Breaking News
Design How-To

How to accelerate genomic sequence alignment 4X using half an FPGA

More Related Links
View Comments: Newest First | Oldest First | Threaded View
User Rank
re: How to accelerate genomic sequence alignment 4X using half an FPGA
gah4   7/7/2011 8:03:19 AM
Interesting article, but a little behind the current needs. Maybe not so far off for proteins, but for DNA the needs are in the 100TCUPS range. Today, a single DNA sequencer produces 3e9 base/day. To compare against the 3e9 base human genome requires 9e18 CUPD, or about 100TCUPS. First, verilog isn't that hard to use. At the scale needed for DNA, 75% isn't good enough. Note also that the DNA data has been growing faster than Moore's law for many years, with a growth rate of about 1%/week. I have been working on systolic arrays in verilog since the XC4000 days. It is even worthwhile to adjust the datapath widths depending on the need for a given data set, down to about six bits for current sequencers. If you don't like verilog, design a symbolic language for systolic array construction, and a program to translate to verilog. Much easier than trying to coerce C into doing something it wasn't meant to do. -- glen

Most Recent Comments
August Cartoon Caption Winner!
August Cartoon Caption Winner!
"All the King's horses and all the KIng's men gave up on Humpty, so they handed the problem off to Engineering."
Top Comments of the Week
Like Us on Facebook Parts Search

185 million searchable parts
(please enter a part number or hit search to begin)
EE Times on Twitter
EE Times Twitter Feed
Flash Poll
David Patterson, known for his pioneering research that led to RAID, clusters and more, is part of a team at UC Berkeley that recently made its RISC-V processor architecture an open source hardware offering. We talk with Patterson and one of his colleagues behind the effort about the opportunities they see, what new kinds of designs they hope to enable and what it means for today’s commercial processor giants such as Intel, ARM and Imagination Technologies.