European exascale applications workshop

School of Mathematics, The University of Manchester, 11th and 12th October 2016

Exascale - all logos

Applications Workshop

Without knowing exactly what future High Performance Computing (HPC) architectures will look like, it is not easy to define the next-generation programming models to go with them. Nevertheless, this is a problem that we need to start tackling now, to ensure applications are ready to exploit the potential of Exascale computing. The underlying and prevalent questions we want to answer will be: Is preserving the investment in legacy applications and opting for an evolution of existing models a viable approach? Or do we need a more drastic approach to fully exploit future architectures?

During the applications workshop at the 11th and 12th of October in Manchester we will stimulate a discussion about communication models and their use in applications. The aim is to increase the scalability of scientific and industrial codes to exascale for those applications that require it. The main audience of the workshop are the developers of relevant applications, representatives of European HPC projects, communication model developers and representatives from industry. The workshop is organised by the European HPC projects INTERTWinE, EPiGRAM, EXA2CT and AllSCALE. It is part of a loose series of workshop, the first of which has been organised by the EPiGRAM project in 2015.

Public Part of the Applications Workshop

We would like to invite the general public, and in particular students and staff members of The University of Manchester, to join us for the public part of the applications workshop:

Location:



Seminar room: Frank Adams 1
Alan Turing Building
School of Mathematics
The University of Manchester
Oxford Road, Manchester M13 9PL
Date: Wednesday 12/10/16
 
13-14h Keynote Presentation (Bill Gropp):
MPI+X on the way to exascale
  As the time to the first Exascale system gets shorter, it is clear that MPI will be an important programming system for those Exascale systems. So we are happy to have Bill Gropp as keynote speaker. Dr. Gropp has played a major role in the development of the MPI message-passing standard. He is coauthor of MPICH, the most widely used implementation of MPI.
14-15h Panel Discussion (Bill Gropp, Mark Bull, Thomas Fahringer, Valeria Bartsch, Tom Vander Aa):
Programming for Exascale, Complete disruption or slow evolution?
  Representatives of European Exascale projects who have undertaken extensive research in the field of future programming models will discuss the future of programming models in the Exascale era.

Agenda of the Workshop for the Invited participants

Tuesday, 11 Oct 2016

09:15h-09:30h Welcome, Valeria Bartsch (Fraunhofer)
09:30h-11:00h

Project presentations (30 mins each) – Valeria Bartsch (chair)

INTERTWinE – Mark Bull, EPCC, INTERTWINE_Presentation

EXA2CT – Tom Vander Aa, imec, EXA2CT_Presentation

EPiGRAM – Stefano Markidis, KTH, EPIGRAM_Presentation
11:00h-11:30h Coffee
11:30h-12:00h

Project presentation, cont’d – Valeria Bartsch (chair)

AllScale – Thomas Fahringer, University of Innsbruck, ALLSCALE_Presentation
12:00h-13:00h

Application presentations (20 mins each) – George Beckett (chair)

iPIC3D – Stefano Markidis, KTH, IPIC3D_Presentation

TAU – Thomas Gerhold, DLR, TAU_Presentation

Ludwig – Kevin Stratford, EPCC, LUDWIG_Presentation
13:00h-14:00h Lunch
14:00h-15:00h

Application presentations (20 mins each) – Jakub Sistek (chair)

Chameleon – Mathieu Faverge, Inria, CHAMELEON_Presentation

Hlib – Ronald Kreimann, Max Planck Inst., HLIBPRO_Presentation

ILUPACK – Matthias Bollhoefer, TU Braunschweig, ILUPACK_Presentation
15:00h-15:30h Coffee
15:30h-16:30h

Application presentation (20 mins each)  – Mark Bull (chair)

Asynchronous Parallel Stochastic Gradient Descent – Mark Silberstein, Technion, ASGD_Presentation

Prototype implementation of an industrial CFD solver based on AllScale architecture – Stéphane Monté, NUMECA, FINEOPEN_Presentation

Domain Decomposition and Localised Filtering for Advection Diffusion Problems – Emanuele Ragnoli, IBM, AMDADOS_Presentation
19:00h

Workshop dinner

Wednesday, 12 Oct 2016

09:00h-10:00h

Application presentations (20 mins each) – Tom Vander Aa (chair)

MACAU – Tom Vander Aa, imec (on behalf of Jaak Simm, KULeuven), MACAU_Presentation

CP2K – Alfio Lazzaro, ETH Zurich, CP2K_Presentation

QuantumEspresso – Carlos Cavazzoni, CINECA, QUANTUMESPRESSO_Presentation
10:00h-10:30h Coffee
10:30h-12:00h Discussion session – Thomas Fahringer (chair)
12:00h-13:00h Lunch
13:00h-14:00h

Open to the public:

Keynote Presentation: MPI+X on the way to Exascale – Bill Gropp, University of Illinois Urbana-Champaign, MPI+X_Presentation
14:00h-15:00h

Open to the public:

Panel discussion “Programming for Exascale: Complete disruption or slow evolution”, Tom Vander Aa, Valeria Bartsch, Mark Bull, Thomas Fahringer
15:00h-15:30h Coffee

Confirmed invited participants

Keynote speaker: Bill Gropp
Representatives from the European Exascale Projects:




Mark Bull (INTERTWinE)

Tom Vander Aa (EXA2CT)

Stefano Markidis (EPiGRAM)

Thomas Fahringer (AllSCALE)

Michael Lysaght (PRACE)

Organising Committee:


Maksims Abalenkovs 

Valeria Bartsch

George Beckett

Thomas Fahringer

Jakub Sistek

Industry Representatives:

Christopher Armstrong (ARM)

Craig Lucas (NAG)

Application Developers:










Mark Silberstein (ASGD)

Stefano Markidis (iPIC3D)

Thomas Gerhold (TAU)

Carlos Cavazzoni (QuantumEspresso)

Mathieu Faverge (Chameleon)

Michele Pottiez (Fine/Open)

Emanuele Ragnoli (AMDADOS)

Jaak Simm (MACAU)

Alfio Lazzaro (CP2K)

Ronald Kriemann (Hlib)

Kevin Stratford (LUDWIG)

Matthias Bollhoefer (ILUPACK) 

Workshop Photograph

Exacsale workshop participants

 

Further Information

For further information about the workshop, please contact us at: info@intertwine-project.eu 

▲ Up to the top