Welcome to CSC

1
Welcome to CSC

CSC the Finnish IT center for science
CSC, March 21, 2006 Juha Haataja Director for
Science support Juha.Haataja_at_csc.fi
2
Computational Science Crossing the Disciplines /
CSC, 21.3.2006

• 9.30 Welcome and Introduction / Juha Haataja,
  CSC
• 10.00 Pipettes and CPUs - the Ying and Yang of
  modern biology /
• Gunnar von Heijne, Stockholm Univ.
• 10.30 Software development and nanosciences /
  Karsten Jacobsen, Technical Univ. of Denmark
• 11.00 break
• 11.15 Case studies A Grand Challenge of
  Computational Fluid Dynamics Simulation of
  Turbulence, Timo Siikonen, TKK New
  computational tools for wave modeling, Tomi
  Huttunen, Univ. of Kuopio
• 12.00 lunch break (cafeteria outside the
  auditorium)
• 13.00 Challenges in data management and analysis
  / Heikki Mannila, Univ. of Helsinki
• 13.30 Case studies Case studies in optimizing
  HPC software, Jan Westerholm, Åbo Akademi
  Perspectives in Computational Earth System
  Sciences, Aike Beckmann, Univ. of Helsinki
• 14.15 coffee
• 14.45 Case study Perspectives in the
  Computational Modeling of Biological Systems,
  Ilpo Vattulainen, Tampere Univ. of Technology
• 15.15 Panel, discussion, conclusions
• 16.15 end of seminar

3
Rise of computational science

• From models to decision-making
• What is computational science?
• What is CSC?
• Collaboration on the global scale
• Scientific software development
• Connecting the knowledge

4
What is computational science?

• Algorithms (numerical and non-numerical) and
  modeling and simulation software
• developed to solve science (e.g., biological,
  physical, and social), engineering, and
  humanities problems
• Computer and information science
• develops and optimizes the advanced system
  hardware, software, networking, and data
  management components needed to solve
  computationally demanding problems
• The computing infrastructure
• supports both the science and engineering problem
  solving and the developmental computer and
  information science

5
Findings of the PITAC report

• Computational Science Ensuring America's
  Competitiveness (President's Information
  Technology Advisory Committee, June 2005)
• Computational science is indispensable to the
  solution of complex problems in every sector
• The strategic significance of computational
  science has not been recognized
• Knowledge of computational science needed to
  solve key problems in science and engineering
• Recommendations
• Multidisciplinary collaboration needed
• Create and direct a multi-decade roadmap

6
FT-2 simulation using Elmfire

• Elmfire developed at TKK and VTT during 2000-2005
• Target research on fusion plasma and reactors
• Runs at IBM eServer Cluster 1600 and HP ProLiant
  DL145 cluster supported by CSC
• Needs more than 10 teraflop/s in the future
• J.A. Heikkinen, S.J. Janhunen, T.P. Kiviniemi,
  and F. Ogando, Full f Gyrokinetic Method for
  Particle Simulation of Tokamak Transport,
  submitted to Phys. Rev. E.

Source J.A. Heikkinen (VTT) and T.P. Kiviniemi
(TKK)
7
The supercomputer procurement at CSC

• CSC is currently orgazing a supercomputer
  procurement
• Total budget 10 million euros
• Target for first installation end of 2006
• Benefits
• Unique cross-disciplinary service (few similarly
  cross-disciplinary centers exist)
• Emphasis on expert support in HPC (including code
  optimization and parallelization)
• Connections to existing and future international
  HPC and grid projects (DEISA, EGEE II, NDGF etc.)

8
Drug discovery
Source Antti Poso, Univ. of Kuopio
9
Experimental, theoretical and computational
science
10
Hemodynamics with Elmer

• Cardiovascular diseases are the leading cause of
  deaths in western countries
• Reduced elasticity of arteries results in higher
  risk for several cardiovascular events
• Modeling of blood flow poses a challenging case
  of fluid-structure-interaction
• Modeling and pictures by Esko Järvinen at CSC

Figures Esko Järvinen, CSC
11
(No Transcript)
12
Challenges are increasing in computational science
ocean
land

• multiphysical models ? complex geometries ?
  multiscale modeling ? ....
• efficient computational methods are needed
  (e.g., domain decomposition based solvers)
• data management and analysis becomes more
  challenging
• scientific software development uses increasingly
  advanced techniques

13
Need to collaborate on the global level

• Big changes in scientific research
• resources are distributed and accessed on the
  network
• researchers and resources interact with each
  other
• New possibilities in research
• global collaborative science
• discoveries and innovations
• networks of people
• multidisciplinary, multi-institutional,
  international
• IT for science combines local and global
  strengths
• experts and institutions work together (often ad
  hoc)
• connect software, databases and instruments
• need both scientific and IT expertise

14
(No Transcript)
15
International Review of Research Using HPC in the
UK (December 12, 2005)

• Create a more balanced HPC infrastructure between
  computational technologies and intellectual
  resources
• Strengthen the computational infrastructure
• systematically deploy leading-edge capability
  systems, large-scale capacity computing, and
  resources deployed widely at universities
• support and develope a state-of-the-art
  applications software infrastructure encompassing
  algorithms, data management and analysis,
  visualization, and best- practices software
  engineering
• Develop human resources in HPC
• Bridge disciplines and build a computational
  science community
• increase interactions and foster collaborations
  between disciplinary groups nationally and
  internationally

16
Case Coupled Model Systems

• CSC develops technology intented for coupling
  different kinds of models systems
• Connected to Finnish expertise on modeling
• The CoMS project funded by the Finnish Technology
  Development Agency Tekes

17
Case M-grid, towards a national machine room

• Joint cluster procurement for materials research
• Seven universities, HIP and CSC
• Infrastructure funding from the Academy of
  Finland
• Procudement and management of the clusters
  coordinated by CSC
• Created a grid environment for medium-scale
  computational needs
• First large-scale grid environment in production
  in Finland
• Each partner does what they know best

18
What next?

• Develop competence in computational science
• Support knowledge transfer
• Coordinate efforts
• gt Formulate a long-term roadmap!

19
From information deluge to unmapped seas

• Information deluge
• Each researcher is like an island in a sea of
  information.
• Connecting
• Coordination, cross-disciplinary, IT skills,
  central data archives and computational
  resources, computer networking.

• Grand challenges
• Science is a drop in the sea of the unknown.
• Pioneering
• Grand challenges, data access, international
  services.

20
Seminar on Finnish IT infrastructures for Science
(May 11th at CSC)

• Tutkimuksen haasteita infrastruktuureille
  Miten tieteen tietotekniikka yhdistää osaamisen?
• Gridin tarjoamat mahdollisuudet ja haasteet,
  Prof. Risto Nieminen
• Ympäristötieteiden infrastruktuurit, Prof.
  Markku Kulmala
• Data ja laskenta tutkimuksen tietotekniikan
  haasteet,Prof. Heikki Mannila
• Announcement and registration in April
• Contact Juha.Haataja_at_csc.fi

21
Computational Science Crossing the Disciplines /
CSC, 21.3.2006

• 9.30 Welcome and Introduction / Juha Haataja,
  CSC
• 10.00 Pipettes and CPUs - the Ying and Yang of
  modern biology /
• Gunnar von Heijne, Stockholm Univ.
• 10.30 Software development and nanosciences /
  Karsten Jacobsen, Technical Univ. of Denmark
• 11.00 break
• 11.15 Case studies A Grand Challenge of
  Computational Fluid Dynamics Simulation of
  Turbulence, Timo Siikonen, TKK New
  computational tools for wave modeling, Tomi
  Huttunen, Univ. of Kuopio
• 12.00 lunch break (cafeteria outside the
  auditorium)
• 13.00 Challenges in data management and analysis
  / Heikki Mannila, Univ. of Helsinki
• 13.30 Case studies Case studies in optimizing
  HPC software, Jan Westerholm, Åbo Akademi
  Perspectives in Computational Earth System
  Sciences, Aike Beckmann, Univ. of Helsinki
• 14.15 coffee
• 14.45 Case study Perspectives in the
  Computational Modeling of Biological Systems,
  Ilpo Vattulainen, Tampere Univ. of Technology
• 15.15 Panel, discussion, conclusions
• 16.15 end of seminar