@ARTICLE{26583204_53707454_2012, author = {А. Ostroverh and А. Tsyrkov and А. Krishtop and G. Tsyrkov}, keywords = {, parallel computing processes, intelligent communications medium, invariant information model, cross-functional organizational structuretechnical and economic forecasting}, title = {Methods of construction high performance computing environment process control systems}, journal = {}, year = {2012}, number = {1(19)}, pages = {62-69}, url = {https://bijournal.hse.ru/en/2012--1(19)/53707454.html}, publisher = {}, abstract = {Alexander Ostroverh- Deputy Head of Department, Khrunichev State Research and Production Space Center.Address: 18, Novozavodskaya str., Moscow, 121309, Russian Federation. E-mail: tias@matias.ru Alexander Tsyrkov - Head of Complex, Khrunichev State Research and Production Space Center.Address: 18, Novozavodskaya str., Moscow, 121309, Russian Federation. E-mail: tias@matias.ru Alexey Krishtop - Head of Laboratory, MATI - Russian State Technological University.Address: 3, Orshanskaya str., Moscow, 121552, Russian Federation.E-mail: tias@matias.ru Georgy Tsyrkov - Associate Professor, MATI - Russian State Technological University.Address: 3, Orshanskaya str., Moscow, 121552, Russian Federation.E-mail: tias@matias.ru The article looks into a business case associated with designing a high-performance computing environment for managing production of complex technical products in a high-tech machine engineering company. The target function for evaluating the high-performance computing environment is generalized in accordance with enlarged cost targets and performance time. The article deals primarily with generalizing and formulating the principles of designing a production management system on the basis of the limitations exist in the real machine engineering enterprise.In the case studied, the production management system is based on the CALS technology, with a high degree of heterogeneity of software and hardware at various stages of the production process in the enterprise, not employing any standard ERP platform or its modules. At the same time, the case shows the application of up-to-date principles of forming cross-functional structures, the centralization of planned-economy aspects of the given production segment, as well as the multiplex approach to setting up data exchange channels. In general, the configuration is described as a self-learning.In order to organize the multiplex information environment, it is necessary to employ patterns of structure-parametric modeling. It should be noted that this method has been tested and approved of as a useful for solving many practical engineering problems. The core of the modeling complex - structural-parametric modeler (SP-modeler) - is focused on modeling objects in the corporate information system. In its turn, setting up the control-management system on the multiplicity principles required finding a solution for the optimization of data exchange and processing algorithms, as well as performing analytical and modeling calculations.The article also thoroughly studies various methods parallelization of data exchange and processing in accordance with the general configuration of the production processes and the corresponding data. The study carried out results in the analysis, summarization and description of the approaches and organizational scheme of high-performance computing environments within the production management system, taking into account the various limitations and specific features of a real machine engineering enterprise.}, annote = {Alexander Ostroverh- Deputy Head of Department, Khrunichev State Research and Production Space Center.Address: 18, Novozavodskaya str., Moscow, 121309, Russian Federation. E-mail: tias@matias.ru Alexander Tsyrkov - Head of Complex, Khrunichev State Research and Production Space Center.Address: 18, Novozavodskaya str., Moscow, 121309, Russian Federation. E-mail: tias@matias.ru Alexey Krishtop - Head of Laboratory, MATI - Russian State Technological University.Address: 3, Orshanskaya str., Moscow, 121552, Russian Federation.E-mail: tias@matias.ru Georgy Tsyrkov - Associate Professor, MATI - Russian State Technological University.Address: 3, Orshanskaya str., Moscow, 121552, Russian Federation.E-mail: tias@matias.ru The article looks into a business case associated with designing a high-performance computing environment for managing production of complex technical products in a high-tech machine engineering company. The target function for evaluating the high-performance computing environment is generalized in accordance with enlarged cost targets and performance time. The article deals primarily with generalizing and formulating the principles of designing a production management system on the basis of the limitations exist in the real machine engineering enterprise.In the case studied, the production management system is based on the CALS technology, with a high degree of heterogeneity of software and hardware at various stages of the production process in the enterprise, not employing any standard ERP platform or its modules. At the same time, the case shows the application of up-to-date principles of forming cross-functional structures, the centralization of planned-economy aspects of the given production segment, as well as the multiplex approach to setting up data exchange channels. In general, the configuration is described as a self-learning.In order to organize the multiplex information environment, it is necessary to employ patterns of structure-parametric modeling. It should be noted that this method has been tested and approved of as a useful for solving many practical engineering problems. The core of the modeling complex - structural-parametric modeler (SP-modeler) - is focused on modeling objects in the corporate information system. In its turn, setting up the control-management system on the multiplicity principles required finding a solution for the optimization of data exchange and processing algorithms, as well as performing analytical and modeling calculations.The article also thoroughly studies various methods parallelization of data exchange and processing in accordance with the general configuration of the production processes and the corresponding data. The study carried out results in the analysis, summarization and description of the approaches and organizational scheme of high-performance computing environments within the production management system, taking into account the various limitations and specific features of a real machine engineering enterprise.} }