[{"authors":null,"categories":null,"content":"Bluejeans Hardware Features\nBeowulf is a multi-computer architecture which can be used for parallel and distributed computations.\nBluejeans (Bj) is a Beowulf HPC cluster based of DSA-LabMNCP, composed by 36 workingnodes and 4 servicenode connected together via switch Ethernet 1000 Mb/s dedicated.\nWorkingnode features:\nn.1 CPU Intel Dual Core 2,6 GHz;\nn.1 1 GB RAM;\nn.1 Ethernet connection 1000 Mb/s.\nServicenodes Features:\nn.1 CPU Intel Dual Core 2,6 GHz;\nn.2 1 GB RAM;\nn.2 Ethernet connection 1000 Mb/s.\nBj: Front view Bj: side view\n              The servicenodes has been configurate for export the followed services:\nUser Authentication;\nNFS Server;\nData Storage and Data Backup;\nSSH login server;\nFollowed you can see the shema hardware of the DSA-LabMNCP/sHPC-Bluejeans:\nIn total the Bj is composed by 40 machines with 80 core available for parallel.\nand distributed computation.\nThe data storage server export to the servicenodes 8 hard disk 2 Tera (mirrored) dedicated for output simulation storage and data backup.\nThe goal of DSA Bluejeans cluster is provide computational resource and distributed environment at the DSA research activities and the DSA-LabMNCP Team can run their batch jobs and distributed compute under the resource manager Torque (PBS).The Torque scheduler provide the followed features:\n Fault Tolerance Additional failure conditions checked/handled Node health check script support; Scheduling Interface; Scalability; Usability;  Also, you can see the Bj runtime today performance at this link.\n","date":-62135596800,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":-62135596800,"objectID":"8005c811699c22aa7644bf65a38e134a","permalink":"/hardware/bluejeans/","publishdate":"0001-01-01T00:00:00Z","relpermalink":"/hardware/bluejeans/","section":"hardware","summary":"Bluejeans Hardware Features\nBeowulf is a multi-computer architecture which can be used for parallel and distributed computations.\nBluejeans (Bj) is a Beowulf HPC cluster based of DSA-LabMNCP, composed by 36 workingnodes and 4 servicenode connected together via switch Ethernet 1000 Mb/s dedicated.\nWorkingnode features:\nn.1 CPU Intel Dual Core 2,6 GHz;\nn.1 1 GB RAM;\nn.1 Ethernet connection 1000 Mb/s.\nServicenodes Features:\nn.1 CPU Intel Dual Core 2,6 GHz;\nn.2 1 GB RAM;","tags":null,"title":"Cluster HPC Bluejeans","type":"hardware"},{"authors":null,"categories":null,"content":"On September 28, 2009, a workstation Genesis GE-i940 Tesl, based on both GPGPU* and nVidia/CUDA** Technologies has been installed at DSA/LabMNCP.\nIt is a testbed for developing advanced simulation in the following research field:\n Stochastic simulation; Molecular Dynamics; Atmospheric and climate modeling; Weather forecast investigation; Grid/Cloud Hybrid Virtualization;  *\n“GPGPU stands for General-Purpose computation on Graphics Processing Units, also known as GPU Computing. Graphics Processing Units (GPUs) are high-performance many-core processors capable of very high computation and data throughput. See more here.”\n**\n“NVIDIA® CUDA™ is a general purpose parallel computing architecture that leverages the parallel compute engine in NVIDIA graphics processing units (GPUs) to solve many complex computational problems in a fraction of the time required on a CPU. See more here. “\n   Hardware          Mainboard Asus x58/ICH10R 3 PCI-Express x16, 6 SAT, 2 SAS, 3+6 USB   CPU i7-940 2,93 133 GHz fsb, Quad Core 8 Mb cache   RAM 6 x 2Gb DRR 3 1333 DIM   Hard Disk 2 x 500 Gb SATA 16Mb cache 7.200 RPM   GPU 1 Quadro FX5800 4Gb RAM    2 x Tesla C1060 4 Gb RAM       Software      OS: GNU/Linux CentOs 5.3 64 Bit   Driver: nVidia Cuda 180.22 Linux 64bit   VMware: VMware-server-2.0.2       OUTPUT of First Test:        Serial simulation(ms) GPU(ms)   execution time for malloc 0.02 175.21 ms   execution time for RndGnr 51430.92 2283.19   execution time for init 275.48 0.31   execution time for computing 391391.12 329.19 ms   execution time for I/O 56822.77 64740.54 ms   execution time for GPU/CPU  198.43 ms    Output using GPU,  device 0 : Quadro FX 5800 device 1 : Tesla C1060 device 2 : Tesla C1060\nSelected device: 2 \u0026lt;\u0026lt;\u0026lt;\u0026lt;\u0026lt;\u0026lt;\u0026lt;\u0026lt;\u0026lt;\u0026lt;\u0026lt;\u0026lt;\u0026lt;\u0026lt;\u0026lt;\u0026lt;\u0026lt;\u0026lt;\ndevice 2 : Tesla C1060 major/minor : 1.3 compute capability Total global mem : -262144 bytes Shared block mem : 16384 bytes RegsPerBlock : 16384 WarpSize : 32 MaxThreadsPerBlock : 512 TotalConstMem : 65536 bytes ClockRate : 1296000 (kHz) deviceOverlap : 1 deviceOverlap : 1 MultiProcessorCount: 30\nUsing 1048576 particles 100 time steps\n\n","date":-62135596800,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":-62135596800,"objectID":"e0355645e6bac4892de226ed4d33e530","permalink":"/hardware/genesis/","publishdate":"0001-01-01T00:00:00Z","relpermalink":"/hardware/genesis/","section":"hardware","summary":"On September 28, 2009, a workstation Genesis GE-i940 Tesl, based on both GPGPU* and nVidia/CUDA** Technologies has been installed at DSA/LabMNCP.\nIt is a testbed for developing advanced simulation in the following research field:\n Stochastic simulation; Molecular Dynamics; Atmospheric and climate modeling; Weather forecast investigation; Grid/Cloud Hybrid Virtualization;  *\n“GPGPU stands for General-Purpose computation on Graphics Processing Units, also known as GPU Computing. Graphics Processing Units (GPUs) are high-performance many-core processors capable of very high computation and data throughput.","tags":null,"title":"Genesis GE-i940 Tesla","type":"hardware"},{"authors":null,"categories":null,"content":"GreenJeans is new experimental HPC Cluster/Beowulf of DSA build up with the aim to create both economy and enviroment sustainable solution for the Scientific HPC field.\nGreenJeans Making of\u0026hellip;\nOver the Green we have installed the followed software:\n CUDA(Driver / Toolkit / SDK) SDK Java Sun MPICH4 V1 MPICH4 V2 MPI2-VMI Eucalyptus(KVM/QEMU Hypervisor) Ganglia Torque  Every work node of GreenJeans have installed nVidia GeForce GTX Ti\n Device 0: “GeForce GTX 560 Ti”\nCUDA Driver Version: 4.0\nCUDA Runtime Version: 4.0\nCUDA Capability Major/Minor version number: 2.1\nTotal amount of global memory: 1072889856 bytes\nMultiprocessors x Cores/MP = Cores: 8 (MP) x 48 (Cores/MP) = 384 (Cores)\nTotal amount of constant memory: 65536 bytes\nTotal amount of shared memory per block: 49152 bytes\nTotal number of registers available per block: 32768\nWarp size: 32\nMaximum number of threads per block: 1024\nMaximum sizes of each dimension of a block: 1024 x 1024 x 64\nMaximum sizes of each dimension of a grid: 65535 x 65535 x 65535\nMaximum memory pitch: 2147483647 bytes\nTexture alignment: 512 bytes\nClock rate: 1.64 GHz\nConcurrent copy and execution: Yes\nRun time limit on kernels: No\nIntegrated: No\nSupport host page-locked memory mapping: Yes\nCompute mode: Default (multiple host threads can use this device simultaneously)\nConcurrent kernel execution: Yes\nDevice has ECC support enabled: No\nDevice is using TCC driver mode: No\n deviceQuery, CUDA Driver = CUDART, CUDA Driver Version = 4.0, CUDA Runtime Version = 4.0, NumDevs = 1, Device = GeForce GTX 560 Ti\n","date":-62135596800,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":-62135596800,"objectID":"f31147029ae3471e0132a33597ade3e7","permalink":"/hardware/greenjeans/","publishdate":"0001-01-01T00:00:00Z","relpermalink":"/hardware/greenjeans/","section":"hardware","summary":"GreenJeans is new experimental HPC Cluster/Beowulf of DSA build up with the aim to create both economy and enviroment sustainable solution for the Scientific HPC field.\nGreenJeans Making of\u0026hellip;\nOver the Green we have installed the followed software:\n CUDA(Driver / Toolkit / SDK) SDK Java Sun MPICH4 V1 MPICH4 V2 MPI2-VMI Eucalyptus(KVM/QEMU Hypervisor) Ganglia Torque  Every work node of GreenJeans have installed nVidia GeForce GTX Ti\n Device 0: “GeForce GTX 560 Ti”","tags":null,"title":"GreenJeans","type":"hardware"},{"authors":null,"categories":null,"content":"The Grid Analysis and Display System (GrADS) is an interactive desktop tool that is used for easy access, manipulation, and visualization of earth science data. The format of the data may be either binary, GRIB, NetCDF, or HDF-SDS (Scientific Data Sets). GrADS has been implemented worldwide on a variety of commonly used operating systems and is freely distributed over the Internet. For more information, follow the link to the GrADS website: http://www.iges.org/grads/.\nGrADS is widely used by the computational environmental scientist community thanks to the scripting language support (GrADS script) and to the external plugin feature (Open GrADS).\nThe DSA-LMNCP contribution to the GrADS world is a Java wrap enabling Java applications to use GrADS as a back end data analysis and rendering.\nThe project is really a “work in progress” and open to any kind of external contribution.\nThe snapshot archive is downloadable here.\nDownlad Jgrads\n","date":-62135596800,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":-62135596800,"objectID":"acb94d1c9a9da98176a592464abea803","permalink":"/download/jgrads/","publishdate":"0001-01-01T00:00:00Z","relpermalink":"/download/jgrads/","section":"download","summary":"The Grid Analysis and Display System (GrADS) is an interactive desktop tool that is used for easy access, manipulation, and visualization of earth science data. The format of the data may be either binary, GRIB, NetCDF, or HDF-SDS (Scientific Data Sets). GrADS has been implemented worldwide on a variety of commonly used operating systems and is freely distributed over the Internet. For more information, follow the link to the GrADS website: http://www.","tags":null,"title":"jGrADS: Java Wrap sull’analisi di rete e del sistema di visualizzazione (GrADS) ","type":"download"},{"authors":null,"categories":null,"content":"Ph.D. Student in Applied Mathematics, Malek Ashtar University of Technology, Isfahan, Iran.\nThe main topic of my research is meshless methods based on radial basis functions. My goal is the modification, development and accelerates the meshless methods by some parallel procedure and fast algorithms.\nResearch interest:\n Numerical Analysis Meshless Methods (Local and strong forms) Kernel Based Approximation techniques Fractional Differential Equations Adaptive Computational Techniques Fast Numerical Methods High-performance scientific computing.  Papers\n An adaptive sparse meshless technique in greedy algorithm framework to simulate an anomalous mobile-immobile transport model (submitted) H. R. Ghehsareh, M. Raei, A. Zaghian, Application of meshless local Petrov- Galerkin technique to simulate two-dimensional time fractional Tricomi-type problem (Under review) H. R. Ghehsareh, M. Raei, A. Zaghian, Numerical simulation of a modified anomalous diffusion process with nonlinear source term by a local weak form meshless method, Engineering Analysis with Boundary Elements, 98 (2019) 64-76. H. R. Ghehsareh, A. Zaghian, M. Raei, A local weak form meshless method to simulate a variable order time-fractional mobile–immobile transport model, Engineering Analysis with Boundary Elements, 90 (2018) 63-75.  ","date":-62135596800,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":-62135596800,"objectID":"f4d24d878214fedefe49e6dcc43ca610","permalink":"/staff/marzie-raei/","publishdate":"0001-01-01T00:00:00Z","relpermalink":"/staff/marzie-raei/","section":"staff","summary":"Ph.D. Student in Applied Mathematics, Malek Ashtar University of Technology, Isfahan, Iran.\nThe main topic of my research is meshless methods based on radial basis functions. My goal is the modification, development and accelerates the meshless methods by some parallel procedure and fast algorithms.\nResearch interest:\n Numerical Analysis Meshless Methods (Local and strong forms) Kernel Based Approximation techniques Fractional Differential Equations Adaptive Computational Techniques Fast Numerical Methods High-performance scientific computing.","tags":null,"title":"Marzie Raei","type":"staff"},{"authors":null,"categories":null,"content":"Software developed:  InstrumentService and AVL AbstractInstrumentFramework and AVL**  ** AVL requires the installation of Ascom Platforms for the use of the telescope.\nThe goal of this Thesis-project is the creation of a software system for secure sharing and aggregation of data acquisition tools, geographically distributed, for engineering and scientific applications. To do this, was choosen the technology made available by computational grids “web service-based” using the software Globus-toolkit Toolkit 4″.\nABSTRACT INSTRUMENT The use of grid technology to control instruments for acquisition and retrieve the data implies the need of develop a standard methodology of interface between the different types of hardware. During the development of the thesis-project has been implemented the framework AbstractInstrument (AIF), used for virtualization of the instruments, through the use of standard interfaces that provide a high level of interation, common to all instruments. Thanks to this approach, any instrument can be hadled through a device driver of hight level.\nINSTRUMENT SERVICE To manage a virtualized instrument through a computational grid, was developed, using the Globus Toolkit version 4 (GT4), developed by the Mathematics and Computer Science Division of the Argonne National Laboratory (MCS / ANL) and the Computation Institute of the European university of Chicago (UOC-CI), scientific institutions of global relevance which are ongoing collaboration, the secure grid-web-service Instrument Service (IS) which, through the functionality offered from the AIF allows access, control and sharing of tools across the virtualized Grid. The IS can interface any instrument to the grid automatically publishing it on the Index Service, standard component of GT4, the metadata relating to each instrument and eventually the values of the current measure acquired by the sensors. This feature, fully configured configurable in terms of information published, allows the Resource Broker Service (RBS), a component developed at the Department of Applied Sciences, to search tools as well as other grid resources through a query written with the description language resource ClassAd, used by Condor and by gLite and considered the de facto standard in this type of applications.\nAVL In order to show what is actually possible by using the components developed, was realized a Virtual laboratory dedicated to astronomical applications (AVL). AVL actually supports robotic telescopes and weather stations that can be used in applications of computational grid integrating also other components such as services for the distribution of multidimensional environmental-data.\nInstrumentService.zip\nAbstractInstrument.zip\n","date":-62135596800,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":-62135596800,"objectID":"5c1ce42a6161b7968b89c1e04752a949","permalink":"/download/quadro/","publishdate":"0001-01-01T00:00:00Z","relpermalink":"/download/quadro/","section":"download","summary":"Software developed:  InstrumentService and AVL AbstractInstrumentFramework and AVL**  ** AVL requires the installation of Ascom Platforms for the use of the telescope.\nThe goal of this Thesis-project is the creation of a software system for secure sharing and aggregation of data acquisition tools, geographically distributed, for engineering and scientific applications. To do this, was choosen the technology made available by computational grids “web service-based” using the software Globus-toolkit Toolkit 4″.","tags":null,"title":"Servizio Instrument ed astratti strumento quadro","type":"download"},{"authors":null,"categories":null,"content":"SQLH and Hyrax\nPre requirements: LIBPQ .\nThe software developed is a Hyrax plugin, useful to add SQL Query capabilities to the BES server. It consists in a fully functional SQL handler that you can customize and expand. In this ALPHA release you can:\n Use OLFS to set constraints Set complex SQL query into the dataset file (Join, union) Set constraints into the dataset Set database password access into the dataset file OR using constraints  BES Software\nBES is a high-performance back-end server software framework that allows data providers more flexibility in providing end users views of their data. The current OPeNDAP data objects (DAS, DDS, and DataDDS) are still supported, but now data providers can add new data views, provide new functionality, and new features to their end users through the BES modular design. Providers can add new data handlers, new data objects/views, the ability to define views with constraints and aggregation, the ability to add reporting mechanisms, initialization hooks, and more.\nOPeNDAP provides the tools to build these new modules that can then be dynamically loaded into the BES.\nHyrax\nHyrax is the next generation server from OPeNDAP. It utilizes a modular design that employs a light weight Java servlet (aka OLFS) to provide the public-accessible client interface, and a back-end daemon, the BES to handle the heavy lifting. The BES uses the same handlers that are used with Server3 (also know as the CGI Server) but loads those at run time.\nBenefits:\n The servlet architecture is faster, more robust, and secure than CGI invoked Perl scripts. A single installation can handle multiple data representations (hdf4, hdf5, netcdf, et c.) THREDDS catalog functionality. A prototype SOAP interface for OPeNDAP data services.  OLFS: The Hyrax Front End\nThe OPeNDAP Lightweight Frontend Servlet (OLFS) provides the public-accessible client interface for Hyrax. The OLFS communicates with the Back End Server (BES) to provide data and catalog services to clients. The OLFS implements the DAP2 protocol and supports some of the new DAP4 features. We hope that other groups will develop new front end modules that will implement other protocols.\nNew Features:\n Provides THREDDS Catalogs responses Prototype SOAP interface.  SQLH: SQL Handler\nIt’s an SQL handler used to connect databases to an OpeNDAP Hyrax (bes) server. Written in C++, it uses libpq to query DB. It implements many interfaces useful to give you an easy way to modify and to use it with other ODBC libraries. It is composed by tree Basic component:\nThe SQLTable class used to load the requested file. Alternatively:\n A DAS object A DDS object A DataDDS object (a flat SQLSequence of strings)  The SQLFilterC(onstraint)E(xpression):\nUsed to parse the selected dataset and (or) the contraint expressions specified by the user (using the OLFS). It actually remove ALL the constraint expression from the URL and use it to build a filtered SQL query. So the filter operation is done by the SQL server and no constraints will be passed to the BES. You can easely change this behaviour.\nThe SQLConnector\nIt’s a component used to manage data transfer from the database (read-only). It’s composed by the following two components:\n The SQLResultSet Specifies a common interface used to get values from the accessed database. Its methods are used in the SQLTable and SQLFilterCE.\n The SQLConnection Specifies a common interface used to open and close a connection to the accessed database. Its methods are used in the SQLTable and SQLFilterCE using SQLConnector.\n  Enjoy!\nSQLHandler.tar.gz [SQLH-UMLs.zip(/zip/SQLH-UMLs.zip)\n","date":-62135596800,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":-62135596800,"objectID":"5739d2a2526b3aa53eabb7814d246de5","permalink":"/download/sqlhandler/","publishdate":"0001-01-01T00:00:00Z","relpermalink":"/download/sqlhandler/","section":"download","summary":"SQLH and Hyrax\nPre requirements: LIBPQ .\nThe software developed is a Hyrax plugin, useful to add SQL Query capabilities to the BES server. It consists in a fully functional SQL handler that you can customize and expand. In this ALPHA release you can:\n Use OLFS to set constraints Set complex SQL query into the dataset file (Join, union) Set constraints into the dataset Set database password access into the dataset file OR using constraints  BES Software","tags":null,"title":"SQL Handler","type":"download"}]