2/3/2004 - BAE Systems and Celoxica Ltd. announced a strategic partnership to deliver next generation technology and design solutions exploiting compact, yet highly scalable parallel processing elements within electronic systems. The two companies will join expertise to supply advanced system design tools and prototyping solutions for real-time Image Processing, Pattern Matching and other massively data intensive applications in the industrial automation, bio-informatics, medical, biometrics, electronic security and defence markets.
The partnership builds on previous collaborations between the two companies and is the dividend of research and development efforts of both BAE Systems and Celoxica. The new solutions integrate LEARRNN (Logic Enabled Asynchronous Rapid Robust Neural Network) weightless neural network technology, supplied by BAE Systems, into the System FPGA design environment supplied by Celoxica through its latest DK Design Suite of system design tools.
As demand and pressure increases for real-time comparisons, fault detection and data interpretation in complex applications such as biometric identification, machine vision, and transaction fraud detection, designers are looking for new solutions over and above existing technology. Based on a pure logic implementation of Weightless Neural Networks, LEARRNN offers distinct and unique advantages over other Neural Network technologies.
“We’re seeing enormous growth and interest for improved performance in the Imaging and Security markets,” said Bob Flint, Director of BAE Systems Ventures Ltd, “particularly where applications need economic and compact implementations of massively parallel pattern matching on incomplete data, or real-time signal or video image processing. This partnership with Celoxica can help us reach these markets by providing the design path to implementation of LEARRNN in hardware.”
Weightless Neural Network technology is ideal for use in real-time filtering and interpretation of 2D and 3D digital data. It can be used to remove noise or unwanted data from a complex signal, or to assist in the interpretation of the information being conveyed by the signal itself. Specific advantages include very high-speed processing with the ability to handle massively parallel data (for example 3D data). Implementations are able to filter noisy data, and the technology is readily extensible to pattern recognition, fault/difference detection and diagnosis/interpretation.
Driven by the design aims of scalability, resilience and high speed, applications can be implemented in Boolean logic without clocks or arithmetic processes, and can store exemplars with just one-pass training. This increases the speed of operation and produces a system that is inherently fault tolerant. The weightless neural network approach is applicable to the manipulation of a wide range of data, particularly where the data is noisy and difficult to interpret.
“Currently available solutions simply don’t provide the efficiencies and performance customers need to remain competitive,” Bob Flint continued. “By combining our LEARRNN technology with the Celoxica DK design environment we can provide our customers with a clear competitive advantage and the capability to differentiate their products and services significantly.”
Leveraging Celoxica’s tools and Software Compiled System Design methodology, the partnership will improve design productivity by exploiting system level design techniques and the consistent level of design abstraction that C-based design, verification and synthesis can offer. Algorithms can be partitioned, evaluated and compiled directly to efficient hardware using Celoxica’s DK Design Suite. All basic LEARRNN functions and building blocks are available as a toolbox for MATLAB and can be simulated with custom IP designed in the DK design environment using Celoxica’s Nexus-PDK co-simulation manager.
“Bringing together expertise and experience from both companies makes a lot of sense and will deliver an unbeatable combination to those choosing to use LEARRNN, “said Jeff Jussel, VP Marketing for Celoxica. “Linking our proven DK Design Suite and market segment expertise with BAE Systems’ technology provides a compelling solution for companies and designers developing the next generation products with the most challenging requirements and design schedules.”
About BAE SYSTEMS
BAE Systems is a systems company, innovating for a safer world. BAE Systems employs nearly 90,000 people including Joint Ventures, and has annual sales of around £12 billion. The company offers a global capability in air, sea, land and space with a world-class prime contracting ability supported by a range of key skills. BAE Systems designs, manufactures and supports military aircraft, surface ships, submarines, radar, avionics, communications, electronics, guided weapon systems and a range of other defence products. BAE Systems is dedicated to making the intelligent connections needed to deliver innovative solutions.
An innovator in system-level electronic design automation (EDA), Celoxica supplies the design technology, IP and services that define Software-Compiled System Design, a methodology that exploits higher levels of design abstraction to dramatically improve silicon design productivity. Celoxica’s products address hardware/software partitioning, co-verification and C-based synthesis to reconfigurable hardware. Established in 1996, Celoxica offers a proven route from complex software algorithms to hardware, and provides an ideal design environment for System FPGA with significant productivity advantages for digital signal processing applications such as imaging, electronic security and communications.
Celoxica, Handel-C and the Celoxica logo are trademarks of Celoxica, Ltd. LEARRNN is a trademark of BAE Systems plc.
For sales enquiries contact: Americas – firstname.lastname@example.org; APAC – email@example.com; Europe, Middle East, Africa – firstname.lastname@example.org.
Previous Page | News by Category | News Search
If you found this page useful, bookmark and share it on: