- Allocations-admins - mailman3.cels.anl.gov

[LCRC Accounts] Yearly Allocation Request from uncertainty-climate
by accounts＠lcrc.anl.gov 13 Sep '10

13 Sep '10

Hello, A yearly allocation for the LCRC cluster has been requested with the following updated information: Submitter/PI: Emil M Constantinescu Project Name: uncertainty-climate Division: MCS Project title: Uncertainty in weather forecasting Associated funding: Other Systems: NFS @ MCS Science: The theoretical part of the project consists in developing uncertainty models for numerical weather and climate prediction systems. Fusion is used to run several scenarios and obtain forecasts and ensemble-based uncertainty information about ambient variables by running weather/climate simulation codes. One objective is to use this information is used in energy management applications such as stochastic optimization control loops that provides optimal operational point for energy systems. A second objective is to develop robust data assimilation methods (i.e., correct model predictions with observations) in order to calibrate weather and climate forecasts for effective operation and planning of energy systems. Project description: The aim is to create very accurate uncertainty estimates of the weather and climate forecasts to be used for optimal management of energy systems such as wind- and solar-power generation, building management and power grid planning. The purpose of using Fusion is to execute several scenarios and obtain forecast and uncertainty information about ambient variables by running high resolution weather/climate simulation codes (such as WRF). This information is used in a stochastic optimization control loop that provides an optimal operational point for energy systems. For the practical part of the project, we will develop and enhance an uncertainty quantification framework (C or FORTRAN, MPI, XML, Python) to run WRF in a realistic fashion and illustrate theoretical findings and provide realistic results. We consider running several WRF instances (each on one node) at the same time. Partial results are collected on a separate node and processed before they are sent back to each specific instance. The instances will continue to run and produce new results. In terms of parallelism, this application framework is embarrassingly parallel since each instance run independently. In FY11, we need to run data assimilation experiments with high resolution and obtain more samples in order to achieve a better level of accuracy. The improved weather forecast accuracy simulations allow us to find a more robust operational point for the operation of energy systems (wind power for electricity generation and building heating and cooling). Project URL: Current FY Hours Used: undetermined amount New FY Requested allocation: 400000 Justification: Note: I used the FY10 form to make updates for renewal and saved them. After that I realized that i need to use a different form... so now renewal for FY11 and FY10 are the same. Sorry about that. Thank You, The LCRC Accounts System

1 0

[LCRC Accounts] Yearly Allocation Request from Atmos_Chem
by accounts＠lcrc.anl.gov 13 Sep '10

13 Sep '10

Hello, A yearly allocation for the LCRC cluster has been requested with the following updated information: Submitter/PI: Beth Drewniak Project Name: Atmos_Chem Division: ER Project title: Process Scale Modeling of Atmospheric Chemistry and Transport Associated funding: DOE,BLM,NASA Other Systems: NERSC Science: The scientific aim of the proposed modeling activity is to further the understanding of the chemistry and distribution of aerosols over certain regional domains and calculate the heating rates as a result of increased aerosol loading primarily over India. This modeling effort supports the Ganges Valley Aerosol Project (GVAX) a one-year field study funded by DOE BER for the years 2011-2012. Project description: 1. What are the sources and sinks of the aerosols comprising the “Asian brown cloud,” and what effects does it have on clear-sky radiation fields, as well as on cloud micro- and macrophysical properties? What effect does the diurnal PBL cycle in the Ganges valley have on the atmospheric distribution of aerosols? Recent studies indicate that nearly 50 90% of the measured black carbon aerosols that give rise to brown clouds over the Indian Ocean are from biomass burning (residential cooking and agricultural burning), but little quantitative information is available regarding how properties of these aerosols — such as light scattering and absorption and droplet nucleating properties — evolve as a function of time near the source Ganges valley region. Very little is also known about the vertical structure of the distribution of aerosols and the fraction of the aerosols that are available for regional- and global-scale transport from the source region in the Ganges valley. The one-year long AMF deployment is uniquely positioned to make continuous, high frequency measurements through the atmospheric column of aerosol optical properties, radiative transfer and cloud properties. In addition, to bring additional focus on resolving some of the more regional and chemical composition relate d questions inherent in this objective, we propose to conduct a winter intensive field operation starting from January 15th 2012 to March 15th 2012, at a time where aerosol loading over this part of India is known to be the highest. During this intensive, the G 1 will add the additional capability of generating in situ measurements and vertical profiles within the PBL and above to generate estimates of the amounts of aerosols that are transported into the mid troposphere and to regional scales. Additional resources such as MAOS, a surface site below in the valley from the location of AMF and a number of supporting surface sites operated by ISRO/IISc will be available. A more detailed description of this intensive is provided below. 2. What effect do increasing aerosols in the Ganges Valley have on the ISM? What effect do aerosols have on shortwave radiative forcing, mid-tropospheric heating and convective activity over continental India? With the AMF, we will be in a position to make continuous, extended estimates of heating rates within cloud over continental India by measuring cloud properties, drop sizes, and LWC and making complete radiative balance measurements of clouds. The uncertainty in parameterization of deep convective activity in climate models has a profound impact on the calculated climate sensitivity to rising CO2. Deep convective activity is common during much of the summer and pre-monsoon period; the initial onset of monsoon over the southern tip of India is initiated by land-sea circulation and then enhanced and driven by deep convective activity over the northern Indian plains. Specific to this study would be an opportunity to measure the influence of aerosols on deep convection. This highly complex, little understood problem in modeling deep convection is complementary to the elevated heat pump hypothesis. Measurement of aerosols from surface supersites to identify their chemical and phys ical characteristics will provide key constraints to the estimates of radiative forcing heating rates from these aerosols in the PBL during the winter and early spring intensive as discussed above. A second intensive operating period, during the pre-monsoon, is planned for the period extending from April 15th 2011 to June 15th 2011 for the purpose of coupling the surface measurements made by the AMF with measurements of in situ cloud properties and aerosol measurements made the IISc/ISRO Beechcraft-20, as well as ongoing Indian activities (Aerosol Radiative Forcing over India; ARFI) to obtain a regional footprint. This would be the first ever such effort to synthesize the various measurements into a combined data product over this region that would be available for model validation and evaluating current cloud level aerosol heating hypotheses. We plan to use a combination of models for the project, including the CCSM4.0, GEOS-CHEM for aerosol transport and IMAGE model for aerosol cloud interactions. Project URL: Current FY Hours Used: undetermined amount New FY Requested allocation: 50000 Justification: Thank You, The LCRC Accounts System

1 0

[LCRC Accounts] Project Allocation Request
by accounts＠lcrc.anl.gov 13 Sep '10

13 Sep '10

Hello, A change in allocation has been requested: Requester: svasquez (Silverio Vasquez) Project: ESMF Title: Earth System Modeling Framework Description: The goal of the Earth System Modeling Framework project is to build high performance, flexible software infrastructure to increase ease of use, performance portability, interoperability, and reuse in climate, numerical weather prediction, data assimilation, and other Earth science applications. The framework is intended for use by individual researchers as well as major operational and research centers. ESMF products include an API specification, a reference implementation accompanied by extensive documentation, and an exhaustive regression test suite that includes 4000+ unit tests, system tests, and examples. The ESMF software consists of a superstructure for coupling and exchanging data between models components (e.g., atmosphere, ocean) and an infrastructure consisting of 1) data structures for representing grids and fields and 2) an optimized, portable set of low-level utilities. The data constructs and low-level utilities can be used by the coupling superstructure and may also be used separately to compose scientific applications. ESMF uses a mixture of C++ and F90. Most of the low level classes are implemented in C++ and have Fortran bindings. We have worked out a detailed interoperability specification which allows F90/C to interact on multiple platforms. Communication is handled using a combination of pthreads and MPI which allows ESMF to support both pure-distributed, pure-shared, and hybrid parallel programming models. In FY11, we will continue to use the extensive compiler set on Fusion as a development platform and to run the ESMF regression test suite. Current: undetermined amount Justification: The ESMF code is regression tested 6 times weekly on the intel, g95 and gfortran compilers. Each test run uses an average of 40 core hours. This calculates to 720 hours a week. Since we expect to add more software and tests during the year, we expect this amount to increase to close to 1000 hours weekly. Therefore we are requesting 52000 hours for regression testing and an additional 8000 hours for ESMF code development and debugging. Requested: 60000 A specific reason has been given: The ESMF code is regression tested 6 times weekly on the intel, g95 and gfortran compilers. Each test run uses an average of 40 core hours. This calculates to 720 hours a week. Since we expect to add more software and tests during the year, we expect this amount to increase to close to 1000 hours weekly. Therefore we are requesting 52000 hours for regression testing and an additional 8000 hours for ESMF code development and debugging. This needs to be approved and the final allocation amount decided upon. Thank You, The LCRC Accounts System

1 0

[LCRC Accounts] Yearly Allocation Request from ESMF
by accounts＠lcrc.anl.gov 13 Sep '10

13 Sep '10

Hello, A yearly allocation for the LCRC cluster has been requested with the following updated information: Submitter/PI: Silverio Vasquez Project Name: ESMF Division: MCS Project title: Earth System Modeling Framework Associated funding: DOD, NASA, NSF, NOAA Other Systems: external: NERSC, NCAR, ORNL, NCSA, NASA Science: The Earth System Modeling Framework (ESMF) enables the coupling of model components in high performance Earth system modeling codes. ESMF component interfaces have been implemented in most of the prominent climate and weather codes in the U.S.: CCSM, WRF, NASA GEOS-5, EMC GFS/NEMS, NRL COAMPS and the GFDL MOM4 ocean. In general, componentization with ESMF has been implemented at the level of major physical domains, where simulated interactions require inter-component data communications (e.g. atmosphere, ocean). This implementation is compatible with codes such as WRF that impose their own interoperability conventions at the level of individual physics parameterizations. The codes listed above, plus the full GFDL modeling system under FMS, can now operate within a component-based, hierarchical architecture in which large-scale components can be called as subroutines. This was not the case when ESMF began. The current level of ESMF adoption represents a major and necessary step towards enabling U.S. centers to exchange large-scale components, and is an excellent basis for further advances in code interoperability. The recent technical focus has been on continued development of ESMF grid classes and a regridding capability. These are built on a base of two flexible data representations that were constructed in previous years: an unstructured finite element mesh library and a general array class designed to support fields on logically rectangular grids. At this point ESMF can produce non-conservative interpolation weights, in parallel or serial, 2D or 3D, for grids that can be represented by triangles or quadrilaterals. Users can also read in external interpolation weights, from a package such as SCRIP, if they need an interpolation option ESMF does not provide. Project description: The goal of the Earth System Modeling Framework project is to build high performance, flexible software infrastructure to increase ease of use, performance portability, interoperability, and reuse in climate, numerical weather prediction, data assimilation, and other Earth science applications. The framework is intended for use by individual researchers as well as major operational and research centers. ESMF products include an API specification, a reference implementation accompanied by extensive documentation, and an exhaustive regression test suite that includes 4000+ unit tests, system tests, and examples. The ESMF software consists of a superstructure for coupling and exchanging data between models components (e.g., atmosphere, ocean) and an infrastructure consisting of 1) data structures for representing grids and fields and 2) an optimized, portable set of low-level utilities. The data constructs and low-level utilities can be used by the coupling superstructure and may also be used separately to compose scientific applications. ESMF uses a mixture of C++ and F90. Most of the low level classes are implemented in C++ and have Fortran bindings. We have worked out a detailed interoperability specification which allows F90/C to interact on multiple platforms. Communication is handled using a combination of pthreads and MPI which allows ESMF to support both pure-distributed, pure-shared, and hybrid parallel programming models. In FY11, we will continue to use the extensive compiler set on Fusion as a development platform and to run the ESMF regression test suite. Project URL: http://www.earthsystemmodeling.org Current FY Hours Used: undetermined amount New FY Requested allocation: 60000 Justification: The ESMF code is regression tested 6 times weekly on the intel, g95 and gfortran compilers. Each test run uses an average of 40 core hours. This calculates to 720 hours a week. Since we expect to add more software and tests during the year, we expect this amount to increase to close to 1000 hours weekly. Therefore we are requesting 52000 hours for regression testing and an additional 8000 hours for ESMF code development and debugging. Thank You, The LCRC Accounts System

1 0

[LCRC Accounts] Yearly Allocation Request from Climate_Emulator
by accounts＠lcrc.anl.gov 13 Sep '10

13 Sep '10

Hello, A yearly allocation for the LCRC cluster has been requested with the following updated information: Submitter/PI: Robert Jacob Project Name: Climate_Emulator Division: MCS Project title: Building a climate emulator for use in energy policy analysis Associated funding: NSF, UC-Argonne Seed Grant. Other Systems: pending application with NCAR Science: State-of-the-art general circulation models (GCMs) provide the best estimates of future climate change. While they are valuable tools for assessing impacts of climate change for a given scenario of future greenhouse gas emissions, they cannot be used in iterative economic models since a single multi-century simulation can take weeks to months to run and the iterative models require hundreds of thousands of iterations. For this reason, we aim to develop a statistical emulator that will accurately and efficiently reproduce GCM forecasts of regional climate change for an arbitrary trajectory of CO2 concentrations by using a library of pre-computed GCM output. Library based statistical emulation has not been possible to this date because there is no consistent and comprehensive library of climate model runs corresponding to a range of scenarios of future CO2 concentrations. Computing time on Fusion will enable us to build this unique resource by performing a series of s imulations using the Community Climate System Model (CCSM3). Since library-based statistical emulation is currently not possible, the most common approach to emulating regional climate uses linear pattern scaling, in which changes in regional temperature and precipitation are assumed to be linearly proportional to the change in global mean temperature (GMT). That is, regional climate is effectively assumed to depend solely on GMT, and not on the rate or duration of warming. Our preliminary analysis of CCSM3 output from the Climate Model Intercomparison Project (CMIP3) archive shows that this assumption is not valid for temperature projections over time scales of multiple centuries and breaks down for projections of precipitation even over decadal time scales. The CMIP3 archive, however, is not sufficient for full statistical emulation, since it is both restricted in number of runs and inconsistent in its treatment of regional climate forcing. The carefully designed library we propose here will be used to develop and test a number of alternative approaches for emulating climate model responses, including (i) nearest neighbor interpolation, (ii) a variant of pattern scaling that accounts for the rate of warming, and (iii) principal component analysis (PCA). The intended outcome of this research is a new climate emulator that will aid in the evaluation of energy policies for greenhouse gas mitigation. We anticipate that there will be a wide interest in a user-friendly, computationally tractable emulation tool from the policy, economics, and climate communities. The library itself will also be made public for further use by the community Project description: Building a new climate library is a prerequisite for developing and testing climate emulation techniques due to the limitations of existing climate libraries. We are applying here for computing time to build a library of CCSM3 runs with a large number of scenarios, multiple realizations of each scenario, and multi-century simulations. This would represent a substantial advance over the library of CCSM3 runs in the CMIP3 archive, which contains only three scenarios for future CO2 concentration growth, does not study CO2 growth beyond 2100, and considers different regional sulphate aerosol emissions for each scenario, complicating any understanding of regional climate patterns expected in a warming world. By addressing these limitations, the consistent and comprehensive climate library we propose here will make possible new emulation methods. CCSM's atmosphere component is the Community Atmosphere Model. CAM currently supports three different numerical cores for solving the basic equations: a spectral transform method, a Semi-Lagrangian scheme, and a finite-volume scheme. The ocean component is a finite-difference model called the Parallel Ocean Program and was developed at Los Alamos. CCSM's sea ice and land surface models are each among the most complex and complete models in existence. CCSM requires Fortran90 and C compilers and the MPI and NetCDF libraries to run. In order to make the project computationally tractable we will use the low-resolution (T31) version of CCSM3. Scenarios of future CO2 concentration will follow a tree design. We will consider three scenarios between 2010 and 2050, branch to six scenarios between 2050 and 2100, and branch again to 12 scenarios between 2100 and 2200. Finally, we will consider six scenarios between 2200 and 2500, for the study of very long-term effects. The tree design allows significant savings to be made in computing resources as multiple scenarios can use the same restart files, and allows examination of hysteresis (i.e., the importance of CO2 history) for scenarios with overlapping branches. Multiple realizations of each scenario (obtained by running the model with different initial conditions) are required to allow robust separation of true signals from internal variability in the climate system. The full library will consist of 10 realizations with historical CO2 forcing between 1870 and 2010, 10 realizations of each scenario between 2010 and 2200, and five realizations for scenarios between 2200 and 2500, when we expect long-term trends to be more pronounced. Finally, we will conduct a detailed exploration on the importance of natural variability by considering a much larger number of realizations (30) for a single 200 year scenario. This yields a total of (140*10) + (40*3*10) + (50*6*10) + (100*12*10) + (300*6*5) + (200*1*(30-10)) = 30,600 model years. We have begun assembling this library of CCSM3 runs using a now expired TeraGrid allocation and a late-FY10 LCRC allocation of 75,000 core hours. These resources have enabled the simulation of 6.500 model years to date. Therefore, we have 24,100 model years remaining. We have sufficient disk space to store the output offsite. Project URL: Current FY Hours Used: undetermined amount New FY Requested allocation: 900000 Justification: Our initial late-FY10 LCRC allocation of 75,000 core hours has allowed us to perform thousands of years of simulations using CCSM3 at T31. This provides us with detailed statistics on the time required to run this model configuration on Fusion. Using 8 nodes (64 processors) takes an average of 35 minutes for a simulation of one model year. Therefore, we are requesting 24,100*64*(35/60) ~= 900,000 CPU hours to complete the full suite of simulations. Several simulations can be run at the same time to scale the project out to the full machine if available. CCSM3 was the recipient of a great deal of tuning and evaluation by combined DOE and NCAR computational scientists so we do not expect further improvements in the per node performance. Thank You, The LCRC Accounts System

1 0

[LCRC Accounts] Yearly Allocation Request from NuclearStructure
by accounts＠lcrc.anl.gov 13 Sep '10

13 Sep '10

Hello, A yearly allocation for the LCRC cluster has been requested with the following updated information: Submitter/PI: Richard Chasman Project Name: NuclearStructure Division: PHY Project title: Optimization of many-body wavefunctions Associated funding: DOE/Office of Nuclear Physics Other Systems: desktop Science: The purpose of this work is to obtain a good description of low-lying nuclear states, going beyond a mean-field description. It becomes necessary to go beyond the mean-field approximation when the nucleus of interest is transitional in some sense, e.g. transitional between a spherical shape and a deformed shape. The need to go beyond a mean-field description is particularly important for excited states. Project description: I am developing a code to handle neutron-proton pairing together with deformation. When the code is developed, I will need time on fusion for running. Project URL: Current FY Hours Used: undetermined amount New FY Requested allocation: 5000 Justification: Thank You, The LCRC Accounts System

1 0

[LCRC Accounts] Yearly Allocation Request from cfd_enginemodeling
by accounts＠lcrc.anl.gov 10 Sep '10

10 Sep '10

Hello, A yearly allocation for the LCRC cluster has been requested with the following updated information: Submitter/PI: Sibendu Som Project Name: cfd_enginemodeling Division: ES Project title: Computational Fluid Dynamics Modeling of Petrodiesel and Biodiesel for Compression Ignition engine applications Associated funding: LDRD Project Other Systems: Science: This project will use an Eulerian-Lagrangian Computation Fluid Dynamics (CFD) software called CONVERGE to model various diesel engine processes like internal nozzle flow, spray behavior, combustion and emission processes. The goal is to capture the differences between diesel and biodiesel fuels from a CFD standpoint since computer based modeling is more cost effective than performing full scale laboratory experiments. Project description: 1) Finite volume schemes 2) LES and RANS turbulence models used 3) Detailed chemical kinetic models for diesel and biodiesel fuels 4) CONVERGE software is being used - 24 licenses available 5) Scales well up-to 24 processors 6) Detailed spray models used for different spray processes Project URL: http://www.transportation.anl.gov/ Current FY Hours Used: undetermined amount New FY Requested allocation: 75000 Justification: Thank You, The LCRC Accounts System

1 0

[LCRC Accounts] Project Request: RadiationDamage
by accounts＠lcrc.anl.gov 10 Sep '10

10 Sep '10

Hello, A new project on the LCRC cluster has been requested. Please forward the information on to the LCRC Allocation sub-committee. Applicant's name: Dilpuneet Aidhy Applicant's institution: ANL Applicant's division: MSD Project Name: RadiationDamage Project title: Atomistically capturing radiation-damage phenomenon to elucidate microstructure evolution in nuclear materials Associated funding: EFRC - Center for Microstructure Science of Nuclear Fuel Other Systems: None Science: The scientific challenge to accurately predict the behavior and lifetime of the nuclear fuel lies in informing, the currently employed empirically-derived engineering-based continuum level models, by comprehensive atomic and electronic level understanding of radiation damage phenomenon, such as, point defect clustering, dislocation dynamics, void migration, fission-gas release, bubble formation, swelling, etc. Computationally, these extreme length scale methods can be bridged via mesoscale methods, such as, phase field and kinetic Monte Carlo, which are technically capable in providing a mathematical framework for the continuum-level methods from the atomistically-calculated input parameters, such as defect mobilities. In this project, we envision a bottom-up approach, where we uncover key mechanisms of the radiation-induced atomic level phenomenon that can be systematically fed into mesoscale methods using molecular dynamics (MD) and density functio nal theory (DFT) calculations. Objectives: In this proposal, we envision to incorporate stress and temperature gradient effects to our existing point defect models. In particular, the project will elucidate: • thermo-migration of voids in BCC Mo and UO2 as model metallic and oxide fuel materials • interaction of point defect with GBs, voids and line dislocations • saturation of GBs with point defects from source/sink strength viewpoint Project description: Computational Methods: Molecular dynamics (MD) simulation Density Functional Theory (DFT) Phase-Field (PF) Methods Programming model: MD : using the HELL code and the LAAMPS code DFT: VASP Phase-field: home grown code Technical details: In preliminary work, we have the MD approach to UO2, MgO and Mo to understand kinetic evolution of irradiation induced point defects (or Frenkel pairs, FP). Using MD simulation, we elucidated clustering mechanism of oxygen FPs in UO2, where we found that, in the presence of FPs on both uranium and oxygen sublattices, oxygen vacancies formed Schottky clusters with uranium vacancies. ([1] Aidhy et al. PRB 80 104107 (2009)) And, the sequestered oxygen interstitials combined together to form cuboctahedral clusters , which have been experimentally observed before in UO2+x systems. ([2] Bevan et al., J. Sol. St. Chem. 61 1 (1986)) In contrast, individually present oxygen FPs recombined and annihilated completely, thus healing the material fully. We observed similar FP behavior in MgO, where defects present simultaneously on both sublattices (i.e., Mg and O) formed clusters on interstitial sites with rocksalt lattice symmetry ([3] Aidhy et. al., Scri. Mat. 60 691 (2009)). These sim ulations not only reproduced the previously observed clusters but also elucidated the underlying clustering mechanisms. We also elucidated the effect of grain boundaries (GB) on point defect evolution (BCC Mo as model materials) by linking the sink/source strength of GBs calculated from MD simulations to the mesoscale level rate-theory equations to predict dominant defect annihilation mechanisms.([4,5], [4] Millett et. al., IJMR 100 4 (2009), [5] Aidhy et. al., unpublished) With these simulations, we have built a firm methodology to which we can now include more intricate phenomenon such as migration of voids, bubble formation, etc. Performance on Fusion: For test calculation of MD simulation: 1000 ps simulation of 12000 atoms on 1 node cost ~30 hrs. Size of calculations: For MD: between 10000 atoms to 400000 atoms for few nanoseconds (both ionic and metallic systems) These are very standard MD calculations used widely. Expected number of project members: currently there is only one post-doc. In future, we will add another member. Project URL: http://www.inl.gov/efrc Requested allocation: 150000 Justification: The requester has used 6179 hours of their initial startup project. In addition to approving an initial amount, please specify a Category and Subcategory for this project. For a list of the current selection of approved categories, please see: https://wiki.lcrc.anl.gov/wiki/Processes/Categories Once the Allocation committee has approved the project, please go to the Project Management page to create it: https://accounts.lcrc.anl.gov/projects.php Thank You, The LCRC Accounts System

1 0

[LCRC Accounts] Yearly Allocation Request from Comp_Nanocatalysis
by accounts＠lcrc.anl.gov 08 Sep '10

08 Sep '10

Hello, A yearly allocation for the LCRC cluster has been requested with the following updated information: Submitter/PI: Julius Jellinek Project Name: Comp_Nanocatalysis Division: CSE Project title: Theoretical/Computational Studies of the Fundamentals of One-Component and Alloy Metal Based Nanocatalysis Relevant to Alternative Sources of Energy Associated funding: BES/DOE Other Systems: NERSC, ~400000 core-hours Science: The central objective of this project is to characterize the chemical reactivity and catalytic functionality of one-component and alloy metal nanoparticles as a function of their size, structure and composition, as well as the material and morphology of their supports. The reactions considered will be those of relevance to alternative sources of energy. Project description: The subject of this project is structural, electronic, and chemical reactivity properties of one-component and alloy metal nanocatalysts, both gas phase and deposited on oxide supports. The complexity of these systems stems from the fact that even for a one-component catalytic cluster or nanoparticle the number of different possible structural forms (isomers) is very large. For the case of alloys particles, the complexity increases even further as each isomeric form gives rise to a large number of secondary structures (homotops) that differ by the placement of the different types of atoms between the sites of that isomer. As a consequence, the configuration space that has to be searched is immense. We are using the parallelized version of the NWChem package installed on Fusion. The current work focuses on pure Pt and Mo and mixed Pt/Mo clusters. So far we have covered the size range from 2 to 13 atoms and made some test runs on a 55 atom cluster. The cas es of low or no symmetry are particularly problematic. Even for a system as small as a 13-atom Pt cluster we repeatedly encountered references to “segmentation error”. The largest number of processors we have used so far is 128. The reason for this is not lack of scalability in performance, but simply the desire to save the very limited resources we have been allocated so far. Our performance tests indicate that we could use profitably hundreds or even thousands of processors in runs on systems with about 50-100 atoms. Our plans are to extend our studies to this size range and beyond. In addition, we plan to add in the computations also the atomistic representation of the supports. The expected number of project members is 4. Project URL: http://www.anl.gov/catalysis-science/index.html Current FY Hours Used: undetermined amount New FY Requested allocation: 990000 Justification: The allocation request is based on estimates made on Fusion and on Carver at NERSC. Tests on Pt_12Mo cluster with the Stuttgart pseudopotential and a basis set of 38 contracted functions per atom performed with 32, 64, and 128 processors indicate full parallel scaling efficiency. Tests on a 55 atom cluster with 128 and 200 processors give a comparable scaling. Thank You, The LCRC Accounts System

1 0

[LCRC Accounts] Yearly Allocation Request from Lattice-QCD
by accounts＠lcrc.anl.gov 08 Sep '10

08 Sep '10

Hello, A yearly allocation for the LCRC cluster has been requested with the following updated information: Submitter/PI: Donald Sinclair Project Name: Lattice-QCD Division: HEP Project title: Lattice simulations of Conformal and Walking Technicolor. Associated funding: DOE, NSF Other Systems: NERSC Cray XT4, "Franklin" NERSC Cluster, "Carver/Magellan" NICS Cray XT5, "Kraken" Science: We are using our simulation methods developed for lattice QCD to study extensions of the standard model of High Energy Physics, in which the Higgs sector is strongly interacting, and the Higgs fields are composites. We are particularly interested in QCD-like theories whose pions play the role of the Higgs field in giving masses to the W and Z gauge bosons. Such theories are called Technicolor theories. We are particularly interested in Walking Technicolor theories, whose slowly evolving couplings evade some of the technical difficulties associated with the phenomenology of Technicolor theories and Extended Technicolor theories, by making a wide range of length/momentum scales natural. These studies are expected to become more important as we enter the LHC era. Our initial studies are of the thermodynamics of such theories since these provide a simple way of measuring the scales of confinement and of chiral symmetry breaking. Our previous work on N_t=4,6,8 lattices (for 2 flavours) has indicated that larger N_t's are needed to access continuum physics. For 3 flavours our work on N_t=4,6 lattices also indicates the need for larger N_t's. Project description: We perform simulations to evaluate the functional integrals of QCD-like theories formulated on a discrete space-time lattice to enable determination of the non-perturbative aspects these theories. These include the properties of these theories at non-zero temperature including the scales of confinement and chiral symmetry breaking, and such zero temperature properties as spectra, decay constants and the running of the gauge coupling constant. We are particularly interested in those theories where the coupling constant evolves particularly slowly, since these are candidate 'Walking Technicolor' theories. Related to these are theories with an infrared fixed point. Of special interest is to differentiate these two different types of behaviour. We are performing simulations of QCD-like theories which are models for Walking or Conformal Technicolor. We have been studying theories which are essentially QCD but with colour-sextet rather than colour-triplet quarks. We hope to measure the running of the QCD coupling constant. For 2 or 3 flavours, 2-loop perturbation theory predicts an infrared fixed point. For 2 flavours, it is possible that a chiral condensate forms before this fixed point is reached. If so the fixed point is avoided, the theory is confining and chiral symmetry breaks spontaneously. However, there is a region where the coupling constant evolves very slowly. These are the properties required for a walking technicolor theory. Simulations we have performed so far at finite temperature suggest that this theory does indeed walk. This is an ongoing program at NERSC. We wish to run small-lattice simulations of the 3-flavour theory, which we expect to have conformal, rather than walking behaviour on Fusion, for comparison. If we are correct, the properties of the 2 and 3 flavour theories should look rather different. Our sextet quark codes are based on our earlier triplet quark codes and use the RHMC simulation method. The Rational Hybrid Monte Carlo (RHMC) is a stochastic molecular dynamics algorithm. The functional integral of QCD is written as a partition function of a classical field theory evolving in a fictitious time. The determinant of the Dirac operator raised to a fractional power is calculated by introducing bosonic fields (pseudofermions), and sandwiching this Dirac operator raised to minus said fractional power between them. This fractional power of the Dirac operator is approximated to machine accuracy by a rational approximation. After defining this this theory on a discrete space-time lattice, the inversions required by the partial-fraction expansion of the rational approximation are performed using Krylov space methods, in particular a multi-shift extension of the conjugate gradient algorithm. A global Metropolis Monte-Carlo accept/reject step applied at the end of each t rajectory removes discretization errors introduced by the numerical integration of these stochastic equations of motion. We parallelize the code by assigning a fixed number of adjacent lattice sites to each MPI task. Network bandwidth ultimately limits how small a chunk of the lattice can be assigned to each task. Benchmarks of this code on Fusion for the 24^3*12 lattice planned for some of our FY2011 runs give: 24 cores = 46 Gflops 48 cores = 95 Gflops 72 cores = 148 Gflops 96 cores = 220 Gflops 144 cores = 369 Gflops 288 cores = 784 Gflops We plan to run the 24^3*12 lattice jobs on 288 cores (36 nodes). Our smaller lattice 16^3*8 runs would be run on 128 cores (16 nodes). Project URL: http://www.hep.anl.gov/dks Current FY Hours Used: undetermined amount New FY Requested allocation: 2500000 Justification: This allocation request is based on having one 288 core job or two 128 core jobs running at all times. Our project is expected to use time at NERSC and NICS as well. LCRC has the advantage that it provides facilities for jobs which have medium levels of parallelism such as the 128 and 288 core jobs indicated here, which are discriminated against on larger machines such as Franklin and Hopper(stage 2) at NERSC, and Kraken at NICS. Carver at NERSC provides a similar platform to Fusion, but the large number of users and potential users mean that we cannot rely on it to provide the bulk of our computing at NERSC. Numbers given in the description section indicate that scaling is excellent. We have plans to increase the level of parallelism for this code, but this is probably irrelevant to our Fusion use. For codes which have smaller portions of the lattice on each core, such as our 16^3*8 code, performances of around 3 Gflop/core are sustained. Thank You, The LCRC Accounts System

1 0