[Bioclusters] Use of FPGA based BLAST accelerators?

[Elia Stupka]

Dear Chris,

    very interested in your mail, since I never got a chance to play with
the TimeLogic systems... I have a few questions.

It seems clear that for things like hmmer, genewise,etc. there is clear
advantages to using their systems... sorry for my ignorance what I am not
clearbut if have a new algorithm coded yesterday, how will it perform "as
is" on these systems, and how much work (and how cooperative is TimeLogic)
in making it screaming fast? What worries me is that on all systems with
enough coding sweat (as discussed earlier on this lsit) one can speed up
algorithms quite substantially (see hmmer 6x speed up on Altivech)...

> Our usual model is to prototype a processing pipeline on the cluster,
> and then migrate the CPU intensive portions to the TimeLogic.  This
> combines the flexibility of the cluster with the power of the
> specialized machine.

Would you agree that in order to run things practically you wouldn't really
want to "just" have the TimeLogic machine and not the normal cluster? Also,
how much extra nuisance/time do you think this two-way approach asks for?
i.e. in a standard system there is no lag between prototyping and running,
because if it works on one of the cluster CPUs, it will work on the rest (of
course with a good sysadmin)

Finally, at our end, we like to have a fully automated pipeline that works
with LSF/SGE/etc and takes care of each job, input, output, etc. How
easy/difficult do you see this sort of system on the TimeLogic machine? i.e.
is it a black box with its proprietory buttons that I need to push, or can I
go in and build and use my own buttons?

Sorry, lots of questions, but I am really trying to understand how it would
be to have one as part of a large analysis cluster and the only little (bad)
experience I had was with the paracel box James Cuff described when I was at
Ensembl, which really cannot be plugged in with the rest of the system.....

Thanks,

Elia