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

[LCRC Accounts] Project Request: HSCD
by accounts＠lcrc.anl.gov 08 Sep '10

08 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: Shashi Aithal Applicant's institution: ANL Applicant's division: MCS Project Name: HSCD Project title: High Speed Combustion and Detonation simulations Associated funding: DoE/ASCR Other Systems: Ranger, BG/P (ALCF) Science: The primary goal of this project is to develop a fundamental understanding of the physical processes that bring about a transition from deflagration (slowly moving laminar flames) to detonation (fast moving turbulent flames) in combusting mixtures of fuels and oxygen. The key objective of this work is to conduct first-principles Direct Numerical Simulations for predicting Deflagration-to-Detonation transition (DDT) in reacting turbulent mixtures. Project description: DDT is inherently a multi-physics/multi-scale phenomena involving a close coupling of multiple physical and chemical processes, occurring over a wide range of spatial and temporal scales. Modeling and simulation of this phenomenon requires a detailed description of the various underlying physical and chemical processes such as chemical kinetics and energy release, heat conduction, species diffusion and interaction of turbulent shocks with boundary layers. A compressible reactive flow Navier-Stokes solver named High-Speed Combustion and Detonation (HSCD) code will be used for the simulation of the DDT phenomena. The HSCD code is a distributed memory parallel adaptive mesh refinement (AMR) reactive flow Navier-Stokes code. A distinct feature of the code is a dynamic cell-by-cell AMR based on a parallel fully threaded tree (FTT) structure. Architecturally, the code consists of three separate layers: (1) the FTT library which provides general services related to all parallel aspects of the code’s execution; (2) the reactive Navier-Stokes AMR code, ALLA, which rides on top of FTT and is responsible for numerical integration and AMR; (3) problem-specific algorithms such as material properties routines (equation of state, kinetics, microscopic transport) and problem initialization routines called from ALLA. Computations in ALLA are organized as a set of global computational steps, with each step followed by communication work which synchronizes data across the domain decomposition boundaries. Application algorithms are programmed in terms of work-functions which are passed to and executed by the global parallel iterator. The code is parallellzed using a hybrid OpenMP/MPI strategy. On each MPI rank the iterator parses the mesh and passes small chunks of cells to a work-function until the entire mesh is processed. The loop over the cells in a chunk is performed inside the work-functions themselves. These loops are parallelized using OpenMP. HSCD scales well up to 131K cores on the BG/P. Based on our simulations on BG/P, it would take about 2 sec/time-step on 1K cores on Fusion and a complete run would require about a 100K time-steps. Thus each run would need about 50K core-hours. We are planning a series of 8-10 parametric cases to identify important operating conditions and flow regimes. The code needs MPI, OpenMP libraries, C and Fortran 90 compilers. It is expected that the project will have 4-5 members including the PI. Project URL: Requested allocation: 450000 Justification: The requester has used 0 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 SepCFD
by accounts＠lcrc.anl.gov 07 Sep '10

07 Sep '10

Hello, A yearly allocation for the LCRC cluster has been requested with the following updated information: Submitter/PI: Kent E. Wardle Project Name: SepCFD Division: CSE Project title: CFD Simulation of Separations Processes Associated funding: - DOE-NE Nuclear Energy Advanced Modeling and Simulation (NEAMS) Program - DOE-NE Fuel Cycle Research and Development (FCR&D) Program Other Systems: - MCS Cosmea Linux Cluster (limited use for development and debugging) - MCS Full-disclosure (SiCortex 5832) - local 8-core workstation for meshing, postprocessing, and small-scale development Science: This project uses the open-source CFD package OpenFOAM (www.opencfd.co.uk) which is based on the finite volume solution method of the Navier-Stokes equations on arbitrary polyhedral, unstructured meshes for computational fluid dynamics simulations. OpenFOAM is capable of simulation of a wide range of physics and the focus in this project is on simulation of turbulent, multi-phase (liquid-liquid and liquid-liquid-air) mixing and separation. The specific application of interest is for liquid-liquid extraction processes employed in recycling of used nuclear fuel. In addition, a small portion of the project seeks to explore advanced CFD methods using a finite-element based unstructured, lattice Boltzmann simulation method. This is promising technique which can bridge the gap between molecular and continuum scales and can potentially allow much more efficient computational scaling than traditional CFD techniques. Project description: This project seeks to apply the open-source CFD package OpenFOAM to simulation of the complex, multiphase flow in liquid-liquid separations devices used for processing nuclear fuel via solvent extraction with a particular focus on simulations of the flow in the annular centrifugal contactor (www.mcs.anl.gov/~wardle/contactorSketch.png) For such processes, complex physical phenomena including free surface flow and liquid-liquid droplet breakup and coalescence must be incorporated. Accurate simulation of such physics requires fine meshes to resolve phase boundaries as well as efficient computational performance to enable transient simulations spanning several seconds of flow time despite the very small time steps that are required (~10-6 s). The specific simulation techniques to be used for the OpenFOAM simulations are based on the Volume of Fluid (VOF) method with Large Eddy Simulation (LES) of turbulence. Additional multiphase simulation methods will also be explored. The code package OpenFOAM uses a C++ object-oriented framework with dynamically linked libraries and high-level solvers to enable the user to write custom solvers and combine various physics models and equations enabling the evaluation of unique combinations of models such as are required for the development of advanced liquid-liquid mixing simulation capability as targeted in this work. Work during FY11 will focus on three-phase (liquid-liquid-air) simulations of the coupled mixing zone and separation zone of the annular centrifugal contactor (see contactor sketch link above). Here is an example of the type of simulations that have been conducted to date and will be built upon as part of this work: www.mcs.anl.gov/~wardle/4vpoly05base_BRC-side-15fps.avi. This simulation shows the turbulent mixing of water (blue), oil (red), and air (cyan) in the annular mixing region of a centrifugal contactor and was performed on ~2000 processors on MCS's SiCortex system. During FY10, large simulations for water-air and water-oil-air flow in a coupled model of the mixing and separating zone were conducted using a the VoF-based multiphase solver. A snapshot from the two-phase water-air simulation showing the flow of water in the contactor as seen from above is shown here: www.mcs.anl.gov/~wardle/fullExt-over-woExts.png. Analysis of the three-phase simulation is ongoin g and will provide valuable insight and provide a basis for the development of advanced multiphase models that will take place during FY11. The results of this work will also be included in a future paper currently being prepared. In addition to this full-scale, liquid-liquid simulation effort, during FY10 a focused project on exploration of a specific design feature of a commercially available contactor design was also conducted with the support of this project allocation. These results will be the core of another paper now in preparation and will also provide valuable direction for improving performance of contactors of this design to places such as Savannah River where they are in use. This project will support nuclear fuel cycle simulations efforts under DOE-NE as well as potential collaborations with DOE-EM sites such as Savannah River where centrifugal contactors and other multiphase separators are in use. The simulations performed here will also support ongoing FY11 experimental work in CSE led by the PI aimed at obtaining detailed multiphase measurements in an engineering-scale annular centrifugal contactor using electrical tomographic techniques. It is hoped that this experimental effort can provide data for validation of advanced multiphase CFD models. Aside from the work outlined above, one additional component of this project seeks to explore 'advanced' CFD techniques which may be more inherently suited to massive-scale parallel computations. One such advanced method is based on an unstructured method for solution of the discrete Boltzmann Equation along with a free-energy based multiphase simulation technique. This method does not require a global pressure solve and depends primarily on nearest neighbor information allowing it to exhibit more ideal parallel scaling. While to-date many of these simulations have been done on other systems (e.g. full-disclosure), a small fraction of the project allocation may also be used for this developmental effort. This project has one PI (Wardle) and may have at most 1-2 additional participating researchers, collaborators, or students. Project URL: Current FY Hours Used: undetermined amount New FY Requested allocation: 600000 Justification: The OpenFOAM code has been found to scale quite well on a number of different computational systems. The Volume of Fluid technique combined with the fine mesh resolution required for droplet and interface tracking necessitate simulation timesteps on the order of microseconds. As such, typical runs require about 100 wall hours for every 1 second of simulated flow time. On Fusion with more recent versions of the code with minor enhancements, we have observed that this can be reduced to ~50 hours/1 second of flow. A plot showing a comparison of the scaling (strong) of a characteristic OpenFOAM simulation (VOF interFoam solver) for a relatively small-size problem (530K tetrahedral cells) is posted here: www.mcs.anl.gov/~wardle/scaling_ann3d_530K_cosmeaFdSurveyorFusion.png. Note that the number of processors has been scaled to 2.0Ghz equivalents. OpenFOAM performs quite well on Fusion. During FY10 work, a number of runs using polyhedral meshes on the order o f 1M cells were done on as many as 128 processors. As new solvers are used the scaling may need to be explored again to ensure the optimum number of processors are used to efficiently use the project allocation. Thank You, The LCRC Accounts System

1 0

[LCRC Accounts] Yearly Allocation Request from SepMD
by accounts＠lcrc.anl.gov 07 Sep '10

07 Sep '10

Hello, A yearly allocation for the LCRC cluster has been requested with the following updated information: Submitter/PI: Kent E. Wardle Project Name: SepMD Division: CSE Project title: Molecular Dynamics Simulation of Separations Processes Associated funding: DOE-NE Nuclear Energy Advanced Modeling and Simulation (NEAMS) Program Other Systems: MCS Full-disclosure (SiCortex 5832) Science: Liquid-liquid extraction is used in a variety of chemical processes relevant to energy technology and most notably is the primary means for recycle of used nuclear fuel. Computational science-based simulation of such processes can lead to key advancements in fuel recycle technology through increased understanding of the scientific foundations of such physical phenomena. Predictive simulation of a complex process such as liquid-liquid extraction requires accurate representation of the fundamental physics and chemistry on a number of scales ranging from the molecular level to the continuum and engineering scale and covering an equally broad range of temporal scales. This project seeks to develop accurate intermolecular potentials incorporating molecular polarizability for use in molecular dynamics simulations of liquid-liquid interfacial phenomena. Electronic structure calculations for key molecular species of interest will be used to calculate polarizabilities. It is anticipated that physically realistic molecular models as developed here will help to move toward greater predictive capability and ultimately enable the design of advanced extractants for efficient separations processes. Project description: Polarizable potentials will be developed using a Drude oscillator implementation which has recently been incorporated in the NAMD code. While polarizable models have been developed for a number of the required species such as water and alkanes, models for extractant molecules are not available. This project has started by developing a polarizable model for a key extractant for nuclear-related separations process, namely, tributyl phosphate (TBP) using the FY10 project allocation. Additional refinements to this model and the development of other models for other extractants and components relevant to solvent extraction for nuclear fuelt recycling will be done during FY12. As before, this will be done using electronic structure calculations with Gaussian. The requested allocation will enable us to perform rigorous calculations and explore a wider configuration space for each molecular species to be explored. In general, these simulations will use only 8-16 processors but require multiple runs for many tens of hours. Following Gaussian simulations, MD model development and testing will be performed in CHARMM and NAMD. CHARMM will be used primarily for system configuration setup and small-scale testing while NAMD will be used for full-scale simulations. The allocation requested here is primarily for the Gaussian work as we are currently running NAMD very successfully on the SiCortex machine Full-Disclosure in MCS. In addition to the model development work and MD simulations, an additional exploratory effort on application of these molecular models for prediction of phase distribution coefficients using the Gibbs ensemble method as implemented in the MPI-parallel code Towhee will also be performed. This is a small exploratory effort at this point but may grow into a separate project in the future. As this project is still in the startup/exploratory phase, the requested allocation is quite small but increased resources may be required in the future. This project has one PI and perhaps 1-2 additional researchers. Project URL: Current FY Hours Used: undetermined amount New FY Requested allocation: 60000 Justification: Thank You, The LCRC Accounts System

1 0

[LCRC Accounts] Yearly Allocation Request from UNEDF_Optimization
by accounts＠lcrc.anl.gov 03 Sep '10

03 Sep '10

Hello, A yearly allocation for the LCRC cluster has been requested with the following updated information: Submitter/PI: Stefan Wild Project Name: UNEDF_Optimization Division: MCS Project title: Optimization of nuclear-energy DFT functionals Associated funding: DOE Scidac UNEDF (Universal Nuclear Energy Density Functional) DOE SciDAC TOPS (Towards Optimal Petascale Solvers) DOE Applied Mathematics Research in Derivative-Free Methods for Complex Simulations DOE SciDAC TOPS (Towards Optimal Petascale Solve Other Systems: 45K CPU-hours on the Cray XT4 (Franklin) at NERSC Science: One of the goals of the UNEDF (Universal Nuclear Energy Density Functional) SciDAC proposal is the improvement of current Density Functional Theory (DFT) functionals. These improvements will give theoreticians better tools for predicting the properties and behavior of atomic nuclei over the entire nuclear table. In this project, an optimization over parameter space (using a newly development derivative-free optimization algorithm) leads to new insights into refining DFT functionals used by the software (HFBTHO), which then requires a new parameter optimization. We plan to continue this work with the next generation of DFT functionals. Our work has generated considerable excitement in the UNEDF project and as a result the request for optimization capabilities is growing. This work is also leading to joint publications in science-based journals. Project description: This DFT parameter optimization is expensive: each simulation of a single nucleus at a particular parameter set using HFBTHO can consume up to .3 CPU-hours. We have recently used OpenMP at the nucleus level, in addition to compiler-based optimization, to bring the wall time per simulation down to 1.5 minutes on a single node of Fusion. Our current optimizations over 13 parameters typically require 200-300 evaluations of about 100 nuclei, which can be done on Fusion in 5-7.5 hours using 100 8-core nodes. This gives us up to 6K CPU-hours for each version of the functional we want to optimize. We note that our current computational needs for a single optimization, 100 nodes for 7.5 hours, yields a significant improvement in wall time over our previous needs of 12 nodes for 90 hours. These optimizations are being done using the derivative-free optimization software POUNDerS developed at Argonne. For prototyping purposes, this software currently requires MATLAB on one master node. Our team of 3 users anticipate doing 40 optimization runs during this project cycle. We also expect to analyze functionals based on more nuclei and depending on more parameters, potentially resulting in further increases to the number of nodes we require. Project URL: http://unedf.org/ Current FY Hours Used: undetermined amount New FY Requested allocation: 240000 Justification: Thank You, The LCRC Accounts System

1 0

[LCRC Accounts] Project Request: QIIME4ANL
by accounts＠lcrc.anl.gov 02 Sep '10

02 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: Jack Gilbert Applicant's institution: ANL Applicant's division: BIO IGSB Project Name: QIIME4ANL Project title: Earth Microbiome Project - QIIME metagenomic data analysis pipeline Associated funding: DOE Other Systems: n/a Science: The scientific objectives of this project are to install and implement a pipeline for the analysis of the 185 16S rDNA amplicon next generation sequencing projects we have produced, and the 159,815 further projects we will produce as part of the Earth Microbiome Project. (1) Perform initial clean-up of 16S-amplicon pyrosequence and illumina data produced both here at ANL and externally. (2) Perform basic statistical analysis and visualization of data Project description: Recently, the first ICiS Workshop was held at Snowbird, Utah, dealing with the concept of Terabase Metagenomics. This is the use of ultra-highthroughput sequencing to extend our sequencing capability for metagenomic science to the trillion base pair level. The workshop comprised 26 international experts who were asked to discuss how the application of this sequencing effort could be best applied to answer ecological questions regarding the global and temporal distribution of microorganisms on this planet. We named this effort the Earth Microbiome Iniative (EMP). This DOE centered initiative aims to acquire 160,000 environmental samples from across the globe and to analyse the metagenomic, metatranscriptomic and taxonomic distribution of the micobial communities in each of these ecosystems. For these 160,000 samples we will sequence more than 10 quadrillion base pairs of data (1 x 10>15 bp). This is an exceptionally ambitious, but entirely plausible. However, to make this possible we need to start exploring the computational infrastructure required to process this much data. One of the big three challenges in this study is the analysis and processing of billions of 16S rDNA fragments across thousands of samples to provide information which can be used to explore the ecological patterns and global dynamics of these ecosystems. To this end we are sequencing 16S rDNA sequences from 5000 samples from soils, marine and lake ecosystems as a pilot study to produce 2-3 billion 16S rDNA sequences. We have adopted a platform called QIIME which is designed to handle these data. However, to manage and process these we need access to the Fusion platform with 450,000 hours. To this end we have Qiime installed on Fusion as a softenv package. This will be the single largest study of microbial diversity and distribution across the world ever. Hence, the sheer scale and potential for ecological inference will mean that this work will be published in Science, providing a high impact study to launch the EMP. With access to the Fusion grid we can enable the first true mega-project and drive forward ecological theory on a global scale, making Argonne the centre of the new computational biology and microbial ecology revolution. Project URL: http://press.mcs.anl.gov/earthmicrobiome/ Requested allocation: 450000 Justification: The requester has used 0 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 LiBattModeling
by accounts＠lcrc.anl.gov 01 Sep '10

01 Sep '10

Hello, A yearly allocation for the LCRC cluster has been requested with the following updated information: Submitter/PI: Roy Benedek Project Name: LiBattModeling Division: CSE Project title: Lithium-Ion Battery Electrode Modeling Associated funding: DOE Office of Vehicle Technologies, EERE (Batteries for Advanced Transportation Technologies [BATT] Program) Other Systems: Franklin, at the National Energy Research Supercomputer Center (NERSC), LBL (43K hrs. allocated in FY2010) Science: First principles local-density-functional-theory methods are applied to calculate physical and electrochemical properties of proposed lithium-battery-electrode materials. This work complements an experimental program in the Battery Department of CSE led by Michael Thackeray. The acid-promoted dissolution of lithium manganate (LiMn2O4) is an obstacle to its commercialization as a cathode material in lithium-ion batteries. Our objective in the simulations described below is to understand the process of lithium manganate dissolution in acid, and assess different strategies to inhibit dissolution, such as doping and coating. Project description: All of our simulations are presently performed at the GGA+U level of density functional theory, with the VASP code (version 4.6). Our license has been upgraded to give access to version 5.2, which opens the possibility of employing the more accurate, but more costly, hybrid XC functional HSE06. Extensions of our previous work being considered for the next FY include: a. Doping of lithium manganate wire (divalent and trivalent substitutions for Mn; F substitutions for O) b. Structure and coating (with ALF3) of Li(Mn1.5Ni0.5)O4 wires: surface and interface properties will be investigated c. Dissolution of lithium manganate surfaces and wires: The acid-promoted dissolution-reaction-energy analyses that were begun in FY2010 will be extended. For example, only the Mn terminated slabs were considered in FY2010. By considering various surface orientations and terminations, the stability of flat surfaces against dissolution will be tested. d. Properties of Li ions and atoms in a-AlF3 coatings: The static and transport properties of Li ions in a-AlF3 coatings will be investigated. Energy of Li atoms and ions in a-AlF3, as well as their activation energies for diffusion, will be calculated. The energy of lithium atoms and ions in the vicinity of the interface between lithium manganate and a-AlF3 coating will be obtained. Project URL: Current FY Hours Used: undetermined amount New FY Requested allocation: 60000 Justification: Thank You, The LCRC Accounts System

1 0

[LCRC Accounts] Project Request: ColSim
by accounts＠lcrc.anl.gov 27 Aug '10

27 Aug '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: Wanming Liu Applicant's institution: ANL Applicant's division: HEP Project Name: ColSim Project title: Study of Photon Collimator Associated funding: DOE Other Systems: Science: For polarized ILC positron source, a photon collimator is required and need to be studied. The objective of this project is to implement a parallized gamma shower simulation based on EGSnrc and simulated the energy deposition in the photon collimator. Project description: A photon collimator is very critical for polarized positron source for future linear collider. This photon collimator needs to be able to handle several tens of kW power from collimated photon beam. It would be even critical to know the detail of energy deposition profile so that a proper cooling can be implemented. Since the collimator needs to be able to stop and absorb the collimated photon beam as much as possible, the size of this device might be too big to be handled by a single PC. For the first phase of this project, we will implement a parallelized gamma shower simulation based on EGSnrc and do some initial simulation studies with simple geometry. Project URL: Requested allocation: 10000 Justification: The requester has used 0 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 Aurora
by accounts＠lcrc.anl.gov 27 Aug '10

27 Aug '10

Hello, A yearly allocation for the LCRC cluster has been requested with the following updated information: Submitter/PI: Ramesh Balakrishnan Project Name: Aurora Division: LCF Project title: Assessing the performance of CFD codes on multicore computing platforms Associated funding: Other Systems: Blue Gene/P (at ALCF) Science: Direct Numerical Simulation (DNS), Large Eddy Simulation (LES), Reynolds Averaged Navier-Stokes (RANS), Detached Eddy Simulation (DES), and Improved Delayed Detached Eddy Simulation (IDDES) of Turbulence Phenomena, Project description: I have not utilized this allocation (as can be seen from the usage). I request that the period for which this allocation is valid be extended by a year. I am not, at this time, requesting more computing time. Project URL: Current FY Hours Used: undetermined amount New FY Requested allocation: 100000 Justification: Thank You, The LCRC Accounts System

1 0

[LCRC Accounts] Project Allocation Request
by accounts＠lcrc.anl.gov 17 Aug '10

17 Aug '10

Hello, A change in allocation has been requested: Requester: robl (Robert Latham) Project: radix Title: Scalable Parallel System Software Description: This is a Computer Science request for time on Fusion for development and evaluation of a family of tools and libraries with a direct benefit to all users of high-performance computing resources. The Radix group in Argonnne's Mathematics and Computer Science division conducts research and development on facilities to better enable efficient Parallel Programming, I/O, and Scientific Understanding. The Fusion system represents an attractive milestone for our software development: projects start out on our breadboard cluster and can further explore scalability in the presence of a high speed network on Fusion before making the jump to BlueGene-scale parallelism. Radix experiments fall into two broad categories. Often, we are concerned with latency. The infiniband network on Fusion provides an excellent low-latency environment for message passing and storage management research while also providing ample bandwidth for I/O and message bandwidth testing. Fusion hosts two quality parallel file systems, GPFS and PVFS. The wider HPC community often asks the Radix file system researchers to compare these file systems head-to-head. Fusion provides us the opportunity to do just that. Hosting two parallel file systems allows us to do more than just drag-race (so to speak). The two file systems have distinct characteristics, and act differently in the broader role of the I/O software stack. On Fusion, we can utilize a single software stack (application, parallel-netcdf, ROMIO), and compare the impact of a change in the underlying file system on overall application behavior. We know from other research, for example, that the two-phase optimization in ROMIO needs to adapt to the underlying file system, and that aligning variables to file system boundaries can yield performance improvements for parallel-netcdf. Fusion compute nodes actually look fairly attractive to active storage research. These nodes have enough disk space to store non-trivial datasets, and powerful enough processors to carry out complex computations without much impact on storage performance. Radix projects also focus on deriving insight both about application behavior as well as scientific insight. 8 core nodes make efforts like in-situ visualization, where for example an application renders a frame of a movie, ever more feasible. Even without fancy graphics accelerators, the processing power on Fusion nodes makes such analysis and visualization feasible, especially if the visualization can take advantage of multiple cores and multiple nodes. Naturally, the radix visualization tools have demonstrated scalability to Fusion sizes and beyond on BlueGene, so making full utilization of Fusion should not be a concern. As should be evident from the variety of planned experiments, the Radix group plans to make full use of the new features Fusion brings to LCRC and the flexibility of a Linux-based cluster. With Argonne employees and student collaborators, we expect around 20 members in FY2010. While Fusion may contain a modest number of compute nodes, the machine still represents an attractive platform for testing: very fast, low latency interconnect; 8 cores per node; .25 TB of storage per compute node. The testing and experiments the Radix group can carry out on Fusion ensures the tools the group develops today will remain relevant even as supercomputers grow. Quite a few radix-developed projects are part of the Fusion software stack: CPU hours for this project yield improvements and benefits not just for the radix group, but for all users of Fusion and indeed users of high-end computational resources worldwide. Current: undetermined amount Justification: Our large allocation is primarily driven by a set of HFS (Hadoop File System) experiments. These experiments use the HFS interface to communicate to a set of PVFS servers. We have consumed a lot of hours getting the scaling right, but we now know that 12 hours on 201 nodes would give us the required 7.5 tb of scratch space, one name node and 200 HFS/MapReduce nodes. We have tried to run this experiment on Magellan and have found it to be too unstable. Requested: 4000000 A specific reason has been given: This particular allocation request extension is for a long HFS (Hadoop File System) experiment. I have provided node counts and hours above in the "large allocation justification". This needs to be approved and the final allocation amount decided upon. Thank You, The LCRC Accounts System

1 0

[LCRC Accounts] Project Allocation Request
by accounts＠lcrc.anl.gov 06 Aug '10

06 Aug '10

Hello, A change in allocation has been requested: Requester: valdes (John Valdes) Project: lcrc-app-engr Title: LCRC Application Engineer support work Description: This project will be used by the LCRC application engineers to do support work for users of the LCRC cluster. This support might be in the form of installation and debugging of both LCRC core and user applications, startup help, etc. In FY2010, we anticipate needing a larger allocation for testing and updating user applications during the transition to the new Fusion cluster. Current: undetermined amount Justification: Requested: 40000 A specific reason has been given: With 2 active App Engineers and software testing for fusion, this project has run out of time for the FY. We request 40K more to cover the negative balance and to give us more time for the remainder of the FY. This needs to be approved and the final allocation amount decided upon. Thank You, The LCRC Accounts System

1 0