Dear Galen,
It actually turns out that there is a problem not only with
MPI_Alltoall_Isend_Irecv, but also with another related operation
insyncol_MPI_Alltoallv-nodes-long-SM.ski (this what seems to be holding
down the FFTs, I checked the source code, and it uses alltoallv):
#/*@insyncol_MPI_Al
Hi Konstantin,
MPI_Alltoall_Isend_Irecv
This is a very unscalable algorithm in skampi as it simply posts N
MPI_Irecv's and MPI_Isend's and then does a Waitall. We shouldn't
have an issue though on 8 procs but in general I would expect the
performance of this algorithm to degrade quite q
Sorry, I forgot to specify everything properly in my previous e-mail:
mpirun -np 8 -mca btl tcp -mca coll self,basic,tuned -mca \
mpi_paffinity_alone 1 skampi41
#/*@insyncol_MPI_Alltoall-nodes-long-SM.ski*/
2 256.4 2.8 8 256.4 2.8 8
32153.0 13.0
Dear Jeff and Galen,
I have tried openmpi-1.1a1r8890. The good news is that it seems like
the freaky long latencies for certain packet sizes went away with the
options they showed up with before. Also, one version of all-to-all
appears to behave nicer with a specified set of parameters. However,
Greetings Konstantin.
Many thanks for this report. Another user submitted almost the same
issue earlier today (poor performance of Open MPI 1.0.x collectives;
see http://www.open-mpi.org/community/lists/users/2006/02/0558.php).
Let me provide an additional clarification on Galen's reply:
Hello Konstantin,
By using coll_basic_crossover 8 you are forcing all of your
benchmarks to use the basic collectives, which offer poor
performance. I ran the skampi Alltoall benchmark with the tuned
collectives I get the following results which seem to scale quite
well, when I have a bit
Hi all,
There seem to have been problems with the attachement. Here is the
report:
I did some tests of Open-MPI version 1.0.2a4r8848. My motivation was
an extreme degradation of all-to-all MPI performance on 8 cpus (ran
like 1 cpu). At the same time, MPICH 1.2.7 on 8 cpus runs more like on
4 (
Hi all,
Please see the attached file for a detailed report of Open-MPI
performance.
Any fixes in the pipeline for that?
Konstantin
__
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around
http://mail.yahoo.com
repor
