Designing a cluster for geophysical fluid dynamics applications Göran Broström Dep. of Oceanography, Earth Science Centre, Göteborg University.

Slides:



Advertisements
Similar presentations
Basics of numerical oceanic and coupled modelling Antonio Navarra Istituto Nazionale di Geofisica e Vulcanologia Italy Simon Mason Scripps Institution.
Advertisements

El Nino. ► ► Every three to eight years, a temporary change in the oceanic and atmospheric conditions occurs over the south Pacific Ocean. ► ► This change.
HIRLAM Use of the Hirlam NWP Model at Met Éireann (Irish Meteorological Service) (James Hamilton -- Met Éireann)
Discretizing the Sphere for Multi-Scale Air Quality Simulations using Variable-Resolution Finite-Volume Techniques Martin J. Otte U.S. EPA Robert Walko.
Ocean surface currents function of wind duration, intensity, and fetch; Coriolis effects; continental position; water density and salinity.
Beowulf Supercomputer System Lee, Jung won CS843.
El Niño Southern Oscillation [ENSO] NORMAL: - Easterly trade winds between ± 30° latitude (Coriolis Force) - Sea Surface Height slant to west - Warm basin.
Climatology Lecture 8 Richard Washington Variability of the General Circulation.
Numerical Modeling in Physical Oceanography
Geophysical Fluid Dynamics Laboratory Review June 30 - July 2, 2009 Geophysical Fluid Dynamics Laboratory Review June 30 - July 2, 2009.
Class #16 Monday, October 4, 2010 Class #16: Monday, October 4 Chapter 8 Oceanography and El Niño/La Niña/ENSO 1.
DARGAN M. W. FRIERSON DEPARTMENT OF ATMOSPHERIC SCIENCES DAY 16: 05/20/2010 ATM S 111, Global Warming: Understanding the Forecast.
Extratropical climate. Review of last lecture Mean state: The two basic regions of SST? Which region has stronger rainfall? What is the Walker circulation?
Oceans. Vertical ocean temperature profile Plimsoll line.
Lappeenranta University of Technology June 2015Igor Monastyrnyi HPC2N - High Performance Computing Center North System hardware.
Impacts of El Nino Observations Mechanisms for remote impacts.
Ocean Response to Global Warming William Curry Woods Hole Oceanographic Institution Wallace Stegner Center March 3, 2006.
Climate and Climate Change. Climate Climate is the average weather conditions in an area over a long period of time. Climate is determined by a variety.
Processor Technology John Gordon, Peter Oliver e-Science Centre, RAL October 2002 All details correct at time of writing 09/10/02.
On the Mechanisms of the Late 20 th century sea-surface temperature trends in the Southern Ocean Sergey Kravtsov University of Wisconsin-Milwaukee Department.
Grid for Coupled Ensemble Prediction (GCEP) Keith Haines, William Connolley, Rowan Sutton, Alan Iwi University of Reading, British Antarctic Survey, CCLRC.
Climate Variability and Change: An Overview Leigh Welling Crown of the Continent Research Learning Center Glacier National Park.
Chapter 4 Ecosystems and the Physical Environment
Dynamic Climate An overview of Climate Oscillations.
By Anthony R. Lupo Department of Soil, Environmental, and Atmospheric Science 302 E ABNR Building University of Missouri Columbia, MO
Experience with COSMO MPI/OpenMP hybrid parallelization Matthew Cordery, William Sawyer Swiss National Supercomputing Centre Ulrich Schättler Deutscher.
Issues in Ocean-Atmosphere-Land-Ice Coupling Ocean Integration in Earth System Prediction Capability Data Assimilation University of Maryland September.
Climate--The average weather over years and longer… Chapter 1 frontispiece. Satellite view looking east from Patagonia over southern Argentina and the.
Preliminary Results of Global Climate Simulations With a High- Resolution Atmospheric Model P. B. Duffy, B. Govindasamy, J. Milovich, K. Taylor, S. Thompson,
Climate Change and Global Warming Michael E. Mann Department of Environmental Sciences University of Virginia Symposium on Energy for the 21 st Century.
Water Year Outlook. Long Range Weather Forecast Use a combination of long term predictors –Phase of Pacific Decadal Oscillation (PDO) –Phase of Atlantic.
1 CUTTING-EDGE CLIMATE SCIENCE AND SERVICES Geoff Love.
Operational sub-regional Long-Range Forecasting Unit at RA VI Regional Climate Center – South-East European Virtual Climate Change Center Vladimir Djurdjevic.
ARGONNE NATIONAL LABORATORY Climate Modeling on the Jazz Linux Cluster at ANL John Taylor Mathematics and Computer Science & Environmental Research Divisions.
GEU 0027: Meteorology Lecture 10 Wind: Global Systems.
Course Evaluation Closes June 8th.
First results from the isopycnic ocean carbon cycle model HAMOCC & MICOM/BCM Karen Assmann, Christoph Heinze, Mats Bentsen, Helge Drange Bjerknes Centre.
Regional Oceanography II OEAS 604 Lecture Outline 1)Pacific Ocean circulation 2)Antarctic circulation 3)Climate cycles 4)Atmosphere-ocean coupling Chapters.
The Character of North Atlantic Subtropical Mode Water Potential Vorticity Forcing Otmar Olsina, William Dewar Dept. of Oceanography, Florida State University.
Oceans and climate: an OCCAM perspective Andrew C. Coward 1 Large Scale modelling team James Rennell Division for Ocean Circulation and Climate Southampton.
Ocean Response to Global Warming/Global Change William Curry Woods Hole Oceanographic Institution Environmental Defense May 12, 2005 Possible changes in.
El Niňo. El Nińo: A significant increase in sea surface temperature over the eastern and central equatorial Pacific that occurs at irregular intervals,
Advances in Fundamental Climate Dynamics John M. Wallace et al.
Presented by LCF Climate Science Computational End Station James B. White III (Trey) Scientific Computing National Center for Computational Sciences Oak.
Cluster Computers. Introduction Cluster computing –Standard PCs or workstations connected by a fast network –Good price/performance ratio –Exploit existing.
Climate Change and Global Warming Michael E. Mann Department of Environmental Sciences University of Virginia Waxter Environmental Forum Sweet Briar College.
18th September 2012Clim. of the month, summer Climate of the month July-August 2012 Isabel Andreu-Burillo with the help of Virginie, Fabian, Mingu.
Climate Prediction: Products, Research, Outreach Briefing for NOAA’s Science Advisory Board March 19, 2002 National Weather Service Climate Prediction.
Manchester Computing Supercomputing, Visualization & eScience Zoe Chaplin 11 September 2003 CAS2K3 Comparison of the Unified Model Version 5.3 on Various.
Warm Up: Label all the continents and oceans on your maps. South Southern Ocean North What is this body of water? Gulf of Mexico Caribbean Sea.
Climatic Changes. Standards 4d: Students know the differing Greenhouse conditions on Earth, Mars and Venus; the origins of those conditions; and the climatic.
1 The Interactions of Aerosols, Clouds, and Radiation on the Regional Scale.
Chapter 4 Ecosystems and the Physical Environment.
Brief introduction about “Grid at LNS”
National Taiwan University, Taiwan
Global scale circulation
Overview of the COSMO NWP model
An overview of Climate Oscillations
Global Climate Modelling Department “Physical Climate System” Mojib Latif, Erich Roeckner and Uwe Mikolajewicz.
Ocean Circulation.
Oceanic Influences on Climate
A Mechanistic Model of Mid-Latitude Decadal Climate Variability
Tore Furevik Geophysical Institute, University of Bergen
Seasonal Predictions for South Asia
How will the earth’s temperature change?
2.3.1(iii) Impacts of El Nino
Case Studies in Decadal Climate Predictability
Impacts of El Nino Observations Mechanisms for remote impacts.
El Niño/ La Niña (ENSO) The cycle is the consequence of slow feedbacks in the ocean-atmosphere system acting alongside the strong air-sea interaction processes.
Understanding and forecasting seasonal-to-decadal climate variations
Presentation transcript:

Designing a cluster for geophysical fluid dynamics applications Göran Broström Dep. of Oceanography, Earth Science Centre, Göteborg University.

Our cluster (me and Johan Nilsson, Dep. of Meterology, Stockholm University) Grant from the Knut & Alice Wallenberg foundation (1.4 MSEK) 48 cpu cluster Intel P Ghz 500 Mb 800Mhz Rdram SCI cards Delivered by South Pole Run by NSC (thanks Niclas & Peter)

What we study

Geophysical fluid dynamics Oceanography Meteorology Climate dynamics

Thin fluid layers Large aspect ratio

Highly turbulent Gulf stream: Re~10 12

Large variety of scales Parameterizations are important in geophysical fluid dynamics

Timescales Atmospheric low pressures: 10 days Seasonal/annual cycles: years Ocean eddies:0.1-1 year El Nino: 2-5 years. North Atlantic Oscillation: 5-50 years. Turnovertime of atmophere: 10 years. Anthropogenic forced climate change: 100 years. Turnover time of the ocean: years. Glacial-interglacial timescales: years.

Some examples of atmospheric and oceanic low pressures.

Timescales Atmospheric low pressures: 10 days Seasonal/annual cycles: years Ocean eddies:0.1-1 year El Nino: 2-5 years. North Atlantic Oscillation: 5-50 years. Turnovertime of atmophere: 10 years. Anthropogenic forced climate change: 100 years. Turnover time of the ocean: years. Glacial-interglacial timescales: years.

Normal state

Initial ENSO state

The ENSO state

Timescales Atmospheric low pressures: 10 days Seasonal/annual cycles: years Ocean eddies:0.1-1 year El Nino: 2-5 years. North Atlantic Oscillation: 5-50 years. Turnovertime of atmophere: 10 years. Anthropogenic forced climate change: 100 years. Turnover time of the ocean: years. Glacial-interglacial timescales: years.

Positive NAO phase Negative NAO phase

Positive NAO phase Negative NAO phase

Timescales Atmospheric low pressures: 10 days Seasonal/annual cycles: years Ocean eddies:0.1-1 year El Nino: 2-5 years. North Atlantic Oscillation: 5-50 years. Turnovertime of atmophere: 10 years. Anthropogenic forced climate change: 100 years. Turnover time of the ocean: years. Glacial-interglacial timescales: years.

Temperature in the North Atlantic

Timescales Atmospheric low pressures: 10 days Seasonal/annual cycles: years Ocean eddies:0.1-1 year El Nino: 2-5 years. North Atlantic Oscillation: 5-50 years. Turnovertime of atmophere: 10 years. Anthropogenic forced climate change: 100 years. Turnover time of the ocean: years. Glacial-interglacial timescales: years.

Ice coverage, sea level

What model will we use?

MIT General circulation model

General fluid dynamics solver Atmospheric and ocean physics Sophisticated mixing schemes Biogeochemical modules Efficient solvers Sophisticated coordinate system Automatic adjoint schemes Data assimilation routines Finite difference scheme F77 code Portable

MIT General circulation model Spherical coordinates“Cubed sphere”

MIT General circulation model General fluid dynamics solver Atmospheric and ocean physics Sophisticated mixing schemes Biogeochemical modules Efficient solvers Sophisticated coordinate system Automatic adjoint schemes Data assimilation routines Finite difference scheme F77 code Portable

MIT General circulation model

Some computational aspects

Some tests in INGVAR (32 AMD 900 Mhz cluster)

Experiments with 60*60*20 grid points

Experiments with 120*120*20 grid points

MM5 Regional atmospheric model

Choosing cpu’s, motherboard, memory, connections

Specfp (swim)

Run time on different nodes

Choosing interconnection (requires a cluster to test) Based on earlier experience we use SCI from Dolphinics (SCALI)

Our choice Named Otto SCI cards P GHz (single cpus) 800 Mhz Rdram (500 Mb) Intel motherboards (the only available) 48 nodes NSC (nicely in the shadow of Monolith)

Otto (P GHz)

Scaling Otto (P GHz) Ingvar (AMD 900 MHz)

Why do we get this kind of results?

Time spent on different “subroutines” 60*60*20120*120*20

Relative time Otto/Ingvar

Some tests on other machines INGVAR: 32 node, AMD 900 MHz, SCI Idefix: 16 node, Dual PIII 1000 MHz, SCI SGI 3800: 96 Proc. 500 MHz Otto: 48 node, P Mhz, SCI ? MIT, LCS: 32 node, P Mhz, MYRINET

Comparing different system (120*120*20 gridpoints)

Comparing different system (60*60*20 gridpoints)

SCI or Myrinet? 120*120*20 gridpoints

SCI or Myrinet? 120*120*20 gridpoints (60*60*20 gripoints) (ooops, I used the ifc Compiler for these tests)

SCI or Myrinet? 120*120*20 gridpoints (60*60*20 gripoints) (ooops, I used the ifc Compiler for these tests) (1066Mhz rdram?)

SCI or Myrinet? (time spent in pressure calc.) 120*120*20 gridpoints (60*60*20 gripoints) (ooops, I used the ifc Compiler for these tests) (1066Mhz rdram?)

Conclusions Linux clusters are useful in computational geophysical fluid dynamics!! SCI cards are necessary for parallel runs >10 nodes. For efficient parallelization: >50*50*20 grid points per node! Few users - great for development. Memory limitations, for 48 proc. a’ 500 Mb, 1200*1200*30 grid points is maximum (eddy resolving North Atlantic, Baltic Sea). For applications similar as ours, go for SCI cards + cpu with fast memory bus and fast memory!!

Experiment with low resolution (eddies are parameterized)

Thanks for your attention