SIAM banner

SIAM Activity Group on Linear Algebra

SIAG LA digest

Jan 4, 2016 • David Bindel

Happy New Year! This edition of the SIAG LA digest will be my last as secretary of the SIAG; as of January 1, Jennifer Pestana is the new secretary.

CONTENTS


New SIAG/LA Officers elected

Dear SIAG Linear Algebra (LA):

The newly elected officers for your SIAG are:

SIAG/LA/Linear Algebra

Chair: James Nagy

Vice Chair: Alison Ramage

Program Director: Melina Freitag

Secretary: Jennifer Pestana

Their term begins on January 1, 2016.

Best wishes to the new officers and thank you very much.

Also a big thank you to the others who participated in the election by running for office as well as handling the nomination committee.

Sincerely,
SIAM


New from SIAM: An Introduction to Domain Decomposition Methods

Announcing the December 14, 2015, publication by SIAM of:

An Introduction to Domain Decomposition Methods:
Algorithms, Theory, and Parallel Implementation Victorita Dolean, Pierre Jolivet, Frédéric Nataf

The purpose of this book is to offer an overview of the most popular domain decomposition methods for partial differential equations (PDEs). These methods are widely used for numerical simulations in solid mechanics, electromagnetism, flow in porous media, etc., on parallel machines from tens to hundreds of thousands of cores. The appealing feature of domain decomposition methods is that, contrary to direct methods, they are naturally parallel. The authors focus on parallel linear solvers.

The authors present all popular algorithms, both at the PDE level and at the discrete level in terms of matrices, along with systematic scripts for sequential implementation in a free open-source finite element package as well as some parallel scripts. Also included is a new coarse space construction (two-level method) that adapts to highly heterogeneous problems.

Audience

This book is intended for those working in domain decomposition methods, parallel computing, and iterative methods, in particular those who need to implement parallel solvers for PDEs. It will also be of interest to mechanical, civil, and aeronautical engineers, physical and environmental scientists, and physicists in need of parallel PDE solvers.

About the Authors

Victorita Dolean is currently a Reader in the Department of Mathematics and Statistics, University of Strathclyde, Glasgow, United Kingdom. She has been a research assistant at the CMAP (Center of Applied Mathematics) at the École Polytechnique in Paris, assistant professor at the University of Evry and at the University of Nice, and visiting professor at the University of Geneva. Her research has been oriented toward practical and modern applications of scientific computing by developing interactions between academic and industrial partners and taking part in the life of the scientific community as a member of the Board of Directors of SMAI (Society of Industrial and Applied Mathematics in France).

Pierre Jolivet is a scientist at CNRS in the Toulouse Institute of Computer Science Research, France, working mainly in the field of parallel computing. Before that, he was an ETH Zurich Postdoctoral Fellow of the the Scalable Parallel Computing Lab, Zurich, Switzerland. He received his Ph.D. from Université de Grenoble, France, in 2014 for his work on domain decomposition methods and their applications on massively parallel architectures. During his Ph.D., Pierre visited the Oak Ridge National Laboratory, Tennessee, for three months in 2013. One of his papers was nominated for the best paper award at Supercomputing 2013.

Frédéric Nataf is a senior scientist at CNRS in Laboratory J.L. Lions at University Pierre and Marie Curie, France. He is also part of an INRIA team. His field of expertise is in high performance scientific computing (domain decomposition methods/approximate factorizations), absorbing/PML boundary conditions, and inverse problems. He has coauthored nearly 100 papers and given several invited plenary talks on these subjects. He developed the theory of optimized Schwarz methods and very recently the GENEO coarse space. This last method enables the solving of very large highly heterogeneous problems on large scale computers. The related paper was nominated for the best paper award at Supercomputing 2013.

Contents

Preface; Chapter 1: Schwarz methods; Chapter 2: Optimized Schwarz methods; Chapter 3: Krylov methods; Chapter 4: Coarse spaces; Chapter 5: Theory of two-level additive Schwarz methods; Chapter 6: Neumann-Neumann and FETI algorithms; Chapter 7: Robust coarse spaces via generalized eigenproblems: The GenEO method; Chapter 8: Parallel implementation of Schwarz methods; Index.

2015 / x + 238 pages / Softcover / ISBN 978-1-611974-05-8
List Price $79.00 / SIAM Member Price $55.30 / Order Code OT144

To order or for more about this book, including links to its table of contents, preface, and index, please visit http://bookstore.siam.org/OT144/.

For more information about SIAM books, journals, conferences, memberships, or activities, contact:

SIAM
Society for Industrial and Applied Mathematics
3600 Market Street, 6th floor
Philadelphia, PA 19104-2688
800-447-SIAM (US and Canada)
215-382-9800
Fax 215-386-7999
service@siam.org
www.siam.org


Graph Algorithm Building Blocks (GABB) 2016: An IPDPS workshop

The Basic Linear Algebra Subprograms, introduced over 30 years ago, had a transformative effect on software for linear algebra. With the BLAS, researchers spend less time mapping algorithms onto specific features of hardware platforms and more time on interesting new algorithms.

Would it be practical to define an analogous set of basic building blocks for graph algorithms? Can we define a core set of mathematical primitives from which we can build most (if not all) important graph algorithms? If we can agree on the mathematical foundations, how would these interact with the data structures used in graph algorithms and result in an API the graph algorithms research community could support?

These questions will be the topic for the second “Graph Algorithms Building Blocks” workshop. Our goal is an interactive workshop where the full range of issues behind “Graph Algorithms Building Blocks” will be explored. We want an interactive “workshop” so papers that report preliminary results and unproven but interesting ideas will also be considered.

Submission and Dates

Submitted manuscripts may not exceed ten (10) single-spaced double-column pages using 10-point size font on 8.5x11-inch pages (IEEE conference style), including figures, tables, and references (see IPDPS Call for Papers for more details). Papers shorter than 10 pages are welcome. All papers will be peer-reviewed. Proceedings of the workshops are distributed at the conference and are submitted for inclusion in the IEEE Xplore Digital Library after the conference.

Important dates:

  • paper submissions due Jan 22, 2016,
  • author notification: Feb 5, 2016,
  • camera-ready due Feb 26, 2016.

Workshop web page: http://graphanalysis.org/workshop2016.html
Submission page (EDAS): http://edas.info/newPaper.php?c=21807.

Program Committee

  • Tim Mattson, Intel Corp. (Chair)
  • David A. Bader, Georgia Institute of Technology
  • Jonathan Berry, Sandia National Labs
  • Aydın Buluç, Lawrence Berkeley National Lab
  • John Gilbert, UC Santa Barbara
  • Jeremy Kepner, MIT Lincoln Labs
  • Chris Long, US Department of Defense
  • Kamesh Madduri, Penn State University
  • Henning Meyerhenke, Karlsruhe Institute of Technology
  • John Owens, University of California, Davis
  • Fabrizio Petrini, IBM
  • Sivan Toledo, Tel-Aviv University

ACM PASC16- Platform for Advanced Scientific Computing Conference 2016

The Platform for Advanced Scientific Computing (PASC) is inviting submissions for the Papers Session of its next conference (PASC16) cosponsored by the Association for Computing Machinery (ACM) and SIGHPC to be held from June 8 to 10, 2016 at the SwissTech Convention Center, located on the campus of the EPFL in Lausanne, Switzerland.

PASC’s structure enables efficient communication between various areas arranged in the following eight domain-specific tracks:

  • Climate and Weather
  • Computer Science and Mathematics
  • Emerging Domains
  • Engineering
  • Life Sciences
  • Materials
  • Physics
  • Solid Earth

The PASC papers program is soliciting high-quality contributions in all of these areas. Papers will be presented during the PASC16 Conference and published in the ACM Digital Library associated with the PASC conference series. Areas of interest include (but are not limited to):

  • Implementation strategies for computational science applications
  • Programming languages and models for science domains
  • Tools for application development
  • Domain-specific libraries or frameworks
  • Use of heterogeneous or advanced computing for scientific applications

To ensure the highest quality contributions, the ACM publication process includes multiple stages of review.

Authors should first submit an abstract for their paper; the abstracts will not be reviewed but are necessary for logistical reasons.

Authors will be invited to submit their full paper (pdf) shortly after abstract submission. These papers will undergo a first round of review, and the strongest papers will be selected for a second round of review. Authors will have the opportunity to revise their manuscripts based on feedback prior to the second round of reviews.

Submissions will be evaluated double blind (authors should not be listed in the (pdf) paper; references to own previous work should be in the third person). Papers should be in the ACM proceedings format and should be no more than 10 pages in length (including everything besides references, for which there is no limit). We suggest using “An Author”, “Another Author”, etc. in the ACM templates, and leaving affiliations and contact details blank.

SUBMISSION DEADLINES

To ensure a timely dissemination of research results, contributors are required to work according to the following schedule:

  • By January 15, 2016: Provide authors’ data, and title and abstract of the paper (maximum 500 words)
  • By January 22, 2016: Provide the full paper (maximum 10 pages)
  • By February 26, 2016: First review notification
  • By March 11, 2016: Provide the revised paper (if the submission has been invited to go forward to the second round of review)
  • By April 7, 2016: Final review notification

REVIEW PROCESS

The review process is organized in tracks. The editor of each track selects appropriate reviewers who are experts in the relevant area.

Submissions will be evaluated double blind: you must not include any information on the full paper (pdf) that reveals the identities of the authors or their affiliations.

Contributions are to be submitted using the PASC16 online system.

PAPERS CHAIRS

  • Torsten Hoefler (ETH Zurich, Switzerland)
  • David Keyes (King Abdullah University of Science and Technology, Saudi Arabia)

EDITORIAL BOARD OF THE SCIENTIFIC TRACKS

  • Climate & Weather: Michael Wehner (Lawrence Berkeley National Laboratory & University of California, USA)
  • Solid Earth: Jeroen Tromp (Princeton University, USA)
  • Life Sciences: Ioannis Xenarios (Swiss Institute of Bioinformatics, Switzerland)
  • Physics: George Lake (University of Zurich, Switzerland)
  • Computer Science & Mathematics: David Keyes (King Abdullah University of Science and Technology, Saudi Arabia)
  • Engineering: George Biros (The University of Texas, USA)
  • Materials: Mark van Schilfgaarde (King’s College London, UK)
  • Emerging Domains: Omar Ghattas (The University of Texas, USA)

SIAM Workshop on Network Science (NS16)

SCOPE

Network science has a rich history, tracing its roots back through sociology, chemistry, biology, electrical engineering, computer science, and topology all the way to Euler and the Königsberg bridges nearly 300 years ago. Network science studies the mathematical structure of the graphs that arise in these diverse fields, and the design, analysis, and applications of algorithms that compute with and on them. The goal of the SIAM Network Science workshop is to promote cross-fertilization among the communities that study and apply networks and graphs, both inside and outside SIAM.

NS’16 will be co-located in Boston with the 2016 SIAM Annual Meeting (July 11-15) and the SIAM Conference on Life Sciences (July 11-14).

KEY DATES

Submissions deadline: February 25, 2016
Notification: March 31, 2016
Preregistration and hotel deadline: June 13, 2016
SIAM NS’16: July 15-16, 2016

SUBMISSION

Submission site: http://easychair.org/conferences/?conf=siamns2016
Latex style: http://www.siam.org/meetings/ns16/siam-wns-style.zip

We invite contributions focused on all aspects of mathematical, algorithmic, and computational techniques in network science.

The workshop will feature three formats for contributed presentations: 25-minute talks, 5-minute lightning talks in the “Ignite” format (5 minutes, 20 slides, 15-second auto-advance), and posters. The program committee will consider all submissions for all three presentation formats.

Submit a 2-page abstract using the LaTeX style files provided on the NS’16 web site through the SIAM NS’16 EasyChair site. Accepted abstracts will be posted online in late May 2016; authors will be required to submit LaTeX source adhering to the template in April 2016.

IGNITE TALK FORMAT

The Ignite format is new for NS’16; see http://www.ignitetalks.io for a description. We especially want to use this format to highlight work by early-career researchers. If you feel your submission is particularly appropriate for an Ignite talk, please let us know using the checkbox on the EasyChair submission page.

ORGANIZATION

Workshop Co-Chairs:

  • John Gilbert, UC Santa Barbara
  • Blair D. Sullivan, NC State

Program Committee:

  • Ulrik Brandes, Konstanz
  • Bailey Fosdick, Colorado State
  • Assefaw Gebremedhin, Washington State
  • David Gleich, Purdue
  • Aric Hagberg, Los Alamos Lab
  • David Kempe, USC
  • Jeremy Kepner, MIT Lincoln Lab
  • Carl Kingsford, Carnegie Mellon
  • Christine Klymko, Lawrence Livermore Lab
  • Dan Larremore, Santa Fe Institute
  • Chris Long, US Department of Defense
  • Vince Lyzinski, Johns Hopkins
  • Aleksander Madry, MIT
  • Vahab Mirrokni, Google
  • Fabrizio Petrini, IBM
  • Cynthia Phillips, Sandia Labs
  • Lev Reyzin, UI Chicago
  • Johan Ugander, Stanford
  • Sebastiano Vigna, Milano

Modeling and OPtimization: Theory and Applications (MOPTA)

We cordially invite you to contribute to the fourteenth annual “Modeling and OPtimization: Theory and Applications” (MOPTA) conference, organized at Lehigh University, Bethlehem, Pennsylvania, August 17-19, 2016.

MOPTA aims at bringing together a diverse group of people from both discrete and continuous optimization, working on both theoretical and applied aspects. There will be invited plenary talks from distinguished speakers as well as organized sessions and contributed talks, spread over three days. Our target is to provide a diverse set of exciting new developments from different areas of optimization research while at the same time providing a setting which will allow increased interaction among the participants. We aim to bring together researchers from both the theoretical and applied communities in the framework of a medium-scale event who do not usually have the chance to interact in larger meetings.

We welcome contributed talks, and we are also very interested in hearing from people who could organize a session centered around one of their topics of expertise. MOPTA sessions have 3 speakers, with 30 mins per presentation including Q&A’s. Speakers will be kindly asked to submit their abstract on the MOPTA website. The deadline for abstract submission is May 20.

Confirmed plenary speakers (so far):

  • Michael P. Friedlander (University of California, Davis)
  • Jacek Gondzio (The University of Edinburgh)
  • Michael Pinedo (New York University)

Submissions for next SIAM-LA digest

The next SIAM-LA Digest is due to be sent out on Feb 01, 2016. Please send any postings for the next Digest to siam-la at siam.org. Only SIAG/LA members may submit postings. To contact the list owner, send an email to siam-la-owner at siam.org.