BEGIN:VCALENDAR VERSION:2.0 PRODID:-//CERN//INDICO//EN BEGIN:VEVENT SUMMARY:An Analysis of FFTW and FFTE Performance DTSTART;VALUE=DATE-TIME:20121017T150000Z DTEND;VALUE=DATE-TIME:20121017T153000Z DTSTAMP;VALUE=DATE-TIME:20261011T062958Z UID:indico-contribution-100-177@indico.ipb.ac.rs DESCRIPTION:Speakers: Josip Jakic (Scientific Computing Laboratory\, Insti tute of Physics Belgrade)\nOne of the most frequently used algorithms in e ngineering and scientific applications is Fast Fourier Transform (FFT). It s open source implementation (Fastest Fourier Transform of the West\, FFTW ) is widely used\, mainly due to its excellent performance\, comparable to the vendor-supplied libraries. On the other hand\, even if not yet in a f ully production state\, FFTE (Fastest Fourier Transform of the East) keeps up with FFTW\, and outperforms it for very large transform sizes. Here we present results of the performance and scalability tests of FFTW and FFTE libraries. Comparison is done using different compilers and parallelizati on approaches on Curie and Jugene supercomputers.\n\nhttps://events.saifa. rs/event/291/contributions/177/ LOCATION:National Library of Serbia URL:https://events.saifa.rs/event/291/contributions/177/ END:VEVENT BEGIN:VEVENT SUMMARY:Number Theory Algorithms on GPU Cluster DTSTART;VALUE=DATE-TIME:20121017T160000Z DTEND;VALUE=DATE-TIME:20121017T163000Z DTSTAMP;VALUE=DATE-TIME:20261011T062958Z UID:indico-contribution-100-181@indico.ipb.ac.rs DESCRIPTION:Speakers: Dobromir Georgiev (Institute of Information and Comm unication Technologies\, BAS)\nMany algorithms from Number Theory and thei r implementation in software are of high practical\nimportance\, since th ey are the building primitives of many protocols for data encryption and\n authentication of Internet connections. Number theory algorithms are also the basic part of\ncryptanalytic procedures. Many of these algorithms can be parallelized in a natural way. In this\npaper we describe our efforts t o develop a software package that implements various Number Theory\nalgori thms on GPU clusters and in partial our implementations of integer factori zation\nusing NVIDIA CUDA on clusters equipped with NVIDIA GPUs. Also we r eport results of our\nexperiments regarding the performance of our implem entation.\n\nhttps://events.saifa.rs/event/291/contributions/181/ LOCATION:National Library of Serbia URL:https://events.saifa.rs/event/291/contributions/181/ END:VEVENT BEGIN:VEVENT SUMMARY:On HPC for Hyperspectral Image Processing DTSTART;VALUE=DATE-TIME:20121017T143000Z DTEND;VALUE=DATE-TIME:20121017T150000Z DTSTAMP;VALUE=DATE-TIME:20261011T062958Z UID:indico-contribution-100-182@indico.ipb.ac.rs DESCRIPTION:Speakers: Mihnea Dulea (National Institute of Physics and Nucl ear Engineering)\nHyperspectral image processing still requires nowadays c onsiderable computational and storage resources\, beyond the available one s for a single server. In particular\, clustering of images gathered from the current satellites can be done in a reasonable time only using high pe rformance computing facilities.\n\nIn this paper we discuss the latest app roaches for fuzzy clustering techniques that are adapted to work on hundre ds of processors as well as the pre-processing techniques for data splitti ng and reading. Comparisons between different techniques are based on impl ementations for BlueGene/P.\n \nThe paper exposes part of the concepts pr esented in [1] and [2]\,\nas well as experimental results.\n\n[1] D. Petcu \, D. Zaharie\, S. Panica\, A. S. Hussein\, A. Sayed\, H. El-Shishiny\, “Fuzzy Clustering of Large Satellite Images using High Performance Compu ting\,” In Proceedings of SPIE Volume 8183\, SPIE Remote Sensing Confere nce: High-Performance Computing in Remote Sensing\, http://dx.doi.org/10.1 117/12.898281\, 2011\n\n[2] A.C. Toma\, S. Panica\, D. Zaharie\, D. Petcu \, "Computational Challenges in Processing Large Hyperspectral Images"\, s ubmitted to Grid\, Cloud & High Performance Computing Science"\, RO-LCG 20 12\, October 2012\n\nhttps://events.saifa.rs/event/291/contributions/182/ LOCATION:National Library of Serbia URL:https://events.saifa.rs/event/291/contributions/182/ END:VEVENT BEGIN:VEVENT SUMMARY:Performance and scalability evaluation of short fragment sequence alignment applications DTSTART;VALUE=DATE-TIME:20121017T153000Z DTEND;VALUE=DATE-TIME:20121017T160000Z DTSTAMP;VALUE=DATE-TIME:20261011T062958Z UID:indico-contribution-100-194@indico.ipb.ac.rs DESCRIPTION:Speakers: Gergely Windisch (Obuda University - Hungary)\nMappi ng short fragments to open access eukaryotic genomes at a very large scale presents a data processing challenge to the scientific world. The large v olume of data processing requires an immense amount of computing power for the tools to provide feasible response time\, which is essential for the researchers. The main tool used for such an application is BLAST which is the one we use in our portlets developed at Obuda University. There are wa ys for the scientists to use the BLAST algorithm either executing it local ly or using a web based BLAST tool. Usually the scope of usability of thes e solutions is limited because of the lack of computing power available. T he portlets developed at Obuda University are served by a web server but c omputation takes place in a massively parallel supercomputing environment. The type of the algorithm and the data it has to process make it an ideal candidate for a highly parallel execution on the HPSEE infrastructure. Th e portlets are available to the scientist community on the Bioinformatics eScience Gateway hosted at OU and powered by gUSE/WS-PGRADE technology\, w hile the backend is being served by the Hungarian HPSEE Infrastructure by NIIF. \nLately most of our work has been focused on evaluating the perform ance and scalability of our applications by profiling\, analyzing the resu lts of the tests and improving the performance of both the portlets and th e server-side massively parallel algorithm by environment optimization usi ng the data collected during the testing phase. \nIn this paper we will de scribe the two portlets (Deep Aligner and Disease Gene Mapper)\, discuss t he issues and challenges during the development and the performance analys is and present our results on the performance and scalability of the appli cations.\n\nhttps://events.saifa.rs/event/291/contributions/194/ LOCATION:National Library of Serbia URL:https://events.saifa.rs/event/291/contributions/194/ END:VEVENT END:VCALENDAR