Подготовка специалистов в области технологий программирования для многопроцессорных систем с массовым параллелизмом
The book contains selected papers that were presented on PhD Summer schools on Scientific Computing jointly organized by Waterford Institute of Technology, Lomonosov Moscow State University, Kyiv National Taras Shevchenko University, Saint-Petersburg State University and Nanjing University of Technology. The schoold were mainly organized in teleconference mode and linked researchers and PhD students from several countries.
Now we have the need for methodics of teaching the topic "parallel computing" in secondary school. The paper presents a three-year experience of the author in this field: a methodical approach, the selection of materials, the business games, experience of tasks on parallel computing at the contest "TRIZformashka", classes of tasks, examples of tasks, program executors, texts for propaedeutic textbook on informatics.
The problems of identifying latent parallelism in the algorithm by explicitly max (the construction of stacked-parallel form of the algorithm graph) and implicit (the method of streaming - DATA-FLOW - calculations), the development of parallel programs in the MPI-paradigm programming and quantitative research strength calculations for the acceleration parallelization on the parameters of a multiprocessor system and the quality of parallel programs. The manual is practical and can be used by students to prepare for the performance of laboratory and practice of the works, of course and diploma projects. Generated by network applications ra-operability in a multiprocessor environment, architecture MPP (Massively Par-allel Processing); particularly on Linux-cluster computing IT department MGUPI 4. Before working to understand whole con-SPECT lectures on 'Parallel Computing'.
In this paper, we present results of a computational evaluation of goMapReduce parallel programming model approach for solving distributed data processing problems. In some applications, particularly data center problems, including text processing the programming models can aggregate significant number of parallel processes. We first discuss the implementation of these approaches using both Linux and Plan9 operating system and conduct a comparative scalability study of the both. From these results, we empirically show that, in practical implementation and evaluation of a goMapReduce model, the degree of OS's support for distributed processing encountered in solving the resulting word counting problem is crucial. We conclude that the goMapReduce approach under Plan9 may be useful in developing a heuristic approach for the data center problems.
The manual sets out the requirements of science and industry, leading to use of multicomputer systems and multiprocessor systems, which inevitably use the principle of parallel computing, background and state of the art, describes the main approaches to the organization of multiprocessor computer systems, development of parallel algorithms for the numerical solution of problems and parallel programming techniques.
The research subject is the computational complexity of the probabilistic neural network (PNN) in the pattern recognition problem for large model databases. We examined the following methods of increasing the efficiency of a neuralnetwork classifier: a parallel multithread realization, reducing the PNN to a criterion with testing of homogeneity of feature histograms of input and reference images, approximate nearestneighbor analyses (BestBin First, directed enumeration methods). The approach was tested in facialrecognition experiments with FERET dataset.