict the complex three-dimensional turbulent heat behaviour
through the superheater and economiser regions of power generation
boilers.

Salary is in the range $ Aust 35,000 - 39,000 (approx.)

Further information from Dr Eddie Leonardi, (+61 2) 697-4252,
fax (+61 2) 663-1222, eddie @ cfd.mech.unsw.oz.au prior to
July 28; or from Professor G de Vahl Davis, (+61 2) 697-4099,
fax (+61 2) 663-1222, gdvd @ cfd.mech.unsw.oz.au after that date
(also by e-mail intermittently prior to then).

Please submit written application, quoting Reference Number 039,
and including business and private telephone numbers, fax and
e-mail numbers, together with the names, addresses (and preferably
fax numbers) of at least two referees, CV and transcript, to:
Recruitment Officer, Staff Office, UNSW, PO Box 1, Kensington,
NSW 2033.


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

End of NA Digest

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