[Beowulf] RE: [Bioclusters] FPGAin bioinformatics clusters (again?)

[Joe Landman]

Mike Davis wrote:
> But BLAST is only a small part and argueably the easiest part of 
> genomics work. The advantages of parallelization and/or smp come into 
> play when attempting to assemble the genome. Phred/Phrap can do the work 
> but starts to slow even large machines when your talking 50k+ of 
> sequences (which it wants to be in one folder). A quiz for  the Unix 
> geeks out there, what happens when a folder has 50,000 files in it. Can 
> you say SLOOOOOOOOOWWWW?

Depends upon the file system.  If you use something like NFS, NTFS, FAT, 
HFS, ext*, etc, yes, it will be slow.  Building/bundling fast file 
systems is hard, and some dominant North American linux vendors have not 
chosen their supported offerings with HPC end users issues in mind, so 
you are stuck with their choices.

With regards to accelerating applications, it might help to know which 
ones are on peoples top 10 lists.  We have been working with a number of 
groups and have built accelerated BLAST and HMMer code as a result of 
this work.  So far we have heard of interest in 
ClustalW/tcoffee/muscle/...  and few others.



-- 
Joseph Landman, Ph.D
Founder and CEO
Scalable Informatics LLC,
email: landman at scalableinformatics.com
web  : http://www.scalableinformatics.com
phone: +1 734 786 8423
fax  : +1 734 786 8452
cell : +1 734 612 4615