PRACE Second Implementation Phase (PRACE-2IP) project

Twenty one PRACE members participate in PRACE-2IP, which extends and complements the work undertaken in PRACE-1IP.

- Project facts

Start date: 01/09/2011
Duration: 24 months
Budget: 35.4 €M
EC contribution: 18 €M
Grant agreement number: RI-283493

- PRACE-2IP Public Deliverables

- Objectives

The purpose of the PRACE RI is to provide a sustainable high-quality infrastructure for Europe that can meet the most demanding needs of European HPC user communities through the provision of user access to the most powerful HPC systems available worldwide at any given time. In tandem with access to Tier-0 systems, the PRACE-2IP project will foster the coordination between national HPC resources (Tier-1 systems) to best meet the needs of the European HPC user community. To ensure that European scientific and engineering communities have access to leading edge supercomputers in the future, the PRACE-2IP project evaluates novel architectures, technologies, systems, and software. Optimizing and scaling of application for Tier-0 and Tier-1 systems is a core service of PRACE.

- Action plan

PRACE-2IP is organised in twelve interrelated Work Packages that produce the project results in parallel. To structure the work they are grouped in four pillars: Tier-1 integration, Technology & Industry, Applications and Training and Dissemination. Actions directly visible to the users of PRACE include regular training events, scientific conferences and publishing a PRACE magazine focusing on the results from Tier-0 and Tier-1 projects. A special focus will be on the establishment of PRACE Advanced Training Centres (PATC) in Europe. In addition the deployment of additional Tier-0 systems and including Tier-1 systems in the European HPC ecosystem as well as a periodic call for proposals, petascaling of applications and creation of a petascaling guide will be covered by the project.

Networking activities

  • The Work Packages of the pillar Tier-1 integration will assist PRACE Research Infrastructure in implementing Tier-1 Services, building on successful work of the DEISA projects.
  • This includes DECI (Distributed European Computing Initiative), a cross-national access programme to national Tier-1 HPC resources.
  • Dissemination and training will be key activities.
  • Training material will be available through a training portal.
  • The HPC ecosystem relations include collaborations with the scientific communities, topical and national HPC centers to propose a model for cross-national access to Tier-1 resources. Dedicated focus is on the requirements of and solutions for industrial HPC users.

Service activities

  • Applications are at the core of the PRACE services for the users.
  • Building on the experience and know-how assembled in the preparatory and first implementation phase, experts from the project partners will support users of Tier-1 and Tier-0 systems to port applications – including those selected in the peer review processes.
  • They will work with the code owners to scale and optimise the applications to utilise the hundredth of thousands of processors.
  • A special focus will on be the refactoring of community codes both scientific and industrial in order to optimally map applications to future supercomputing architectures.
  • The associated techniques to come from efficient Tier-1 usage to Tier-0 scaling will be taught in the PRACE training events and published on the PRACE web site as best practice guides. The operation of the distributed infrastructure will deploy comprehensive common services at the Tier-1 level and integrate those Tier-1 resources and services in the PRACE distributed Tier-0 infrastructure.

Joint Research activities

  • To stay at the forefront of HPC and prepare for future generations of multi-Petaflop/s systems and software therefore, the PRACE-2IP project aims to prototype novel scalable computing platforms with the potential for ultra-high compute performance.
  • In addition to the research on hardware also novel programming techniques are covered by the PRACE-2IP project.
  • The work will focus on auto-tuned runtime environments, and automatic techniques to be applied in parallel programming, performance tools and file systems.
  • Partnerships with industry and research organizations will contribute to fostering European HPC developments.

The project is managed by Forschungszentrum Jülich (FZJ).

- Collaboration with other EC funded projects

EUDAT
HP-SEE
GÉANT
EGI Inspire
HPC Europa2
EMI

- Project Partners

  • FZJ – Forschungszentrum Jülich GmbH (Germany, Coordinator)
  • GCS – Gauss Centre for Supercomputing (GCS) e.V. (Germany)
  • GENCI – Grand Equipement National de Calcul Intensif (France)
  • EPSRC – The Engineering and Physical Sciences Research Council (United Kingdom)
  • BSC – Barcelona Supercomputing Center – Centro Nacional de Supercomputacion (Spain)
  • CSC-Tieteellinen Laskenta OY (Finland)
  • ETHZ – Eidgenössische Technische Hochschule Zuerich (Switzerland)
  • NCF – Stichting Nationale Computerfaciliteiten (The Netherlands)
  • JKU – Johannes Kepler Universitaet Linz (Austria)
  • SNIC – Vetenskapsradet (Sweden)
  • CINECA Consorzio Interuniversitario (Italy)
  • PSNC – Instytut Chemii Bioorganicznej Pan WPoznaniu (Poland)
  • SIGMA – UNINETT AS (Norway)
  • GRNET – Greek Research and Technology Network S.A. (Greece)
  • UC-LCA – Faculdade de Ciencias e Tecnologia da Universidade de Coimbra (Portugal)
  • NUI Galway – National University of Ireland,Galway (Ireland)
  • UYBHM – Istanbul Technical University,Ayazaga Campus (Turkey)
  • CaSToRC – The Cyprus Institute (Cyprus)
  • NCSA – National Centre for Supercomputing Applications (Bulgaria)
  • VSC-TUO – VSB – Technical University of Ostrava (Czech Republic)
  • IPB – Institute of Physics Belgrade (Serbia)
  • NIIF – Nemetzi Informacios Infrasrtuctura Fejesztese Intezet (Hungary)

Share: Share on LinkedInTweet about this on TwitterShare on FacebookShare on Google+Email this to someone