Intel® Cluster Studio

Intel® Parallel Studio XE 2016 Beta program has started!

The Intel® Parallel Studio XE 2016 Beta program is now available!

In this beta test, you will have early access to Intel® Parallel Studio XE 2016 products and the opportunity to provide feedback to help make our products better. Registration is easy through the pre-Beta survey site.

This suite of products brings together exciting new technologies along with improvements to Intel’s existing software development tools:

MPI: polling 'passive' rma operations

Hi,

lately I'm wondering if your implementation of the passive target communication was ever really ment for usage...

Despite the fact that it isn't really passive (since one has to call some mpi functions on the target to get the mpi_win_unlock ever to return), I couldn't even figure out which mpi functions exactly must/can be invoked to achieve the flushing. In the release notes is only written:

Intel MPI, perhost, and SLURM: Can I override SLURM?

All,

(Note: I'm also asking this on the slurm-dev list.)

I'm hoping you can help me with a question. Namely, I'm on a cluster that uses SLURM and lets say I ask for 2 28-core Haswell nodes to run interactively and I get them. Great, so my environment now has things like:

SLURM_NTASKS_PER_NODE=28
SLURM_TASKS_PER_NODE=28(x2)
SLURM_JOB_CPUS_PER_NODE=28(x2)
SLURM_CPUS_ON_NODE=28

Now, let's run a simple HelloWorld on, say, 48 processors (and pipe through sort to see things a bit better):

mpitune get "could not dump the session, because unknown encoding: utf-8"

Hi forum,

I try the following command on a server: (impi 5.0.2.044, icc 2015.2.164)

mpitune -of analysis.conf -application \"mpirun -n 24 -host `hostname` ./myexe\"

It did run for a while but output nothing of analysis.conf. Meanwhile the console output message like:

ERR | Could not dump the session, because unknown encoding: utf-8

I try to change LANG=C, as locale outputs:

LANG=C

LC_CTYPE="C"

...(all other environment variables are "C")

Intel MPI 5.0.2.044 and Windows firewall on localhost

Hi everybody,

We run mpiexec on Windows 7 on multiple network installations of our product in the following manner:

mpiexec.hydra.exe -np x -localroot -delegate -localonly -localhost 127.0.0.1 one_path_to_EXECUTABLE

The problem is with the windows firewall (which is executable path triggering) and alerts, because mpiexec.hydra.exe and pmi_proxy.exe do a listen on 0.0.0.0:port.

Is there an option / environment variable available to tell mpiexec.hydra only listen on 127.0.0.1 instead of all available interfaces if we run it only local on one workstation.

open_hca: getaddr_netdev ERROR: Connection refused. Is ib0 configured ERROR

Dear all,

I am a beginner in linux; i am using quantum espresso software; The cluster we have at University has three types of partition short (for jobs within an hour), long (jobs within 4 days) and superlong (jobs within 10 days). each node has 8 processors; however recently when I am running a job, only the SHORT partition works properly; this is not very useful for me as I need to run longer jobs. when i run the other two (long and superlong) I get several errors: running on more than one node say : 16 processors (2 nodes) producesan error:

Intel MPI Benchmarks Archives

I wish to use Intel MPI benchmarks for performance analysis of my mpich2 implementation. However, I'm using mpich2-1.4.1 version in my cluster. Where can I download the appropriate benchmarks for this version of mpich2? The latest version is not compatible with the mpich2 version I use.

Please help.

Subscribe to Intel® Cluster Studio