ENTERTAINMENT

Xilinx Virtex-4 FPGA-based hardware acceleration opens door to high-resolution 3-D video in real-time: At the IFSEC show today, Xilinx, in collaboration with Dimensional Imaging, Ltd., and the FPGA High Performance Computing Alliance (FHPCA), showcased its latest Virtex-4 FPGA based technology demonstration for high-speed computing. The new technology leverages the high-performance features of Xilinx Virtex-4 devices to accelerate compute-intensive 3D and 4D digital imaging algorithms by up to 12 times, enabling supercomputing performance in a compact platform that communicates with a host PC through a PCI Express(R) interface. Target applications include facial analysis and recognition, biometrics, digital entertainment and advanced orthodontics. Today's news opens the door for non-computer specialists to take advantage of the performance of supercomputer clusters and provides tremendous improvement in multiple applications including homeland security, motion picture production, digital gaming, and advanced medical treatments. Attendees of this week's IFSEC can view the demonstration in action at Silica's Stand #18177, Hall 12, at the National Exhibition Centre (NEC) in Birmingham, United Kingdom. "In 3D imaging applications, FPGA-based computing will enable greater than 12 times faster execution of Dimensional Imaging's DI3D(TM) tools, allowing analyses for applications ranging from medical treatments to security biometrics to be performed significantly faster than a conventional software processing model," said Dr. Colin Urquhart, chief executive officer at Dimensional Imaging. "For image processing on full-motion video, this technology will enable dramatic enhancements in digital entertainment by allowing rapid, accurate modeling of human features to enhance special effects. Other applications will include life-like simulations of real actors in electronic games." Dimensional Imaging's DI3D(TM) technology uses standard digital still cameras (DSCs) to capture one or more stereo pairs of images. The FPGA-based hardware processor uses the DSC data to generate a high resolution 3D surface image, which can be easily recorded or manipulated in a variety of ways. Processing in real-time is achievable, enabling 4D -- or 3D video -- performance at full frame rates.