zcam
zcam

2017 activities poster available here

The 2017 ZCAM program is ready

The 2016 ZCAM program is now finished!

Reunión de Usuarios y Desarrolladores de Métodos de Simulación de Aragón

Electronic Structure Library coding workshop: solvers

Prof. Emilio Artacho, CIC Nanogune Consolider, San Sebastián
Prof. Volker Blum, Duke University
Prof. Fabiano Corsetti, Imperial College London

DATE OF EVENT : 06/06/2016       DURATION : 11 day(s)

LOCATION : ZCAM Campus Actur C/ Mariano Esquillor s/n Edificio I+D 50018 Zaragoza
URL : http://www.cecam.org/workshop-1274.html


SUMMARY :

The main focus of the ESL coding workshops is to get developers together to produce concrete results within the workshop itself, i.e., discussing ideas for libraries and then coding them, either from scratch or by adapting routines from existing code bases. Another important goal is to define the general direction of the ESL as a whole, and to implement any new infrastructure that is needed for this. Most of the basic structure of the ESL has already been set up in the previous two workshops, and is being put into practice in the wiki and the existing libraries.

Expanding on the successful format of previous years, we propose a two-week coding workshop with a team of 15 coders connected with as many different electronic structure codes as possible. The first two days will be dedicated to presentations and discussions amongst a larger number of attendees (the coding team, the Advisory Committee of the ESL, guest speakers, and other invited stakeholders). The initial discussions will focus on two main topics: Firstly, how to bring the ESL forward into a new
phase by enhancing the integration and compatibility of its separate constituent libraries, making it easier for users to package and deploy multiple libraries together. This will build upon some of the decisions taken in the previous workshop (on language and library recommendations and common low-level utilities).

In particular, we will discuss the available options for storing the libraries in one place and performing automatic testing, as well as how to organize oversight mechanisms to ensure high-quality of contributions.
Secondly, we will look at the changing landscape in computer architectures and coding practices, particularly with reference to high-performance computing (HPC), and how this should affect the electronic structure community and the ESL. We aim to inform the discussion by inviting speakers to talk on different aspects of this topic, both from within the electronic structure community and from HPC centres and industry; furthermore, we will hear the report from a previous ESL workshop on the subject of big data for massively parallel architectures. After the discussion days, the rest of the workshop will be dedicated to the coding effort. This year, development of new libraries will revolve around the broad theme of solvers. We plan to work on three specific library bundles, all relating to crucial parts of any density-functional theory code:

-Kohn-Sham eigensolvers: Development for the ESL is already underway in this area. New steps forward will be taken through a new “Electronic Structure Infrastructure” (ELSI) project, in direct coordination with
ESL and also connecting U.S. based researchers to the wider ESL effort. ELSI will bring together several different approaches to solving or circumventing the Kohn-Sham eigenvalue problem, including a density
matrix solver (libOMM), a resolvent-based method (PEXSI), and a scalable diagonalization method (ELPA), and will be open to integrating other open source solvers that are compatible and/or complementary.

-Poisson solvers: similarly to the eigensolvers, the aim will be to implement in a single package several different algorithms of use in different situations, providing a unified and clean interface for the user.
Special attention will go to allowing different FFT back ends to be connected to the library.

-Atomic solvers: this is one of the most notable examples of replicated development in different code bases, with no modern independent library available. The aim will therefore be to create a new library with a modern, coherent list of features, ultimately targeting spherical Hartree-Fock, semilocal and hybrid DFT, and different optional relativistic levels of theory. It will also be connected to LibXC for immediate access to many different exchange-correlation functionals (and, in doing so, will avoid replicated development).