NA Digest, V. 16, # 33

NA Digest Monday, September 05, 2016 Volume 16 : Issue 33

Today's Editor:

Daniel M. Dunlavy
Sandia National Labs

Today's Topics: Subscribe, unsubscribe, change address, or for na-digest archives:

Submissions for NA Digest:

From: Robert van de Geijn
Date: September 01, 2016
Subject: Educational Materials on Optimizing GEMM

Over the years, how to optimize matrix-matrix multiplication has been
a common exercise in courses on high-performance computing, computer
architecture, compilers, and numerical linear algebra.

In the past, some have found the wiki "How-to-optimize-gemm", created
for one of my classes, useful. The wiki server where this was hosted
is no longer functional. We have moved this material to

A somewhat more advanced tutorial, BLISlab, that teaches the basic
principles that underlie BLIS can also be found on github:

(The document that describes the exercise is in file "tutorial.pdf" in
that repository and is also available from arXiv:

We hope you will find these materials useful and are always open to

From: Jiguang Sun
Date: September 01, 2016
Subject: New Book, Finite Element Methods for Eigenvalue Problems

Finite Element Methods for Eigenvalue Problems
Jiguang Sun, Aihui Zhou

July 18, 2016 by Chapman and Hall/CRC
Reference - 343 Pages - 47 B/W Illustrations
ISBN 9781482254648 - CAT# K23972
Series: Monographs and Research Notes in Mathematics

Summary: This book covers finite element methods for several typical
eigenvalues that arise from science and engineering. Both theory and
implementation are covered in depth at the graduate level. The
background for typical eigenvalue problems is included along with
functional analysis tools, finite element discretization methods,
convergence analysis, techniques for matrix evaluation problems, and
computer implementation. The book also presents new methods, such as
the discontinuous Galerkin method, and new problems, such as the
transmission eigenvalue problem.

Features: Offers an up-to-date treatment of the finite element method
and eigenvalue problems that have been significantly expanding
Provides a systematic study of theory, algorithm, and implementation
of finite element methods for eigenvalue problems Includes MATLAB
codes for eigenvalue problems Covers typical eigenvalue problems along
with new methods.

From: Bruce Bailey
Date: August 31, 2016
Subject: New Book, Learning LaTeX, Second Edition

Learning LaTeX, Second Edition, by David F. Griffiths and Desmond J.
Higham. x + 103 pages / Softcover / 978-1-611974-41-6 / List Price
$29.00 / SIAM Member Price $20.30 / OT148

"For years, my only system for training students to LaTeX has been to
hand them a copy of Learning LaTeX and to tell them to start writing.
This has worked without fail for dozens of students. I never would
have thought it possible, but the new edition is a substantial
improvement with the additional coverage of BiBTeX, Beamer, and
posters. Learning LaTeX should be handed to new graduate students in
mathematical sciences along with their office key and ID card." -Don
Estep, Colorado State University.

Here is a short, well-written book that covers the material essential
for learning LaTeX. It includes incisive examples that teach LaTeX in
a powerful yet abbreviated fashion. This is the handbook to have if
you don't want to wade through extraneous material.

This second edition of the well-regarded and highly successful book
includes additional material on the American Mathematical Society
packages for typesetting additional mathematical symbols and multi-
line displays; the BiBTeX program for creating bibliographies; the
Beamer package for creating presentations; and the a0poster class for
creating posters.

To order or for more about this book, including links to its table of
contents, preface, and index, please visit .

From: Petter Bjorstad
Date: August 29, 2016
Subject: Domain Decomposition Methods, Norway, Feb 2017

The DD24 Conference is now open for registration and submission of
Contributed Papers / Posters. The Conference will take place from
Feb. 6th. to Feb. 10th. in Longyearbyen, Svalbard. This is the
Northernmost destination on the planet that can be reached by a
regular, commercial flight. Please see the Conference WEB page for
additional info.

The conference will have 12 invited, plenary speakers and 18 scheduled
minisymposia with more than 100 speakers. (See the web-page)

Join us in the Arctic Night ! (There will be no daylight from the
beginning to the end), home of the polar bears!

The Local Organizing Committee email:

From: Federico Poloni
Date: August 31, 2016
Subject: Matrix Equations and Tensor Techniques, Italy, Feb 2017

The 2017 Matrix Equations and Tensor Techniques workshop will be held
in Pisa, Italy, on 13-14 February 2017. It is the seventh in a series
of biennial workshops on matrix equations and tensor techniques. As in
the previous meetings, the focus will be on the latest developments in
the theory, computation and applications of linear and nonlinear
matrix equations and tensor equations.

Further information and registration:

- Abstract submission: December 4, 2016
- Acceptance notification: December 16, 2016
- Registration: January 27, 2017

From: Roman Hatzky
Date: September 01, 2016
Subject: Position, High Perf Comp, MPI for Plasma Physics

On behalf of the European Fusion Community the Max-Planck Institute
for Plasma Physics (IPP) offers an opportunity to participate in a
group ( for the high level support of European
code developments for magnetic fusion related plasma physics and
material modelling. Experts in high performance computing (HPC) are
invited to apply for a position in this group.

More information can be found under the code number 44/16 at:

From: Nick Gould
Date: August 31, 2016
Subject: Senior NA Position, STFC Rutherford Appleton Laboratory, UK

We have an opening for a senior researcher to join the Numerical
Analysis Group at the STFC Rutherford Appleton Laboratory, near
Oxford, England. The Group's main focus is on topics relating to
large-scale numerical linear algebra and optimization.

The position will involve undertaking novel independent and
collaborative scientific research at an international level in an area
of computational numerical analysis that complements and contributes
to the work of the Group, and is in line with the Group's current
EPSRC research grant "Least Squares: Fit for the Future". The role
will also contribute to developing, supporting and maintaining
software within one or more of the Group's key software projects
(notably HSL, SPRAL and GALAHAD).

Further responsibilities will include developing national and
international research collaborations, working with colleagues on the
development and expansion of projects involving STFC scientific
facilities (notably the ISIS pulsed neutron and muon source, Diamond
Light Source and RAL Space), and the writing of grant applications to
secure future funding for numerical analysis research and software
development projects.

For further details, please visit
or contact us (, by
email for particular queries. The closing date for applications is 09
October 2016, with interviews scheduled on a date to be confirmed.

From: Shan Zhao
Date: August 29, 2016
Subject: Faculty Positions, Computational & PDE Modeling, Univ of Alabama

The Department of Mathematics at The University of Alabama invites
applications for two tenure-track positions at the Assistant Professor
or Associate Professor level, in the area of computational and PDE
modeling, beginning on August 16, 2017. One position is in the areas
of numerical solution of PDEs and computational modeling of
multi-physics processes. The other is in the areas of PDE modeling and
multiscale analysis. For both positions, we especially seek candidates
whose research has potential applications in water related fields,
such as fluid mechanics, transport problems, flood forecasting, flood
inundation mapping, hydrologic analysis, surface runoff, groundwater
modeling, porous and fractured media, and real-time simulation and
forecasting, in order to foster collaborations with the recently
established National Water Center (NWC) in Tuscaloosa

Candidates must possess a doctoral degree in mathematics or a closely
related field by August 16, 2017. Preference will be given to
applicants with postdoctoral experience. Candidates with strong track
record and well documented success in external funding will be
considered at the rank of Associate Professor.

Applicants should complete the online application at The application should
include a letter of application, a current curriculum vita, a research
statement, a teaching statement, and three letters of recommendation
(one of which concerns teaching). The recommendation letters should be
sent electronically to: Applications will be reviewed on
an ongoing basis starting December 1st, and will continue to be
accepted until the position is filled. The University of Alabama is an
Equal Opportunity/Affirmative Action employer.

From: Lili Ju
Date: August 29, 2016
Subject: Postdoc Position, Univ of South Carolina

Postdoc position
Department of Mathematics
University of South Carolina

A postdoctoral position is available immediately in computational
ocean modeling. The project involves the design, implementation, and
testing of advanced numerical algorithms and/or their parallel
implementation for spatially-dependent time stepping methods, for
large time-step schemes, and for variable resolution spatial mesh
generation. The postdoc joining the team would work on one or more of
these research directions. This appointment is renewable for up to
three years.

Regardless of the areas of Ph.D. degree obtained, desirable skills
include algorithm development, testing, and implementation and
sophisticated computing capabilities. Some knowledge of climate
modeling is desirable.

Inquiries and applications should be addressed to
Lili Ju,

If applying, please send, to this email address, a current CV and have
2 letters of reference sent to this address as well.

From: Neha Mistry
Date: September 02, 2016
Subject: Contents, Information and Inference: A Journal of the IMA, 5 (3)

Information and Inference: A Journal of the IMA. 5 (3)
This issue is available at

Super-resolution of point sources via convex programming, Carlos

Detecting the large entries of a sparse covariance matrix in
sub-quadratic time, Ofer Shwartz and Boaz Nadler

Near-optimal estimation of simultaneously sparse and low-rank matrices
from nested linear measurements, Sohail Bahmani and Justin Romberg

From: Prof. David G. Yu
Date: August 31, 2016
Subject: Contents, Statistics, Optimization & Information Computing, 4 (3)

Statistics, optimization & information computing, Vol. 4 No. 3,

A Comparison of Compressed Sensing and Sparse Recovery Algorithms
Applied to Simula Applied to Simulation Data, Ya Ju Fan, Chandrika
Kamath, DOI 10.19139/soic.v4i3.207

A Criterion for Testing Hypothesis about Impulse Response Function,
Iryna Rozora, Yu. V Kozachenko, DOI 10.19139/soic.v4i3.222

A statistical model of macromolecules dynamics for Fluorescence
Correlation Correlation Spectroscopy data analysis, Dmitri Koroliouk,
Vladimir Semenovich Koroliuk, Eleonora Nicolai, Paolo Bisegna, L
Bisegna, Lorenzo Stella, Nicola Rosato, DOI 10.19139/soic.v4i3.219

Learning Unknown Structure in CRFs via Adaptive Gradient Projection
Method Method, Wei Xue, Wensheng Zhang, DOI 10.19139/soic.v4i3.228

On Size Biased Kumaraswamy Distribution, Dreamlee Sharma, Tapan Kumar
Chakrabarty, DOI 10.1 10.19139/soic.v4i3.217

Nonmonotone Spectral Gradient Method for l_1-regularized Least
Squares, Wanyou Cheng, Qingjie Hu, DOI 10.19139/soic.v4i3.230

End of Digest