No best tree generated when I use AVX on an HPC system.
72 views
Skip to first unread message
Ramona Walls
Jul 28, 2015, 7:58:56 PM7/28/15
to raxml
I am running RAxML on the Stampede cluster at TACC. I have compiled three versions: raxmlHPC-HYBRID-AVX, raxmlHPC-MPI-AVX, raxmlHPC-SSE3.
When I conduct multiple searches for the best tree using SSE3, I get a best tree (e.g., RAxML_bestTree.T6). When I do the same exact analysis using MPI-AVX or HYBRID AVX, still get all the output _log, _parsimonyTree, and _result files, but no bestTree file.
Here is an example command:
../raxmlHPC-MPI-AVX -m GTRGAMMA -p 12345 -s dna.phy -# 20 -T 16 -n T6
I have tried this with several input files, including dna.phy, and it is reproducible.
I would really like to be able to use HYBRID AVX, because it is considerably faster in the benchmarks I ran.
When I conduct multiple searches for the best tree using SSE3, I get a best tree (e.g., RAxML_bestTree.T6). When I do the same exact analysis using MPI-AVX or HYBRID AVX, still get all the output _log, _parsimonyTree, and _result files, but no bestTree file.
Here is an example command:
../raxmlHPC-MPI-AVX -m GTRGAMMA -p 12345 -s dna.phy -# 20 -T 16 -n T6
I have tried this with several input files, including dna.phy, and it is reproducible.
I would really like to be able to use HYBRID AVX, because it is considerably faster in the benchmarks I ran.
Alexandros Stamatakis
Jul 29, 2015, 3:06:57 AM7/29/15
to ra...@googlegroups.com
that's weird, but without the associated error message that you get for
the AVX version I can't help you,
alexis
> --
> You received this message because you are subscribed to the Google
> Groups "raxml" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to raxml+un...@googlegroups.com
> <mailto:raxml+un...@googlegroups.com>.
> For more options, visit https://groups.google.com/d/optout.
--
Alexandros (Alexis) Stamatakis
Research Group Leader, Heidelberg Institute for Theoretical Studies
Full Professor, Dept. of Informatics, Karlsruhe Institute of Technology
Adjunct Professor, Dept. of Ecology and Evolutionary Biology, University
of Arizona at Tucson
www.exelixis-lab.org
the AVX version I can't help you,
alexis
> You received this message because you are subscribed to the Google
> Groups "raxml" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to raxml+un...@googlegroups.com
> <mailto:raxml+un...@googlegroups.com>.
> For more options, visit https://groups.google.com/d/optout.
--
Alexandros (Alexis) Stamatakis
Research Group Leader, Heidelberg Institute for Theoretical Studies
Full Professor, Dept. of Informatics, Karlsruhe Institute of Technology
Adjunct Professor, Dept. of Ecology and Evolutionary Biology, University
of Arizona at Tucson
www.exelixis-lab.org
Ramona Walls
Jul 29, 2015, 8:07:03 PM7/29/15
to ra...@googlegroups.com
There are no errors (i.e. the standard error files are empty), but as you can see in the attached standard output files, the SSE run generates a best tree, but the MPI-AVX and HYB-AVX runs stop after writing the results files.
Ramona------------------------------------------------------
Ramona L. Walls, Ph.D.
Scientific Analyst, The iPlant Collaborative, University of Arizona
Research Associate, Bio5 Institute, University of Arizona
Laboratory Research Associate, New York Botanical Garden
Ramona L. Walls, Ph.D.
Scientific Analyst, The iPlant Collaborative, University of Arizona
Research Associate, Bio5 Institute, University of Arizona
Laboratory Research Associate, New York Botanical Garden
--
You received this message because you are subscribed to a topic in the Google Groups "raxml" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/raxml/lRO886sFMQk/unsubscribe.
To unsubscribe from this group and all its topics, send an email to raxml+un...@googlegroups.com.
Alexandros Stamatakis
Jul 30, 2015, 2:48:13 AM7/30/15
to ra...@googlegroups.com
dear ramona,
could you maybe update to raxml v 8.2.0, several bugs have been fixed in
this version,
alexis
On 30.07.2015 02:06, Ramona Walls wrote:
> There are no errors (i.e. the standard error files are empty), but as
> you can see in the attached standard output files, the SSE run generates
> a best tree, but the MPI-AVX and HYB-AVX runs stop after writing the
> results files.
>
> Ramona
>
> ------------------------------------------------------
> Ramona L. Walls, Ph.D.
> Scientific Analyst, The iPlant Collaborative, University of Arizona
> Research Associate, Bio5 Institute, University of Arizona
> Laboratory Research Associate, New York Botanical Garden
>
> On Wed, Jul 29, 2015 at 12:06 AM, Alexandros Stamatakis
> <alexandros...@gmail.com
> <mailto:raxml%2Bunsu...@googlegroups.com>
> <mailto:raxml+un...@googlegroups.com
> <mailto:raxml%2Bunsu...@googlegroups.com>>.
could you maybe update to raxml v 8.2.0, several bugs have been fixed in
this version,
alexis
On 30.07.2015 02:06, Ramona Walls wrote:
> There are no errors (i.e. the standard error files are empty), but as
> you can see in the attached standard output files, the SSE run generates
> a best tree, but the MPI-AVX and HYB-AVX runs stop after writing the
> results files.
>
> Ramona
>
> ------------------------------------------------------
> Ramona L. Walls, Ph.D.
> Scientific Analyst, The iPlant Collaborative, University of Arizona
> Research Associate, Bio5 Institute, University of Arizona
> Laboratory Research Associate, New York Botanical Garden
>
> On Wed, Jul 29, 2015 at 12:06 AM, Alexandros Stamatakis
> <alexandros...@gmail.com
> <mailto:raxml+un...@googlegroups.com
> <mailto:raxml%2Bunsu...@googlegroups.com>>.
> For more options, visit https://groups.google.com/d/optout.
>
>
> --
> Alexandros (Alexis) Stamatakis
>
> Research Group Leader, Heidelberg Institute for Theoretical Studies
> Full Professor, Dept. of Informatics, Karlsruhe Institute of Technology
> Adjunct Professor, Dept. of Ecology and Evolutionary Biology, University
> of Arizona at Tucson
>
> www.exelixis-lab.org <http://www.exelixis-lab.org>
>
>
> --
> Alexandros (Alexis) Stamatakis
>
> Research Group Leader, Heidelberg Institute for Theoretical Studies
> Full Professor, Dept. of Informatics, Karlsruhe Institute of Technology
> Adjunct Professor, Dept. of Ecology and Evolutionary Biology, University
> of Arizona at Tucson
>
>
> --
> You received this message because you are subscribed to a topic in
> the Google Groups "raxml" group.
> To unsubscribe from this topic, visit
> https://groups.google.com/d/topic/raxml/lRO886sFMQk/unsubscribe.
> To unsubscribe from this group and all its topics, send an email to
> raxml+un...@googlegroups.com
> <mailto:raxml%2Bunsu...@googlegroups.com>.
> --
> You received this message because you are subscribed to a topic in
> the Google Groups "raxml" group.
> To unsubscribe from this topic, visit
> https://groups.google.com/d/topic/raxml/lRO886sFMQk/unsubscribe.
> To unsubscribe from this group and all its topics, send an email to
> raxml+un...@googlegroups.com
Ramona Walls
Jul 30, 2015, 12:22:55 PM7/30/15
to ra...@googlegroups.com
Ok. I will do that and see if it fixed the problem.
Thanks!
Enviado de Baja Arizona.
> To unsubscribe from this group and all its topics, send an email to raxml+un...@googlegroups.com.
Thanks!
Enviado de Baja Arizona.
Alexandros Stamatakis
Jul 31, 2015, 3:57:45 AM7/31/15
to ra...@googlegroups.com
yes, thank you,
alexis
alexis
Reply all
Reply to author
Forward
0 new messages