Friday, January 8, 2010

Impulse C biotechnology DNA model sequencing algorithm accelerated on Pico Computing, Xilinx FPGA platform

KIRKLAND, USA: Impulse Accelerated Technologies announced that its CoDeveloper C-to-FPGA compiler has been used to develop and deploy a DNA biotechnology genome search algorithm on a new, modular FPGA acceleration platform created by Pico Computing, of Seattle, Washington.

The Xilinx FPGA-accelerated bioinformatics algorithm, written in C, performs sequence analysis and pair-match scoring of DNA runs with lengths of 25 base pairs. The algorithm is inherently parallel, making it an ideal candidate for acceleration using a large number of FPGA devices.

Within two days of powering up the EX-500 with the first Virtex-6 SX240T FPGA computing modules, Pico Computing engineers had successfully migrated and tested the C-language bioinformatics searching algorithm and demonstrated significant improvements in algorithm speed and throughput over previous generation FPGA products.

“This bioinformatics algorithm created by Pico Computing is a solid example of how higher-level design tools such as Impulse C can speed the development and deployment of complex algorithms”

“This bioinformatics algorithm created by Pico Computing is a solid example of how higher-level design tools such as Impulse C can speed the development and deployment of complex algorithms,” said Brian Durwood, co-Founder and COO of Impulse. “Our Platform Support Package integration with multiple Pico Computing products, including the EX-500, allowed this algorithm to be moved with little difficulty onto platforms ranging from single-FPGA cards, to FPGA clusters based on Xilinx Spartan, Virtex-5 and now the new Virtex-6 FPGAs.”

Using the new, modular EX-500 platform as the foundation, and Impulse C as the programming tool, Pico Computing software and firmware engineers were able to successfully deploy the bioinformatics algorithm with no significant changes needed to the algorithm source code. The Pico EX-500 supports up to six M-501 modules per PCI Express board.

Each module can be configured with a range of FPGAs including the Xilinx Virtex-6 LX240T devices used in this first test application. Pico Computing will demonstrate this algorithm at the Plant and Animal Genome Conference exhibit area on January 10-12, booth #426.

“The M-501 modules with Virtex-6 FPGAs satisfy demands from our customers for increasing levels of computing performance, in applications that range from data security to life sciences and financial modeling,” said Mark Hur, Director of Sales and Marketing for Pico Computing.

“Our government and commercial customers depend on our ability to supply the most advanced acceleration solutions, using the most up-to-date FPGA technologies and the most advanced tools for FPGA programming. We are happy to be partnering with Impulse on this and other applications.”

Impulse is also partnering with the University of Washington and other schools on Tomography, Radiology and many other medical and medical imaging projects.

No comments:

Post a Comment

Note: Only a member of this blog may post a comment.