[Bioclusters] what motherboards are people using for dualAthlon boxes?

[Joe Landman]

I should learn to post after the coffee infusion begins...  the -j switch 
is for parallel makes (which I was in the middle of when I wrote the 
message).  The -a option is correct.

On Fri, 7 Mar 2003, ben wrote:

> Joseph Landman wrote:
> 
> > Possibly on an SMP, and that would be nice to see on this platform.
> > Unlikely on HT.  For BLAST however, the Athlon does nicely, and the -j 2
> > switch gives you 50-90% additional performance (search and algorithm
> > dependent) from what I have tried myself on non-pathological cases.
> >
> 
> Hi ,
> I ' m afraid I didn' t understand .... is -j 2 parameter the one to make
> blast aware of the dual smp aystem ? I' m using ncbi blast on a 2-P3 and I
> thought it was -a 2 .. I' m asking because , after few testing with yeast db
> I don' t see a big difference in running blast with 1 or 2 cpu ; maybe I set
> the wrong parameter ? Shoud I also tell that I' m  using

Depends upon your problem, the machine, and a few other factors.  BLAST is 
memory latency sensitive, and if your memory pipe is not large enough, 
that can become a factor as well.  There are possible issues that can 
cause BLAST to not scale well using the -a switch. 

> blast with > 1 processor with some configs during the compilation ...
> I think I didn't see anything about this
> thanks in advance