GPUs

Unprecedented Performance

GPUs (Graphical Processing Units), offer very high GFLOP/Watt performance. Although traditionally used in laptops/desktops for gaming applications, GPUs feature superior computation and memory bandwidth over general purpose microprocessors. They excel at highly parallel, compute intense applications, such as streaming signal/image applications With recent architectural advancements, the algorithmic scope to which GPUs can be applied has grown dramatically. For traditional signal processing algorithms like the FFT (Fast Fourier Transform), they provide unprecedented performance, particularly performance per watt.

Improving Processing and Exploitation

GPUs can now be leveraged against the processing and exploitation requirements of growing markets like signals intelligence and oil & gas exploration, with outstanding results. Industry performance benchmarks implemented on GPUs in high-performance signal processing applications have shown that GPUs can obtain 20x performance improvement and more over other processors.

Technology options for high-performance computing
        

Sensor Stream Computing Platform
VXS-based development system that harnesses the tremendous processing power of commercial GPUs for high performance, data-parallel computing
Multi-Core Video
Learn about future compute accelerator solution architectures
E-Mail
 for more product related information

By Phone
Toll-free: 866.627.6951
Fax: 978.256.3599

Accelerating Applications

Compute accelerators with multiple processing cores can provide unprecedented levels of performance for computationally intensive applications. Examples of compute accelerating technology include field-programmable gate arrays (FPGAs), graphics processing units (GPUs), and multicore microprocessors such as the Cell BE Processor.

Mercury's software tools and libraries, professional services, and hardware acceleration technologies can significantly accelerate mathematical computations for complex algorithms across these technologies. Mercury works with customers in medical imaging, electronic design automation (EDA), semiconductor, aerospace and defense, seismic processing, and video processing markets to eliminate bottlenecks in their applications.

To connect the hardware together and optimize the processor-to-processor I/O, we employ a variety of switch fabrics including RapidIO, RACE++®, switched PCI Express®, InfiniBand®, and Gigabit Ethernet.
Copyright © 2010 Mercury Computer Systems, Inc. All rights reserved.