**Today's Topics:**

- Enhancements to the NA-NET Whitepages.
- Magnus Hestenes
- Matrix Exponential Code Wanted, #1
- Matrix Exponential Code Wanted, #2
- FFT in 386 Assembly Language Wanted
- PICL Update
- Position Wanted
- Position at University of Leicester
- SERC Research Studentship at Manchester
- Seminar on Incomplete Decompositions
- Table of Contents: SIAM J. Appl. Math.
- Table of Contents: SIAM Computing

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

From: Jack Dongarra and Bill Rosener <rosener@cs.utk.edu>

Date: Thu, 13 Jun 91 15:16:26 -0400

**Subject: Enhancements to the NA-NET Whitepages.**

The following two major features have been added to the NANET whitepages.

1. KEYWORD SEARCHES. Below are 2 examples.

To find out more information on all people listed in the

NANET whitepages who are interested in "parallel".

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

mail to: na.whois@na-net.ornl.gov

Subject: "blank"

Keyword: parallel

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

This does a string comparison on all fields other than the name fields

searching for the pattern "parallel".

2. MULTIPLE LINE ENTRIES. All fields (except first, last and

middle name) can now be more than one line long. Below is

an example of how Jack Dongarra would change his "research:"

to be more than one line.

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

mail to: na.change-wp@na-net.ornl.gov

Subject:

lastname: dongarra

firstname: jack

research: Numerical Linear Algebra

Parallel Processing

Software Tools

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

A <return> character is inserted at the end of each line. Once a

keyword has been read (e.g. "research:"), subsequent lines will be

added to this field until the next keyword is read or the

end-of-message is encountered.

Several minor enhancements and features have also been added to the

NANET-whitepages. For example, the software used to extract the name

from mail sent to "na.whois@na-net.ornl.gov" has been extended to

accept many different forms. For instance, the name can now appear on

the subject line.

For a complete set of instructions describing the NANET and the new

features to the whitepages, send mail to "na.help@na-net.ornl.gov"

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

From: Tony Chan <chan@math.ucla.edu>

Date: Tue, 4 Jun 91 13:09:36 -0700

**Subject: Magnus Hestenes**

I regret to inform you that Professor Emeritus Magnus Hestenes

of UCLA died on Friday, May 31, 1991. He had had a heart attack

several days earlier and all three of his children were able to

see him before he passed away peacefully at 2:30pm.

Magnus Hestenes was born in 1906 and received his Ph.D. from the

University of Chicago in 1932. In 1947 he accepted a Professorship

at UCLA (at the time he was an Associate Professor at Chicago)

and he taught at UCLA until his retirement in 1973. He served as

Department Chair from 1950-58.

In addition to his essential role in building the UCLA Mathematics

Department, he had an outstanding research career. He was best

known for his work on the problem of Bolza, for his 1951 paper on

quadratic forms in Hilbert space, and for his development of the

conjugate gradient method. He received the Guggenheim and Fulbright

awards, was a Vice President of the American Mathematical Society,

and was an invited speaker at the 1954 International Congress of

Mathematicians in Amsterdam. During his tenure at Chicago and UCLA,

he supervised the thesis research of 34 students.

Tony Chan

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

From: John Rice <rice@surya.UCSD.EDU.

Date: Mon, 10 Jun 91 15:00:31 PDT

**Subject: Matrix Exponential Code Wanted, #1**

I am looking for code to compute exp(Q). Q is the "generator" for a Markov

process and is non-symmetric and of moderate size (less than, say, 30x30).

In particular, an implementation of the Pade approximation discussed in

Golub and Van Loan (Sec 11-3), would be useful.

Any pointers will be appreciated!

John Rice (jrice@ucsd.edu or jrice@ucsd.bitnet)

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

From: A. Scottedward Hodel <hodel@george.msfc.nasa.gov.

Date: Thu, 13 Jun 91 12:27:37 -0500

**Subject: Matrix Exponential Code Wanted, #2**

I am looking for either pre-written code or literature on the matrix

exponential e^{A h} and its integral \int_0^h{e^{A t}} dt. The application

is in the discretization of a large scale (n=150), possibly sparse continuous

time system

d/dt x = A x + B u

y = C x + D u

into a sampled-data discrete time system. At this point the most reliable

methods I am aware of are diagonal Pade' approximations and Schur

decomposition based methods (Moler and Van Loan, 1978). If any progress has

been made on this problem beyond these methods, I'd appreciate any input.

Pre-written implementations would also be welcome!

A. S. Hodel summer address:

200 Broun Hall 3636A Squaw Valley Dr.

Auburn University, AL 36849 Huntsville, AL 35805

scotte@eng.auburn.edu hodel@george.msfc.nasa.gov

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

From: Daniel Tinswin Lin <tinswind@caen.engin.umich.edu>

Date: Tue, 11 Jun 91 10:35:25 -0400

**Subject: FFT in 386 Assembly Language Wanted**

I am looking for an assembly code 80386 of an FFT. I prefer

it to be compatible with Microsoft languages. I am working

in the Electrical Engineering and Computer Sciences

Department at the University of Michigan, Ann Arbor.

Thank you very much.

E-mail Address: tinswind@caen.engin.umich.edu

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

From: Pat Worley <worley@haven.EPM.ORNL.GOV>

Date: Fri, 14 Jun 91 16:53:44 EDT

**Subject: PICL Update**

A new version of PICL (Portable Instrumented Communication Library) is now

available from netlib@ornl.gov. This backward compatible upgrade includes

new features that improve the timestamps generated when tracing, allow

Intel iPSC/2 and iPSC/860 users to access nonblocking send and receive

commands (isend and irecv), and significantly decrease the overhead

of tracing on the iPSC/2.

This upgrade represents part of a larger upgrade that will be available

in a couple of months. It is being made available now because the new

features are extremely useful when using the Intel iPSC/2 and iPSC/860.

We also have an implementation of PICL for the NCUBE 6400 (also known as

the NCUBE 2) that we are making available at this time.

This upgrade has not been tested as thoroughly as the previous version, nor

is there much documentation for the new features. In consequence, the

original version of PICL will continue to be available from netlib.

The source for the new version has the same name as the original

version, but with an "x" prefix to the name. Thus "xpicl.shar" is the

new version of the "picl.shar" file. A short description of the new

features is contained in the file "upgrade.1.91". A list of all available

PICL source can be obtained by sending the following message to

netlib@ornl.gov:

send index from picl

Pat Worley

worley@msr.epm.ornl.gov

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

From: C. Woodham <AP1CAW@primea.sheffield.ac.uk>

Date: Tue, 11 Jun 91 11:04:32

**Subject: Position Wanted**

Postion Wanted

I am currently finishing a PhD on eigenvalue placement for Variable

Structure Control Systems and am looking for a post-doc post,

preferably in the US. I have a BSc (Hons) in Applied Mathematics,

an MSc in Numerical Analysis, and three years industrial experience

working for GEC Avionics on Inertial Navigation Systems. I would

ideally like a post in the Numerical Analysis field, or in the field

of Mathematical Control. If further information is required, my

full cv is available on request.

No reasonable offer refused !!

Kate Woodham

SERC Research Assistant

Department of Applied & Computational Mathematics,

University of Sheffield,

Sheffield S3 7RH

UK

AP1CAW@uk.ac.sheffield.primea

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

From: Will Light <maa007@central1.lancaster.ac.uk>

Date: Wed, 12 Jun 91 10:49:41 +0100

**Subject: Position at University of Leicester**

University of Leicester

Department of Mathematics

The University of Leicester is planning to build up a research group in

numerical analysis within the mathematics department. At present two

appointments have already been made in this area -- Will Light

(approximation theory) and Mark Ainsworth (finite element methods).

There are two more immediate vacancies. One of these is a permanent

position in numerical analysis, and the other is a one year temporary

position. However, a good candidate for the temporary position could

expect to have her/his appointment made permanent after one or two years.

We are looking for good people in any area of numerical

analysis, but since Leicester has some very strong Science and

Engineering departments, we would particularly welcome applications from

people with interests in applications of numerical analysis to scientific

problems. It is most important that the persons appointed to these

positions are able researchers in numerical analysis.

Salary will be on the scale 11,969 pounds to 17,593 pounds, although

this is an interim scale as UK salaries are currently under review.

There is also a possibility that someone with an established research

record could be appointed to a senior position.

Unfortunately, we are working to a tight deadline, and 5th July will be

the nominal closing date. Further details can be obtained from me via

email, snail mail, or FAX.

email: maa007@cent1.lancs.ac.uk or na.light@na-net.ornl.gov

FAX: (UK code) (0)524 - 381707 (until 30th June)

snail mail: Professor W. Light, Mathematics Department, University of

Leicester, University Road, Leicester LE1 7RH, England.

Will Light

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

From: Nick Higham <MBBGSNH%CMS.Manchester-Computing-Centre.AC.UK@CUNYVM.CUNY.EDU>

Date: Fri, 14 Jun 91 12:59:28 BST

**Subject: SERC Research Studentship at Manchester**

The University of Manchester

Department of Mathematics

Further applications are invited for an SERC Earmarked Research

Studentship in Numerical Analysis on the topic

Numerical Stability of Iterative Methods for

Solving Linear Equations

tenable from October 1991 for research leading to a Ph.D.

The Numerical Analysis Group comprises six members of staff with research

interests including matrix computations, optimization, approximation theory,

ordinary and partial differential equations, and parallel computing.

Excellent computing facilities are available in the department and in the

Manchester Computing Centre.

Applicants, with a background in numerical mathematics, should contact as soon

as possible

Dr N.J. Higham

Department of Mathematics

University of Manchester

Manchester M13 9PL

UK

Tel: 061 275 5822

Fax: 061 275 5819

email: mbbgsnh@uk.ac.mcc.cms

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

From: Gabriel Wittum <wittum@iwr1.iwr.uni-heidelberg.de>

Date: Fri, 14 Jun 91 15:26:13

**Subject: Seminar on Incomplete Decompositions**

GAMM-Fachausschuss

Effiziente numerische Verfahren fr partielle Differentialgleichungen

First Announcement

The GAMM Committee "Efficient numerical methods for PDE" in cooperation with

the Christian-Albrechts-Universit Kiel organizes the

Eighth GAMM-Seminar Kiel

on

Incomplete Decompositions (ILU):

Theory, Technique and Applications

Chairmanship: W. Hackbusch (Kiel), G. Wittum (Heidelberg)

Date: January 24 to 26, 1992

Location: Institut fr Informatik und Praktische Mathematik,

Universitaet Kiel (Germany)

Topics: Incomplete Decompositions used as solvers,

Preconditioners or smoothers,

Factorization strategies/ Modification/ Stability/

Robustness/ Ordering/ Comp. aspects (e.g. Parallelization)

Abstracts: Please send abstracts (10-20 lines) of your lecture by Oct. 31,

1991. Notice of acceptance will be given by Nov. 15. All

participants, whether giving a talk or not, have the

possibility of sending an abstract of their work on the topic

of the conference.

The collection of abstracts will be available at the

conference.

Conference fee:DM 50.P (to be paid after arrival)

The first seven GAMM-Seminars were held at Kiel in 1984, 86, 87, 88, 89, 90,

91. The corresponding proceedings have been published in the series "Notes on

Numerical Fluid Mechanics" (Vieweg, Braunschweig; vols. 10, 16, 21, 23, 30,

31).

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

From: SIAM Publications Department <SIAMPUBS@WILMA.WHARTON.UPENN.EDU>

Date: Mon, 10 Jun 91 10:20 EDT

**Subject: Table of Contents: SIAM J. Appl. Math.**

Table of Contents

SIAM J. Appl. Math. Vol. 51, No. 5, October 1991

Semi-Cherenkov and Bi-Chernekov Problems

Lim Chee-Seng

A Note on Particle-Like Solutions for a nonlinear Complex-Valued

Klein--Gordon Equation

Jon C. Luke

Implications of Viscosity and Strain-Gradient Effects for the

Kinetics of Propagating Phase Boundaries in Solids

Rohan Abeyarante and James K. Knowles

Critical Imperfection of Symmetric Structures

Kazuo Murota and Kiyohiro Ikeda

Thin Plates and Compressive Membrane Solutions I: Global Breakdown

M. E. Brewster

Asymptotic Analysis of a Premixed Polydisperse Spray Flame

I. Silverman, J. B. Greenberg, and Y. Tambour

Transition from Thermal Runaway to Propagating Flames

K. L. Henderson and J. W. Dold

The Structure and Stability of Flame Balls: A Near-Equidiffusional

Flame Analysis

C. J. Lee and J. Buckmaster

A Mathematical Model for Outgassing and Contamination

W. Fang, M. Shillor, E. Stahel, E. Epstein, C. Ly, J. McNeil, and E. Zaron

Nonlinear Hydrodynamic Stability and Spinning Deflagration of

Liquid Propellants

John K. Bechtold and Stephen B. Margolis

Forced Transverse Oscillations in a Simple Spring-Mass System

Lawrence K. Forbes

On Chains of Oscillators Forced at One End

N. Kopell, G. B. Ermentrout, and T. L. Williams

Chaotic Spikes Arising from a Model of Bursting in Excitable Membranes

David Terman

Competition in an n-Vessel Gradostat

H. L. Smith, Betty Tang, and Paul Waltman

Stochastic Properties of the Scalar Buckley--Leverett Equation

Helge Holden and Nils Henrik Risebor

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

From: SIAM Publications Department <SIAMPUBS@WILMA.WHARTON.UPENN.EDU>

Date: Fri, 14 Jun 91 10:34 EDT

**Subject: Table of Contents: SIAM Computing**

Table of Contents

SIAM Journal on Computing, Volume 20, Number 5, October 1991.

The Maximum Size of Dynamic Data Structures

Claire M. Kenyon-Mathieu and Jeffrey Scott Vitter

Time Complexity of Boolean Functions on CREW PRAMs

Miroslaw Kutylowski

Embedding of Grids into Optimal Hypercubes

Mee Yee Chan

PP Is as Hard as the Polynomial-Time Hierarchy

Seinosuke Toda

Selection Networks

Nicholas Pippenger

An Output-Sensitive Algorithm for Computing Visibility Graphs

Subir Kumar Ghosh and David M. Mount

Learning Simple Concepts under Simple Distributions

Ming Li and Paul M. B. Vitanyi

On the Communication Complexity of Solving a Polynomial Equation

Zhi-Quan Luo and John N. Tsitsiklis

The Spectra of Infinite Hypertrees

Joel Friedman

On the Complexity of Learning Minimum Time-Bounded Turing Machines

Ker-I Ko

Minimum Diameter Spanning Trees and Related Problems

Jan-Ming Ho, D. T. Lee, Chia-Hsiang Chang, and C. K. Wong

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

End of NA Digest

**************************

-------