By seizing the opportunities enabled by Grid technology, the European Commission is making significant strides toward its ambitious goal to transform the European Union into the most competitive, knowledge-based economy in the world by the year 2010.
Europe is putting a lot of effort and money into Grid computing to build a next-generation IT infrastructure that enables researchers and businesses to better share knowledge and resources. These investments are creating an environment in which all of Europe's scientific and technological capital can be combined to improve European competitiveness and quality of life.
The European Grid Conference in Amsterdam Science Park in February showcased the progress and prospects of Grid computing in science and industry. The three main tracks of the conference were on scientific topics, business and industry, and on European and national Grid projects. The sessions provided a great overview and insight on the current status of Grid computing in Europe, from the research, development and deployment points of view.
Since 2000, the European Commission has been financing Grid research, then under the fifth research framework program (FP5) with more than 50 million Euros, as part of the European Union's Information Society Technologies (IST) research (http://www.cordis.lu/ist/results). In FP5, the focus was mainly on technology development and application pilots. The main projects in FP5 were the European Data Grid, GridLab, EUROGrid, CROSSGrid and DATATAG.
In FP6, beginning in 2002 and extending to 2006, the European Union is funding Grid research with more than 140 million Euros, following a technology push/application pull approach (i.e., developing the underlying technologies and Grid-enabling “real-world” applications). The crucial enabling infrastructure for the current research Grid projects is GEANT, the world's most powerful research network which links more than 3,000 research institutions across Europe.
In the first phase, the European Union has launched several significant Grid initiatives which help structure the Grid infrastructures in Europe and build upon the already established GEANT infrastructure, most notably EGEE (Enabling Grids for E-Science in Europe) and DEISA (operating a distributed/Grid terascale supercomputing facility across Europe).
In 2004, the EU has approved another 12 FP6 Grid projects, among them four major projects with a focus on Grid infrastructure technologies that will create a critical mass of expertise and resources from across Europe:
- CoreGrid is a “network of excellence” addressing longer term Grid research, creating the foundations for the next-generation Grids toward 2010 and beyond. The project brings together existing Grid research communities by creating virtual centers of excellence.
- NextGrid is an integrated project focusing on the underlying technologies of the Next Generation Grid, aiming to deliver a new Grid architecture by the end of the decade. The work addresses security and business models, taking into account requirements from sectors such as finance and media.
- Akogrimo is developing mobile Grid architectures and services. Building on Europe's strengths in mobile communications, the project will demonstrate a vision of dynamic virtual organizations in pilot applications in e-health and e-learning.
- SIMDAT is an integrated project developing generic Grid technologies for industry in the areas of data integration, collaboration and knowledge discovery. The focus is on the use of Grids to solve complex problems in important sectors such as aerospace, automotive, pharmacology and meteorology.
In addition, seven smaller projects (K-WF Grid, UniGrids, HPC4U, inteliGrid, OntoGrid, DataMining Grid and Provenance) focus on specific targets, such as knowledge extraction, workflows, data mining, collaboration, trust and security.
Finally, there is GridCoord, which has been established to support the coordination of the national Grid programs. There are currently (to my knowledge) 15 different national and regional Grid initiatives in Europe funded with several hundred million Euros at national levels and each with a potential tendency to “re-invent the wheel” of Grid computing. GridCoord is an important initiative to accelerate these initiatives by avoiding duplication and fragmentation.
Among these national projects is, most notably, the UK e-Science Program, initiated in 2000 by the UK Department of Trade and Industry, and led by champions John Taylor and Tony Hey. The key elements of the program are: building a national network of regional Grid centers; developing generic Grid middleware; supporting e-Science projects; and collaborating in international projects. The DTI funding explicitly requires matching industrial contributions. This approach led to an astonishingly early and strong involvement of industry and a high interest in the United Kingdom of spanning e-Science and e-Business with Grid services.
The UK e-Science projects have been supported by investments of more than 118 million pounds so far, resulting in a vast variety of Grid research, middleware, applications and industry projects. Examples include: DAME (Distributed Aircraft Maintenance Environment); GridCast (Grid for television/radio broadcasting); RiskGrid (for risk assessment calculations in financial services); GRIA (Grid Resources for Industrial Applications); MyGrid (a Grid portal for Bioinformatics research); and OGSA-DAI (Grid Resources for data access and integration).
Recently, the Open Middleware Infrastructure Institute (OMII) has been established at the University of Southampton to become the source for reliable, interoperabl, and open-source Grid middleware, ensuring the continued success of Grid-enabled e-Science in the United Kingdom.
As early as in May 2002, British Prime Minister Tony Blair stated that “the UK Grid intends to make access to computing power, data repositories and experimental facilities as easy as the Web makes access to information.”
Other nations are catching up. At GGF12 in Berlin in the spring of 2004, the German Minister for Education and Research announced a five-year, 100-million-Euro e-Science program. Germany started as early as 1996 with the development of UNICORE, a uniform access infrastructure connecting German supercomputer and research centers. Germany is now investing a larger portion of the 100 million Euros in the development of D-Grid, a countrywide general Grid research infrastructure and Grid services initiative, along with several (vertical) community Grids for community-specific middleware and Grid-enabled applications that focus on specific research areas.
Other national and regional Grid initiatives are: the French Grid5000 project and e-Toile projects; the Italian GRID.IT; the DutchGrid and the 40-million-Euro Dutch VL-e project to develop a virtual laboratory for e-Science in the Netherlands; the Grid-Ireland MarineGrid, CosmoGrid and WebCom-G projects; the NorduGrid with its Advanced Resource Connector (ARC) Grid middleware used by SweGrid, Estonian Grid, M-Grid in Finland, the Danish Production Grid and others; the Austrian Grid Initiative with focus on infrastructure, middleware and applications, and its ASKALON Grid application development and computing environment; Poland's PROGRESS Grid, combining resources in Krakow, Poznan, Lodz and other cities via its powerful PIONEER network; and the SEE-Grid for South Eastern European Grid-enabled e-Infrastructure Development, with partners from Greece, Bulgaria, Romania, Turkey, Hungary and Yugoslavia.
So, what makes Europe so different from other national and international Grid research projects? While early Grid initiatives in Europe where mostly unrelated point efforts (as are still many Grid projects around the world today), my impression from the European Grid Conference in Amsterdam is that, first and foremost, Europe now has a long-term, coordinated and shared Grid R&D vision, mission, strategy, roadmap and funding, driven by the European Commission's IST Framework Programmes 5, 6 and 7 (the latter will start in 2006) and hosted by its Directorate Generale (DG) for Information Society.
The 15 national Grid initiatives (which might appear at first glance to be somewhat unrelated) typically consider what already exists on the European and international level and strive to avoid spending money on re-inventing the Grid wheels. In addition, there is GridCoord, which helps to coordinate many of the national and European Grid projects. Estimating the total annual funding is difficult because of the many Grid-related projects in Europe, but it is certainly well over 100 million Euros.
Europe has embraced the notion of a “Worldwide Grid for Research”, as expressed in the e-Infrastructure's Reflection Group White Paper. For Europe, the “Worldwide Grid” will form the basis of the Information and Knowledge Society by providing or enabling many diverse elements: virtual collaborative environments; tools for education and research; planning and simulation tools for complex problem solving; virtual environments for medical treatment; storage and analysis of high-resolution data, pictures and video; providing access to massive scientific databases for disciplines from bio-informatics and bio-chemistry to meteorology, physics and astronomy; and non-scientific databases for cultural heritage, museum collections and many more.
The Grid, for Europe, is far more than resource sharing. It is a big step forward to build the Cyberinfrastructure for a united research community tackling the grand challenges of our universe. It is a coordinated, single economic engine preparing to compete with Asia and the United States. And it is a commitment, through the advancement of next-generation technology, to improve the quality of life for every citizen in Europe.
Acknowledgement
As with every attempt to describe a complex field in simple words, this article lacks many details. For example, this article doesn't mention at all the Grid initiatives of the early adopters in industry. Also, there is a lack of mentioning all the other great Grid projects in Europe, especially the courageous early ones like the FP4 Metacomputing projects in the mid 1990s; Unicore (started in 1996); the White Rose Grid in 2001; the bold commercial start-ups like GridXpert, GridSystems and Gridwise; and certainly all the hard work (mostly behind the scenes) from the European Commission's DG for Information Society.