[Bioclusters] more NCBI BLAST testing on opterons

Lucas Carey bioclusters@bioinformatics.org
Thu, 18 Sep 2003 12:35:05 -0400

I've built it with icc, but I no longer have the .mk file.
I seem to remember ~30 speedup for the more computationally intensive blast searches over gcc -O2 but this is all stuff I haven't touched in six months. 
If you're really searious you'll want to build in profiling info, run a few times, then build with -prof_use, but I never bothered doing this. If anyone does it I'd be interested to know how much speedup you get. I would guess that the math libraries won't help, as they tend to be more for floating point operations.

If I remember correctly, the modified lines should be:
# $Id: linux.ncbi.mk,v 1.28 2003/04/16 15:29:48 beloslyu Exp $
NCBI_LDFLAGS1 = -O3 -wp_ipo -march=pentium4 -tpp7
NCBI_OPTFLAG = -O3 -wp_ipo -march=pentium4 -tpp7

(with -wp_ipo you have to build everything you link to the toolkit with icc -wp_ipo as well)

On Thu, Sep 18, 2003 at 11:20:01AM -0400, Joseph Landman wrote:
>   We are also seeing some interesting affinity bits.  It looks like we
> need a way to pin memories and CPUs together for scheduling for best
> performance.  Has anyone successfully built the NCBI toolkit using the
> ICC?  If so, would you be willing to share your linux.ncbi.mk file? 
> Thanks!
>   Joe
> -- 
> Joseph Landman, Ph.D
> Scalable Informatics LLC
> email: landman@scalableinformatics.com
>   web: http://scalableinformatics.com
> phone: +1 734 612 4615
> _______________________________________________
> Bioclusters maillist  -  Bioclusters@bioinformatics.org
> https://bioinformatics.org/mailman/listinfo/bioclusters