3. Towards Computational
Science Computational
Theoretical
• Evolution of science [1]
Empirical
– From being primarily Empirical to grow an
– Theoretical component and today also include a
– Computational component
• PITAC report 2005
– Defines and emphasizes the advancement of Computational
Science
“Advances in computing and connectivity make it possible
to develop computational models and capture and analyze
unprecedented amounts of experimental and observational
data to address problems previously deemed intractable or
p p y
beyond imagination.”
[1] J. Gray, G. Bell and A. Szalay, Petascale Computational Systems:Balanced CyperInfrastructure in a Data-
Centric World, Report to NSF October 2005
4. Evolution of Computational
Science
S i
• Computational needs continuously
growing
• R
Research communities and projects h
h iti d j t have
become global
• The need to share data and resources
5. Evolution of Computational
Science
• Unified view of
theory, experiments
and numerical
simulation e-Science or
• Advances in
e-Research
e Research
computing and
networking
t ki e-Infrastructure
e Infrastructure
6. Building a Distributed Research
Infrastructure
I f t t
• Sharing resources allows load balancing
• Better use of funding
• Eases collaboration
10. DEISA
• A grid consisting of several European supercomputers,
g g p p p
including CSC's Louhi.
• A European grid project that shares computing
resources for demanding scientific applications
• Yearly applications for computing time (DECI call)
• EU FP6: DEISA May 1st 2004 – April 30th 2008
1st, 30th,
• EU FP7: DEISA2 May 1st, 2008 – April 30th, 2011
• Most DEISA sites are connected via 10 Gb/s lines
• DEISA Global File System - eliminates need for file
staging
13. EGEE
• EGEE processes up to 300,000 j
p p , jobs p day
per y
• more than 80,000 CPU cores
• several tens petabytes of storage
• available 24 hours a day, seven days a week
No. Cores
80000
70000
60000
50000
40000
30000
20000
10000
0
huhti.04
huhti.05
huhti.06
huhti.07
huhti.08
heinä.04
oka.04
tammi.05
heinä.05
oka.05
tammi.06
heinä.06
oka.06
tammi.07
heinä.07
oka.07
tammi.08
lo
lo
lo
lo
15. Grids in Finland?
• CSC is the key p y
y player
• Participation in large European projects
– DEISA
– EGEE
• Influencing the next generation of grid
infrastructures
– PRACE
– EGI
• Coordinates national level grid infrastructure
g
16. Some grid use cases in
Finland
• CERN Large Hadron Collider
– ALICE and CMS experiments
– Computational and storage facilities accessed
through grid
– Finnish resources are hosted at CSC for the
Helsinki Institute of Physics
17. Some grid use cases in
Finland, contd.
• DEISA DECI
– LIPOS - Study of Low Density Lipoproteins
to Prevent Heart Attacks
• http://www.deisa.eu/science/deci/projects2007-2008/Lipos
• DEISA Vi t l C
Virtual Community
it
– Planck satellite first light study
• http://www.esa.int/SPECIALS/Planck/SEM5CMFWNZF_0.html
18. Some grid use cases in
Finland, contd.
• PairsDB updates
– Gaining throughput without impacting local
resources
– http://pairsdb.csc.fi/
20. European Commission
“…for Grids we would like to see
the move towards
long-term sustainable initiatives
less dependent upon
p p
EU-funded project cycles”
Viviane Reding, Commissioner, European Commission, at the
EGEE’06 Conference, September 25, 2006
www.eu-egi.eu 20
21. Evolution
National
European
e-Infrastructure
European
Productive Use
Testbeds Utility
www.eu-egi.eu 21
23. Grids in Europe
www.eu-egi.eu
Each National Grid Initiative
… should be a recognized national body with a single point-
of-contact
… should mobilize national funding and resources
… should ensure the operation of a national e-Infrastructure
p
… should support user communities
… should contribute and adhere to intl. standards and
policies
www.eu-egi.eu
24. EGI
• Ensure pan-European grid coordination
pan European
and integrity
• Ai
Aims f standardization where possible
for t d di ti h ibl
• National Grid Initiatives (NGIs) will operate
National Grid Infrastructures
• EGI will link existing and new NGIs
• Operation starts January 2010
• http://web.eu-egi.eu/
26. PRACE
• Prepares for the creation of a persistent pan-European HPC service
• Requires a periodic renewal of the systems and a continuous upgrade of
the infrastructure
• Aims to provide European researchers with access to petaflop scale
computers and form the top level of the European HPC ecosystem
ecosystem.
• The service will comprise three to five petaflop class supercomputing
centres strengthened by regional and national HPC centres working in
collaboration
• Current phase is undertaking all the necessary technical preparations,
including the evaluation of components, systems and facilities, porting key
scientific applications, and the developing training infrastructure
• Should be in a position to deploy and exploit the first petaflop/s systems in
2009/2010.
• CSC is in charge Dissemination, Outreach and Training
• Funded by the European Commission’s 7th Framework Programme
Commission s Programme.
27. CSC and the Cloud
• Currently working with ”A data intensive
A
customer”
– Close collaboration to find suitable technologies from
Clouds and Grid to meet their needs.
• Enhancing the existing software p
g g products to
better support distributed computing
– Chipster: user friendly DNA microarray analysis
p y y y
software developed by CSC offers an intuitive
graphical user interface to a comprehensive collection
of up-to-date analysis methods
up to date
28. CSC and the Cloud
• Building Internal Expertise
• Following developments
• Keen on involving customers as partners
in projects
• Interested in Customer needs