Demonstration of this application has won main prize during 2nd EGEE User Forum/OGF 20 in Manchester 11.05.2007.
The purpose of porting this application to the grid was exploiting the combined power of a high number of CPUs interactively in order to visualize the behavior of plasmas inside fusion rectors . Numerically the method consists in integrating the equations of movement of a charged particle in the background of the magnetic field inside the reactor. For every individual particle, and at every step of the algorithm, the position, velocity, and other magnitudes of the particle are stored. This is the simulation core of the program. The 3D visualization of those particle trajectories is a very interesting tool for researchers. In particular many structures developing inside the plasma are three dimensional, for example magnetic islands. Also, the possibility of direct observing what happens inside the reactor in a 'real time' simulation provides with an insight impossible to obtain by just measuring average values of the magnitudes defining the plasma. For visualization purposes, all CPUs need to transmit the information related with the position and velocity of the particles in a continuous and synchronized way to the central machine, which after rendering will end the output as a video stream to the clients located at the researcher desktops. The user should then be able to interact with the application roughly once per minute. The interaction consists of changing the simulation parameters, such as electromagnetic fields, or the number of particles being simulated, and also zooming in certain parts of the reactor. In order to show requirements of this application let us go through example usage for the calculations that have been done for the Stellerator TJ-II of the CIEMAT in Madrid.
The graphical display should be such that 20 frames of data are transmitted every second to give impression of continuity. Every second each CPU sends 20 frames, each containing the information of 4000 points. Therefore every CPU will be producing 1 Mbyte/second of information. If we assume we spread the calculation over 3 farms, each with about 30 CPUs, the central cluster node will handle a data rate of 30 Mbytes/second, which has to be transmitted for visualization preparation.
- MD - fusion application use case, best demonstration during
EGEE User Forum / OGF20: md_demo_manchester.avi
- Slides that describes technical aspects of the demo:
- Short Guide presenting how to run this application from Migrating Desktop
- Startup script needed for running this application: fusion-start.sh
- More information about application and vo: int.eu.grid webpage - fusion
General Atomic and Molecular Electronic Structure System (GAMESS) is an application for molecular quantum chemistry (see http://www.msg.ameslab.gov/GAMESS/ for all details about GAMESS)
GAMESS is a program for ab initio molecular quantum chemistry. Briefly, GAMESS can compute SCF wavefunctions ranging from RHF, ROHF, UHF, GVB, and MCSCF. Correlation corrections to these SCF wavefunctions include Configuration Interaction, second order Perturbation Theory, and Coupled-Cluster approaches, as well as the Density Functional Theory approximation. Nuclear gradients are available, for automatic geometry optimization, transition state searches, or reaction path following. Computation of the energy hessian permits prediction of vibrational frequencies, with IR or Raman intensities. Solvent effects may be modeled by the discrete Effective Fragment Potentials, or continuum models such as the Polarizable Continuum Model. Numerous relativistic computations are available, including third order Douglas-Kroll scalar corrections, and various spin-orbit coupling options. The Fragment Molecular Orbital method permits use of many of these sophisticated treatments to be used on very large systems, by dividing the computation into small fragments.
- MD - gamess application use case md_demo_gamess.avi
- Short Guide - tutorial presenting how to run this application from Migrating Desktop gamess.doc
- Example input files needed for running this application: gamess.inp
SemtiKamols project in University of Latvia aims to develop semantic resources and methodologies for automatic meaning extraction from Latvian texts. This ultimate goal requires the lower levels of the language analysis, namely, morphology, syntax and lexical semantics, to be properly implemented in the first place. Besides basic grammar parsing for Latvian, the project utilizes techniques from Semantic Web and related formal logic-based approaches to enable formal reasoning about the facts mentioned in the text. The advantage of such semantic framework is that it avoids the text disambiguation problem at the lower parsing levels, and leaves it to the higher semantic processing layers. The semantic processing layer is being built upon techniques adapted from the frame semantics (FrameNet), ontologies (OntoSem, SUMO) and ACE (Attempto Controlled English).