The Experts below are selected from a list of 91542 Experts worldwide ranked by ideXlab platform
Chen Zongji - One of the best experts on this subject based on the ideXlab platform.
-
real time simulation of flight control system based on rtw and vxworks
Computer Simulation, 2007Co-Authors: Chen ZongjiAbstract:A rapid prototype method of flight control system was introduced based on RTW in Matlab/Simulink and Tornado/VxWorks. A nonlinear longitudinal short-period aircraft model was given. And a dynamic inversion control system satisfied with the C*.criterion was designed by feedback linearization. RTW generates optimized source code and executable modules by using auto generator from the Simulink models. When running in Vxworks, the modules can be monitored real-time and regulated online. In an ether local area net (LAN) environment, the modules were downloaded to the Target Computer based a Pentimum4 personal Computer and run real-time. Compared with the results only in Simulink, the comparison simulation result figures and some conclusions were given.
M. Asunción Castaño - One of the best experts on this subject based on the ideXlab platform.
-
Convolutional neural nets for estimating the run time and energy consumption of the sparse matrix-vector product:
The International Journal of High Performance Computing Applications, 2020Co-Authors: Maria Barreda, Manuel F. Dolz, M. Asunción CastañoAbstract:Modeling the performance and energy consumption of the sparse matrix-vector product (SpMV) is essential to perform off-line analysis and, for example, choose a Target Computer architecture that del...
A. V. Fedorov - One of the best experts on this subject based on the ideXlab platform.
-
Debugging and monitoring of applicationprograms in the BagrOS-4000 real-time operation system based on the Elbrus architecture
CRI «Electronics», 2019Co-Authors: R. G. Gordienko, O. G. Fedorenko, A. A. Demidov, A. V. FedorovAbstract:The article is concerned with the problems of monitoring and debugging of operating system processes, the effectiveness of which in the hard real-time operating system version does not allow any stopping to analyze the state of software and/ or hardware. The paper describes the concept of a debugging and monitoring system developed taking into account this feature in the Sukhoi design bureau for the BagrOS-4000 hard real-time operating system on the Elbrus architectural platform together with the specialists of MCST JSC. The method of non-stop monitoring and data collection in hard realtime processes in the multiprocess multimodular systems is discussed. An approach to the management of debugging Targets in terms of source code using the DWARF debugging information specification is presented. The transition from the instrumental machine to the system server built into the Target Computer is described. Given the rationale for the use of client-server architecture in the debugging and monitoring system for BagrOS-4000. A comparative analysis of the key functionality of the debugging and monitoring system with the existing debugging systems has been carried out; the key aspects of the DMS architecture have been considered. The design of a machine-dependent interface required for the integration of the independent hardware platforms into the BagrOS-4000 system when implementing the system on an integrated avionics module of the onboard complex is discussed. The results of testing of the debugging and monitoring systems are analyzed in terms of efficiency versus the classical method of using the debug console prints when debugging a real-time operating system. Most of the above solutions are universal and have been successfully tested using other microprocessor platforms on multi-threaded application programs of real-time operating systems running on multi-core processors, including the MIPS, Power PC, Intel platforms
Castaño Álvarez, María Asunción - One of the best experts on this subject based on the ideXlab platform.
-
Convolutional neural nets for estimating the run time and energy consumption of the sparse matrix-vector product
'SAGE Publications', 2020Co-Authors: Barreda Vayá Maria, Dolz, Manuel F., Castaño Álvarez, María AsunciónAbstract:Modeling the performance and energy consumption of the sparse matrix-vector product (SpMV) is essential to perform off-line analysis and, for example, choose a Target Computer architecture that delivers the best performance-energy consumption ratio. However, this task is especially complex given the memory-bounded nature and irregular memory accesses of the SpMV, mainly dictated by the input sparse matrix. In this paper, we propose a Machine Learning (ML)-driven approach that leverages Convolutional Neural Networks (CNNs) to provide accurate estimations of the performance and energy consumption of the SpMV kernel. The proposed CNN-based models use a blockwise approach to make the CNN architecture independent of the matrix size. These models are trained to estimate execution time as well as total, package, and DRAM energy consumption at different processor frequencies. The experimental results reveal that the overall relative error ranges between 0.5% and 14%, while at matrix level is not superior to 10%. To demonstrate the applicability and accuracy of the SpMV CNN-based models, this study is complemented with an ad-hoc time-energy model for the PageRank algorithm, a popular algorithm for web information retrieval used by search engines, which internally realizes the SpMV kernel
Maria Barreda - One of the best experts on this subject based on the ideXlab platform.
-
Convolutional neural nets for estimating the run time and energy consumption of the sparse matrix-vector product:
The International Journal of High Performance Computing Applications, 2020Co-Authors: Maria Barreda, Manuel F. Dolz, M. Asunción CastañoAbstract:Modeling the performance and energy consumption of the sparse matrix-vector product (SpMV) is essential to perform off-line analysis and, for example, choose a Target Computer architecture that del...
