Mercury Computer Systems Quadruples Processing Density of Image and Signal Processing Systems - the RACE VME signal and imaging processing system - Product Announcement

Edge: Work-Group Computing Report, Sept 6, 1999

Mercury Computer Systems, Inc. Tuesday announced its new RACE Series VME multicomputer systems based on Motorola's newest PowerPCTM microprocessor, the MPC7400 with AltiVec technology. Providing processing densities of up to 130 GFLOPS per cubic foot, these systems quadruple the computational performance of high-end signal and image processing systems previously available for applications such as medical imaging and military signal processing.

This increase in processing density enables more real-time processing to be performed closer to the source of the data, thereby shortening the time to response. For medical imaging, this could enable advances such as interactive interventional surgery, where the surgeon views the images in real time during an operation. For defense applications, this can lead to forwardly deployed systems performing automated target recognition, moving target classification, and dynamic mission control. For the broadcast industry, this processing density can enable high-accuracy encoding, decoding, and manipulation of MPEG image streams at HDTV resolutions in real time all in a single computer chassis.

"Mercury's expertise in designing high-performance embedded computer systems directly complements the PowerPC with AltiVec's strengths in vector processing," said Will Swearingen, Director of PowerPC Marketing for Motorola's Networking and Computing Systems Group. "Mercury's deep understanding of our AltiVec architecture ensures that these systems will use the MPC7400 to its full capabilities."

Mercury's products based on Motorola's MPC7400 are supported with a full suite of runtime and development software. The most important new software is a set of signal processing functions added to Mercury's Scientific Algorithm Library (SAL), which has been hand-coded and hand-tuned for the AltiVec vector execution unit.

"The full potential of the AltiVec technology can best be realized by keeping the data continuously flowing through the vector unit," commented Richard Jaenicke, Director of Product Marketing for Mercury. "The SAL signal processing functions are carefully and effectively optimized to allow application developers to achieve maximum performance advantage from this impressive new generation of microprocessor technology. This optimization has already resulted in increased interest for our library of signal processing functions and our products as a whole."

The application development environment for RACE VME systems is based on open standards. The VSIPL-Lite library from Mercury allows developers to access the industry-benchmark performance of the SAL library through the VSIPL open standard API. Mercury is also developing a high-performance implementation of the MPI/RT real-time multiprocessor communication standard. Together, these standard APIs provide open access to Mercury's multicomputer systems based on the ANSI/VITA RACEway Interlink standard.

RACE VME systems provide a multiprocessor hardware architecture that balances the vector processing capabilities of the MPC7400 microprocessor with the high-bandwidth, low-latency data transfer capabilities of the RACEway switched communication fabric. Each set of 18 processors can be connected with a communication bandwidth totaling 2.4 gigabytes per second. Up to 288 processors can be connected in a single VME chassis with an aggregate bandwidth of over 38 gigabytes per second. This bandwidth can be directly applied to the most challenging applications such as space-time adaptive processing (STAP), synthetic aperture radar (SAR), and classified adaptive signal processing.

The RACE Series VME systems based on the MPC7400 are Mercury's fourth generation of products using PowerPC microprocessors. These VME systems are the first products in a series to follow the new PowerPC architecture roadmap. Mercury will also field AltiVec-based products for its PCI and proprietary systems. For more information on the RACE architecture and RACE Series VME systems using the MPC7400, see http://www.mc.com. For more information on the MPC7400, see http://www.mot.com/SPS/PowerPC/.> Mercury Computer Systems, Inc., is the leading independent producer of high-performance digital signal and image processing computer systems that transform sensor data to information for analysis and interpretation on a real-time basis. Mercury's products play a critical role in a wide range of defense and medical imaging applications. In air-, sea-, and land-based military platforms, these systems process real-time radar, sonar, and signal intelligence data. Mercury's systems are also used in state-of-the-art medical diagnostic devices, including magnetic resonance imaging (MRI), computed tomography (CT), and the rapidly growing field of digital X-ray.

Mercury has also developed a range of shared storage software products that allow multiple systems and servers to share files and data using Fibre Channel at data rates up to and exceeding 100 MB/sec. These products are marketed for use in the broadcast, entertainment, and digital prepress industries, and for a variety of enterprise computing applications.


 

BNET TalkbackShare your ideas and expertise on this topic

Please add your comment:

  1. You are currently: a Guest |
  2.  

Basic HTML tags that work in comments are: bold (<b></b>), italic (<i></i>), underline (<u></u>), and hyperlink (<a href></a)

advertisement
CXO UnpluggedSmart Business interviews on BNET

See and hear how senior level executives across the Asia Pacific are developing smart business ideas across a variety of sectors. The focus is on the future, and on how businesses need to evolve.

advertisement
  • Click Here
  • Click Here
  • Click Here
advertisement

Content provided in partnership with Thompson Gale