Air Force lab’s new 20-Tflop computer will mimic human brain
- By Defense Systems Staff
- Jan 14, 2014
The U.S. Air Force Research Laboratory Information Directorate’s High Performance Systems Branch has placed an order with GE Intelligent platforms for a High Performance Embedded Computing (HPEC) system designed to simulate information pathways of the human nervous system.
The GE system is designed to support DOD’s High Performance Computing Modernization Program (HPCMP) and will benefit the development of next-generation radar technology such as Gotcha wide-area synthetic aperture radar, Military & Aerospace Electronics reported. The Gotcha wide-area synthetic aperture radar (SAR) processes single radar mode data into data products such as super-resolution 2D imagery, video SAR, 3D SAR imagery, and ground moving target indication (GMTI) with minimum detectable velocity (MDV).
The system is capable of maintaining 20 teraflops, or 20 trillion floating-point operations per second, and is housed in a scalable 6U OpenVPXrack mount chassis that can be supplemented with additional compute nodes and racks.
The HPEC system supplied to AFRL utilizes graphics processing unit (GPU) accelerators created by Nvidia Corp., GE said in an announcement. http://www.businesswire.com/news/home/20140107005879/en/GE-Receives-Order-AFRL-20-Teraflop-Rugged#.UtWT2NJdU25 The GPU units allow for the use of parallel processing, providing real-time processing for high-bandwidth radar data through advanced neuromorphic architectures.
Neuromorphic computing allows for machines to autonomously process complex, real-world information by adaptively learning relevant features in their environments and determining their relationships. The GE system will enable the development and deployment of neuromorphic algorithms for large-scale dynamic data analysis.
Rod Rice, general manager for military and aerospace products at GE Intelligent Platforms, said “GPUs deliver an unbeatable combination of very high performance computing with minimal power consumption and heat dissipation in constricted spaces – characteristics that were key to AFRL awarding GE this order.”