••• 1

European e-Infrastructure activities

Experiences gained The way ahead

Leif LaaksonenCSC, the Finnish IT center for science

European Commission - DG INFSOGÉANT & e-Infrastructure Unit

Moscow 23-25 April 2007

••• 2

HPC in Europe TaskforceTowards a new level of High Performance

Computing facilities for Europe

• Dr. Kimmo Koski• CSC - The Finnish IT Center for Science• http://www.hpcineuropetaskforce.eu/

••• 3

Computational science infrastructure

••• 4

HPC in Europe Taskforce (HET)

• Temporary taskforce shaping the European strategy for petaflop computing

• Founded in June 2006• Strategy work delivered in January 2007• Members from 11 European countries• Chaired by CSC, Finland

• Complete documentation available: http://www.hpcineuropetaskforce.eu/

• EU FP7 project to build the petaflop infrastructure based on HET strategy is currently planned

••• 5

HET Scope

• The upper layers of the pyramid – HPC centers / services– European projects

(HPC/Grid, networking, …)

• Activities which enable efficient usage of upper layers– Inclusion of national

HPC infrastructures– Software development

and scalability issues– Competence

development

••• 6

Basic Ecosystem

• Balanced approach with sufficient peak and strong base of the pyramid

• Access to the extreme computing power for the top European scientists

• Strong support by software and competence development activities

• Methods to allocate cycles and exchange infrastructure and competence based resources

• Inclusion for the whole European Union

National/regional centers, Grid-collaboration

Local centers

EuropeanHPC center(s)

••• 7

HET Recommendations

• Recommendation for the development and operation of a “top end” infrastructure– HET recommends establishment of a small number of European HPC

facilities to provide extreme computing power – exceeding petaflop capability – for the most demanding computational tasks.

• Recommendation for developing the full European ecosystem– HET recommends increased emphasis on the development of the full

HPC ecosystem, including the local infrastructure, national and regional facilities, top-level European computing capabilities and the interoperability of their services.

• Recommendation to enable petascale computing by supporting the development of novel software architectures– HET recommends starting a range of activities aimed at addressing the

key issues in building software that allows exploiting the performance potential of petascale machines in a coherent, efficient, scalable and sustainable manner.

••• 8

HET Recommendations (cont.)

• Recommendation to implement an efficient and highly reliable infrastructure for storing large amounts of data – HET recommends increasing emphasis on permanent and

persistent data repositories as a part of the HPC ecosystem. • Recommendation to support competence development in

computational science through extensive training and education activities – HET recommends supporting extensive training and education

activities which focus on enabling more efficient and higher quality use of the top-end facilities in the long run

• Recommendation to raise the visibility of HPC– HET recommends strong activities to increase the visibility and

improve the publicity of computational science in order to highlight the strategic impact and need of numerical simulation for most areas of science and engineering

••• 9

HET Recommendations (cont.)

• Recommendation to boost collaboration – HET recommends support for collaborative actions with

a target to link the major players in HPC Ecosystem – existing grid and HPC projects, national and regional centers, main computational research groups, funding organizations and potential new planned FP7 efforts – in order to maximize synergy for actions.

• Recommendation to support European industry– HET recommends support for collaborative efforts with

European industrial HPC users and European HPC industry at large – from hardware and software R&D to product design and manufacturing.

••• 10

HET outcome

• Entry in ESFRI list for petaflop computing• Papers:

– Scientific case for European HPC (most work done by previous HPCEUR project)

– Proposal for funding models– Proposal for peer review process– Views for HPC Ecosystem– Summary paper with recommendations

• Good team spirit with a common approach • Basis for practical implementation

– Consortium for ESFRI Preparatory phase– Memorandum of Understanding for European Tier 0 HPC

service

••• 11

Partnership for Advanced Computing in Europe (PACE)

• Dr. Kimmo Koski• CSC - The Finnish IT Center for Science

••• 12

Partnership for Advanced Computing in Europe (PACE)

••• 13

PACE

• Project proposal for preparatory phase, call deadline 2.5.2007• Memorandum of Understanding, 14 countries signed and more to come• PACE consortium with 14 partners (14 countries)

– Austria– Finland– France– Germany– Greece– Italy– Norway– Poland– Portugal– Spain– Sweden– Switzerland– The Netherlands– United Kingdom

••• 14

ESFRI

• Strategy Forum with a consulting role to EU• Wide representation of scientists in various disciplines• Roadmap process for major new European research

infrastructures (range of 10-1000 MEUR for an infrastructure)

• Roadmap published in 2006– 35 projects labeled mature– One of the projects European HPC Service

• Preparatory projects for each project – 1-4 years– Deadline for project call May 2nd, 2007

••• 15

New market for European HPC

• 35 ESFRI list new research infrastructure projects, most of them starting a preparatory phase project late 2007– 1-4 years– 1-7 MEUR * 2 (petaflop computing 10 MEUR * 2)

• Successful new research infrastructures start construction 2009-2011– 10-1000 MEUR per infrastructure

• Existing infrastructures are also growing• Results:

– Growing RI market, considerably rising funding volume – Need for horizontal activities (computing, data, networks,

computational methods and scalability, application development,…) – Real danger to build disciplinary silos instead of searching IT synergy

••• 16

Targets for European HPC collaboration2007 onwards

• Continuation of existing grid projects (DEISA, EGEE …) and development in GEANT2 network infrastructure

• Building European petaflop computing services integrated in the full HPC ecosystem according to the performance pyramid model (PACE)

• Maximal synergy between PACE and DEISA (integration after some time?)

• Interoperability between PACE and EGI/EGEE• Building up research infrastructure services for ESFRI

roadmap• Target to establish an active European community for HPC:

infrastructure, resource sharing, communication and collaboration over country borders

••• 17

e-IRG

• Dr. Leif Laaksonen• CSC - The Finnish IT Center for Science

••• 18

Goals and Background

• Initiated in 2003 during the Greek EU Presidency– First workshop in Spring 2003, Athens

• Mission statement drafted in the Rome meeting– “to support on the political, advisory and monitoring

level, the creation of a policy and administrative framework for the easy and cost-effective shared use of electronic resources in Europe (focusing on Grid-computing, data storage, and networking resources) across technological, administrative and national domains.”

• Starting point: impact of the major investments in the Grid and networking projects depends also on policies governing the use of the e-Infrastructure.

www.e-irg.eu

••• 19

Membership and Mandate

• Delegates appointed by the member state ministries– 28 European countries represented

• Maximum 2 representatives/country• Usually appointed by the ministries of education

– Observers• European commission• Major e-Infrastructure projects

– EGEE, DEISA, ...

• NREN Policy Committee

• Recognised as an advisory body by the EC– In addition to member states themselves

www.e-irg.eu

••• 20

Working methods

• Publish e-IRG White Paper– “Snapshot of the policy and state of the art in technology”– Short- to medium-term scope

• Develop the e-IRG roadmap– Long-term strategic issues– Responses to emerging technologies, paradigm shifts.

• Four annual meetings– Two open workshops

• Host discussions within the community involved with the Grid and research networking

– Part of the e-IRG White Paper process • Technology/policy-axis

– Two delegates meetings• Endorse recommendations, discuss strategic issues

www.e-irg.eu

••• 21

DEISA

• Prof. Victor Alessandrini• IDRIS - CNRS• www.deisa.org

••• 22

DEISA objectives

• To enable Europe’s terascale science by the integration of Europe’s most powerful supercomputing systems.

• DEISA is an European Supercomputing Service built on top of existing national services. This service is based on the deployment and operation of a persistent, production quality, distributed supercomputing environment with continental scope.

• Main focus is High Performance Computing (HPC).

• After three years of operation, DEISA confirms its original motivation of providing a basic vector of integration of HPC resources at the continental scale. The DEISA services have been tailored to enable seamless access to, and high performance cooperative operation of, a distributed park of leading supercomputing platforms in Europe.

• DEISA services are deployed on top of a dedicated high speed network infrastructure connecting computing platforms, using selected middleware. Their primordial objective is enabling capability computing across remote computing platforms and data repositories.

••• 23

BSC Barcelona Supercomputing Centre Spain

CINECA Consortio Interuniversitario per il Calcolo Automatico Italy

CSC Finnish Information Technology Centre for Science Finland

EPCC/HPCx University of Edinburgh and CCLRC UK

ECMWF European Centre for Medium-Range Weather Forecast UK (int)

FZJ Research Centre Juelich Germany

HLRS High Performance Computing Centre Stuttgart Germany

IDRIS Institut du Développement et des Ressources France

en Informatique Scientifique - CNRS

LRZ Leibniz Rechenzentrum Munich Germany

RZG Rechenzentrum Garching of the Max Planck Society Germany

SARA Dutch National High Performance Computing The Netherlands

and Networking centre

Participating Sites

••• 24GEAN

T

AIX distributedsuper-cluster

Vector systems(NEC)

Linux systems(SGI Altix, IBM, …)

THE DEISA SUPERCOMPUTING GRID

Systems interconnected with dedicated 10 Gb/sNetwork provided by GEANT.

••• 25

How is DEISA enhancing HPC services in Europe?

• Running larger parallel applications in individual sites, by a cooperative reorganization of the global computational workload on the whole infrastructure.

• Enabling workflow applications with UNICORE (complex applications that are pipelined over several computing platforms)

– In some cases (whan it makes sense) enabling coupled multiphysics Grid applications.

• Providing a global data management service whose primordial objective is the tight integration of distributed data with distributed computing platforms (this paves the way and is critical to the efficient operation of future shared European supercomputers):

– Enabling efficient, high performance access to remote datasets (with Global File Systems and stripped GridFTP).

– Integrating hierarchical storage management and databases in the supercomputing Grid.

• Deploying portals as a way to hide complex environments to new users communities, and to interoperate with another existing grid infrastructures.

••• 26

Enabling science: DEISA users communities

National users communities have accountson a given site and do not « naturally » see thewhole DEISA environment.

Promotion of DEISA users is done via theExtreme Computing Initiative.

European call for proposals for grand challengesimulations every year in May since 2005.

About 50 grand challenge projects supportedeach year since 2005.

Full information about Extreme Computing projects and reports from terminatedprojects can be found in the DEISA Web server: www.deisa.org

The Extreme Computing Initiative is the current DEISA service provisioning model. This, however, will evolve in FP7.

••• 27

e-IRG and access to research data

• Dr. Leif Laaksonen• CSC - The Finnish IT Center for Science

••• 28

Facing challenges in

• Creating new information and turning it into knowledge– Using money to create knowledge (research).– Computational sciences generate data and knowledge.– Education efforts to support the creation of new

information.– Including new groups into the science process.

• Turning new knowledge into something beneficial for the society– Using innovation to turn knowledge into money

(innovation).– Data storage and access to scientific information.– Educational efforts to support knowledge transfer.

• The goal is to create and foster a service based collaborative science environment.

• The common denominator are e-infrastructures with a clear user policy for resource sharing . Challenging but must be possible!

••• 29

Key components in the e-Infrastructure

• Computational sciences both create and need data.• Data management, repositories, curation and analysis.• Collaborative environments (Virtual Organisations). • Education and training

• Supporting actions– Authentication, authorisation and accounting.– Policy issues

• access to data• resource sharing

National/regional centers,Grid-collaboration

Local centers

EuropeanHPC center(s)

The Performance Pyramid

••• 30

Example: Data exchange (basis of knowledge exchange)

– Publicly funded data are a public good• Should be as open as possible, with as few restrictions as

possible, on a non-discriminatory basis, and available freely or at the lowest possible cost

• Good stewardship of public knowledge– Data are central to the scientific research process

• Data is basis of value chain of science and technology, optimum return on public investment

• Strong value chains of innovation– Data sharing issues are international in scope

• ICT makes multidisciplinary and international collaborations possible

• Key scientific and social problems are global: health, environment

• The creation of value from international cooperation

OECD Follow-up Group on Issues of Access to Publicly Funded Research Data, 2002

••• 31

Core Principle

• Publicly funded research data should be openly available, subject only to compelling superseding considerations and policies

– Limited period of exclusive use by principal investigators,

– Protection of confidentiality and privacy,– National security, and– Respect for intellectual property rights.

OECD Follow-up Group on Issues of Access to Publicly Funded Research Data, 2002

••• 32

Towards sustainable grid-based e-Infrastructures

• FP5: Broad scale test-beds• FP6: Production quality facilities• FP7: Sustainable grid-based e-Infrastructures (utility model)• To support the transition

– e-IRG White Paper– e-IRG Roadmap– Opportunities list – http://www.e-irg.eu/publ/

• Task Force on Sustainable e-Infrastructures (SeI). Final report distributed widely.

• New batch of e-IRG recommendations approved during the Finnish EU Presidency on the 20th November 2006.

••• 33

e-Infrastructure Reflection Group (e-IRG)Task Force on Sustainable e-Infrastructures

• Recommendations:1. Governments and the Commission should develop policies and

mechanisms to encourage increased investment in a more coherent and interoperable way across Europe.

2. The existing e-Infrastructure projects must be superseded by integrated sustainable services at national and European levels.

3. e-Infrastructures must be application-neutral and open to all user communities and resource providers. National funding agencies should be encouraged to fund multi-disciplinary and inclusive infrastructures rather than disciplinary-specific alternatives.

4. e-Infrastructures must inter-operate and adopt international standard services and protocols in order to qualify for funding.

5. The Commission should, within the seventh Framework Programme, develop a pan-European e-Infrastructure which explicitly encourages the further integration of national e-Infrastructure initiatives .

http://www.e-irg.eu/publ/2006-Report_e-IRG_TF-SEI.pdf, 2006

••• 34

Conclusions for fostering an eco-system

• Computational science is indispensable to the solution of complex problems in every sector.

• The strategic significance of computational science has not been fully recognized.

• Knowledge of computational science needed to solve key problems in science and engineering.

• Extend the frontiers of scientific simulation through a new generation of computational models that fully exploit the power of advanced computers and collaboratory software that makes scientific resources available to scientists anywhere, anytime.

• Coordinate education and training efforts, with an emphasis on the efficient exploitation of e-Infrastructures (software, algorithm and middleware).

• Most challenges are not technical but policy issues.

••• 35

Nordic e-Science

• Prof. Juni Palmgren• Stockholm University and Karolinska

Institutet, Stockholm

• Leif Laaksonen

••• 36

Strengthening the Nordic co-operation

• NORDUnet (network) and NORDUnet2/NORDUnet3 (projects).

• Nordic Data Grid Facility service– NDGF, is a collaboration between the Nordic countries

(Denmark, Finland, Norway, Sweden).

• Nordic Computational Grand Challenge Survey.– The primary aim of the survey was to identify ’Grand

Challenge’ scientific problems that are considered important to the research community and whose solutions require access to large scale computational or storage resources, possibly larger than any of the existing supercomputer facilities or larger than the aggregate capacity provided by any of the national infrastructures in the Nordic countries.

• NCM eScience ad hoc workgroup.

••• 37

NCM ad hoc eScience workgroup Dec 2006-Oct 2007

• To produce an eScience strategy document in language adequate for political decision makers, focusing on the point if, how and why the Nordic countries’ research ministries and research councils can cooperate, aiming to make eScience a Nordic strength in a global context.

• Not just in terms of ICT infrastructure, but as an enabling factor for advanced ICT applications, which will serve not only research but society at large.

• Technologies may exist – policy decisions may be lacking.

••• 39

Draft goals/recommendations

Added value by enabling:

• A Nordic eScience eco-system - shared Nordic resources. Market for e-Science resources.

• Cooperation in key Nordic eScience research and application programmes. Clever focusing of funding.

• A Nordic eScience education programme. Big need for education and training.

• A sustainable (virtual) Nordic eScience organisation and funding plan.

• Access to data resources across thematic and country borders.

• A joint Nordic gateway to Europe and beyond.

••• 40

Points of departure

• National HPC and eScience programmes• Ongoing Nordic collaborations

– NORDUnet and Nordic Data Grid Facility (NDGF)– Nordic Grand Challenge Initiative (March 2007)– NCoE’s (NCoEDG, NECCwith eScience components)– Scientist initiated collaborations– ...

• Efforts in Europe and beyond: EU FPs and ESFRI– Existing: e-IRG, Deisa, EGEE, Sirene, FP6 programmes– New: European High-performance computing service (PACE)– New: European Grid Initiative (EGI)– New: Preparatory phase for ESFRI roadmap projects; 35 projects with

eScience interface. FP7 programmes.

••• 41

Further information

More info on e-Infrastructures: www.cordis.europa.eu/ist/rn/