GridCoord d 1



Yüklə 0,69 Mb.
səhifə2/29
tarix27.10.2017
ölçüsü0,69 Mb.
#16992
1   2   3   4   5   6   7   8   9   ...   29

Analysis


This analysis is an attempt to digest and make readable the large amount of information accumulated during the collection phase of the GridCoord survey project. The complete contributions of the GridCoord partners can be found in the AnnexPart II of this compendium. The diversity and heterogeneity of the individual contributions by the GridCoord partners emphasises the need for a comprehensive synthesis of the collected information. The diversity of the contributions is due to the different approaches to Grid Research, to national and cultural traditions, to different funding schemes. This diversity is also a proof and evidence how vibrant Grid research is in Europe.

The analysis starts with a short historical overview of Grid research in Europe, followed by an examination of the major current research directions and motivations. The driving forces are investigated and where feasible the degree of adoption in various field of science, industry and business. The analysis attempts as well to comprehend and interpret the different funding schemes, and provide an approximate figure of the amount of money spent for Grid Research in the surveyed GridCoord countries. Finally, the analysis shortly describes links to industry and business and their involvement.


Historical overview of Grid Research in GridCoord Countries


The name “Grid” was coined in the US in the mid ’90. H, however, this does not mean that Europe was not already active in research in parallel and distributed computing, supercomputing, and metacomputing. Several European countries have a long tradition in the above-mentioned fields and pioneered research activities. The following table gives a historical glimpse at the research approaches that led to the development of Grid Computing in the surveyed countries. We can distinguish between loose, bottom-up approaches pioneered by research communities (computer science and applications) and a more co-ordinated, top-down approach coming from funding authorities. Also, a distinction can be made between infrastructure-oriented approaches which are often more coherent and research-oriented approaches that tackle specific open problems of Grid research. The rapid development of networking technology has allowed for easier coupling of distributed applications and the progress European countries made in deploying advanced research and education networks and their interconnection were instrumental in triggering the uptake of Grid Computing.

Grid computing can be seen as an evolutionary technology, capturing existing technology trends (technology push) like networked computing, clusters and blades that enable new sets and mechanism for delivering ubiquitous and predictable services. In addition there is an application pull, initial coming from several sciencestific disciplines needing access to instruments and databases distributed all over the world. Applications have evolved from once simple, often monolithic, computer-or server-centric binaries to services that are disaggregated and then distributed across the network. However, there is also a revolutionary and radical component/perspective: Grid computing mandates a systemic rather component-centric approach: it requires giving up personal ownership of hardware and using the resources of others, a collectivistic mentality that must will have to learn to considerhas to be learned. Grid computing is designed for change, going far beyond sheer computing power. Hence, the reason why Grid Computing is a radically different approach to using effectively computing power is not only technological, but it also has a very important sociological componentHence, the reason why Grid Computing is radical is mainly sociological.

The following table is a summary of the history chapters from the individual contributions.

History Summary Table

Country

Timing

National Initiatives/Projects

Approaches/
Drivers


Institutions

Austria

Before 2003

Since 2003



Individual Researchers; Austrian Centre for Parallel Computation (ACPC)
Austrian Grid Consortium

Bottom-up

Co-ordinated



Universities


France

Since 1996
Since 1998
Since 2001
Since 2003 2005

Individual Researchers,
RNRT, RNTL programmes,
ACI
Grid5000

Reorganisation in the newly created ANRGrid5000, “National Research Agency”



Computer Science oriented;
Bottom-up, loose co-ordination

INRIA, CNRS, Universities

Germany

Early 1990
Late 1990
Since 2004

Individual research projects
Unicore, Access to HPC systems
D-Grid Initiative

Loose co-ordination
Regional initiatives; Application-driven

Universities, Research Labs

Hungary

Since 2000
Since 2003

Early work on Grids
Hungarian Grid Competence Centre

Bottom-up, Infrastructure-oriented

Universities, MTA SZTAKI

Italy

Since 1990

Since 1994


1995-2000


2000
2002-2004

2002-2005


Early 2003

2005


Parallel Computing and High Performance Computing
HPC

Special programme on Metacomputing for the solution of large-scale problems in engineering


PQE2000 Project
INFN Grid
Grid Computing: enabling technologies and applications for e-science
High-performance Large-scale Distributed Platform
Grid.it, PON
IG-BIGEST (Italian Grid for e-Business, e-Industry e-Government and e-Science & Technology)
Grid Technology Platform

Computer Science oriented;
Parallel and Metacomputing oriented

Infrastructure-oriented



Universities, CNR
INFN
Universities, CNR

Universities


CNR
INFN

INFN, CNR



Netherlands

1993
2003

Distributed ASCI Supercomputer (DAS)
VL-e, LOFAR

DutchGrid



Supercomputing
Science/application oriented
Infrastructure-oriented

Universities and Labs

Poland

Since 2002

Progress, ClusteriX, LDAP, SGIGrid, Vlab

Infrastructure-oriented

Research Labs, Universities

Spain

2000
2002

2004
2004-2007



First Grid Projects
Thematic Networks and Working Groups

First public demo of a Spanish Grid


e-Science in National Research Programme
IRISGrid starts

Application-oriented

Infrastructure-oriented



Individual researchers

Research Labs



Sweden

Since 1990
2002
2003

HPC PDC at KTH
Swedish HPC Metacenter
SWEgrid

Supercomputing

Applications, e-Science



Universities and Research Labs

UK

Since 2001

e-Science initiative
Core e-Science Programme

Science/application oriented

Universities and Research Labs



Yüklə 0,69 Mb.

Dostları ilə paylaş:
1   2   3   4   5   6   7   8   9   ...   29




Verilənlər bazası müəlliflik hüququ ilə müdafiə olunur ©muhaz.org 2024
rəhbərliyinə müraciət

gir | qeydiyyatdan keç
    Ana səhifə


yükləyin