ARINC Inaugurates New Supercomputer
The Advanced Systems Engineering and Integration Division of ARINC Engineering Services today announced it is offering an expanded supercomputing capability at its Annapolis, Maryland headquarters.
A new Linux-based supercomputer has been installed and is now available to support a wide range of engineering analyses for ARINCs external customers. The companys previous supercomputer also remains available.
ARINCs supercomputer facility has performed advanced modeling and analysis of antennas, communications and navigation systems since the 1990s. The company has also developed proprietary software technology to handle demanding analysis requirements and permit the modeling of large systems for commercial, government, and military customers.
Our extensive experience has given ARINC one of the worlds most advanced capabilities for computer modeling of antennas and electromagnetic effects, states Jim Potter, ARINC Director, System Engineering & Analysis.
The new Linux supercomputer is a multiprocessor cluster, consisting of 14 compute [sic] nodes, each with multiple high-speed multi-core processors. The nodes are interconnected by a high-speed local area network and have a total of 448 Gigabytes of available memory for solving computationally large problems.
This arrangement allows parallel processing of large problems such as calculating radar cross-sections, or finding the radiation patterns of antennas mounted on large aircraft and ships, notes Potter.
Engineers realize significant savings by modeling and simulating the real-world performance of antennas and electromagnetic systems on a supercomputer, instead of constructing prototypes and making measurements on physical systems. Computer-based performance modeling optimizes the design of new antennas and their mounting locations early in the design process, ensuring that performance objectives are met when the systems are installed on aircraft, communications towers, ships, and vehicles.
The new supercomputer can now fully implement ARINCs analytical toolset, including the Geometric Theory of Diffraction and the more broadly applicable Method of Moments, Physical Optics, and Finite Element methods, along with hybrid implementations. This significantly increases both the size of the objects that can be modeled and the frequencies at which they can be analyzed. It also greatly reduces the computational time needed to solve simpler problems.
ARINCs supercomputer facility has performed advanced modeling and analysis of antennas, communications and navigation systems since the 1990s. The company has also developed proprietary software technology to handle demanding analysis requirements and permit the modeling of large systems for commercial, government, and military customers.
Our extensive experience has given ARINC one of the worlds most advanced capabilities for computer modeling of antennas and electromagnetic effects, states Jim Potter, ARINC Director, System Engineering & Analysis.
The new Linux supercomputer is a multiprocessor cluster, consisting of 14 compute [sic] nodes, each with multiple high-speed multi-core processors. The nodes are interconnected by a high-speed local area network and have a total of 448 Gigabytes of available memory for solving computationally large problems.
This arrangement allows parallel processing of large problems such as calculating radar cross-sections, or finding the radiation patterns of antennas mounted on large aircraft and ships, notes Potter.
Engineers realize significant savings by modeling and simulating the real-world performance of antennas and electromagnetic systems on a supercomputer, instead of constructing prototypes and making measurements on physical systems. Computer-based performance modeling optimizes the design of new antennas and their mounting locations early in the design process, ensuring that performance objectives are met when the systems are installed on aircraft, communications towers, ships, and vehicles.
The new supercomputer can now fully implement ARINCs analytical toolset, including the Geometric Theory of Diffraction and the more broadly applicable Method of Moments, Physical Optics, and Finite Element methods, along with hybrid implementations. This significantly increases both the size of the objects that can be modeled and the frequencies at which they can be analyzed. It also greatly reduces the computational time needed to solve simpler problems.
.png)
.jpg){kind=logo}
Comments
There are no comments yet for this item
Join the discussion