Re: [OMPI users] Sandy Bridge performance question

It depends on the application you are using. Some are "balanced" - i.e., they run faster if the number of processes is a power of two. You'll see that n8 is faster than n7, so this is likely the situation. On Jun 6, 2013, at 4:10 PM, "Blosch, Edwin L" wrote: > I am

[OMPI users] Sandy Bridge performance question

I am running single-node Sandy Bridge cases with OpenMPI and looking at scaling. I'm using -bind-to-core without any other options (default is -bycore I believe). These numbers indicate number of cores first, then the second digit is the run number (except for n=1, all runs repeated 3 times).

Re: [OMPI users] Set number of processes in another way than -np X

Pvm is not dead. It's "stable". 2013/6/6 Jeff Squyres (jsquyres) > The traditional way to do this stuff in MPI is > > If rank==0 do_master_stuff > Else do_slave_stuff > > Sounds like that pattern should apply to your app. > > Pvm has been dead for years. > > Sent from

[OMPI users] OMPI Coll Framework and RDMA

Hi, I have a quick question. Is there an openib (in btl framework) equivalent in coll framework? I have an MPI application with gatherv and scatterv. I am wondering if I can leverage RDMA capabilities of the underlying Infiniband fabric. Thanks, -- Joba

Re: [OMPI users] Set number of processes in another way than -np X

The traditional way to do this stuff in MPI is If rank==0 do_master_stuff Else do_slave_stuff Sounds like that pattern should apply to your app. Pvm has been dead for years. Sent from my phone. No type good. On Jun 6, 2013, at 9:43 AM, "Ralph Castain"

Re: [OMPI users] Set number of processes in another way than -np X

What wrong with only allowing the rank 0 to execute the code before and after the funcCompScalapack function as indicated in the example below: int main() { //Initialize MPI MPI_Init(NULL,NULL); MPI_Comm_rank(MPI_COMM_WORLD, ); if( 0 == rank ) { //some work that must be

Re: [OMPI users] Oversubcription for mimd/mpmd mpiruns

On Jun 6, 2013, at 9:29 AM, "Nima Aghajari" wrote: > Hello, > first of all thanks for your reply. I tried specifying the --slot-list option > like you proposed. This will unfortunately lead to the same mpirun with 5 > cores. Adding another slot-list command for the second

Re: [OMPI users] Set number of processes in another way than -np X

I honestly don't know - you'd have to look at the PVM docs. You also might look at OpenMP and try doing it with multiple threads instead of processes, though that limits you to running on a single node. On Jun 6, 2013, at 9:37 AM, José Luis García Pallero wrote: >

Re: [OMPI users] Set number of processes in another way than -np X

2013/6/6 Ralph Castain > > On Jun 6, 2013, at 8:58 AM, José Luis García Pallero > wrote: > > 2013/6/6 Ralph Castain > >> should work! >> > > Thank you for your answer. > > So I understand that MPI_Comm_spawn() is my function. But I see

Re: [OMPI users] Set number of processes in another way than -np X

On Jun 6, 2013, at 8:58 AM, José Luis García Pallero wrote: > 2013/6/6 Ralph Castain > should work! > > Thank you for your answer. > > So I understand that MPI_Comm_spawn() is my function. But I see in the > documentation that the first argument is

Re: [OMPI users] Set number of processes in another way than -np X

2013/6/6 Ralph Castain > should work! > Thank you for your answer. So I understand that MPI_Comm_spawn() is my function. But I see in the documentation that the first argument is char* command, and command is the name of the program to spawn, but I not want to execute an

Re: [OMPI users] Set number of processes in another way than -np X

should work! On Jun 6, 2013, at 8:24 AM, José Luis García Pallero wrote: > 2013/6/6 Ralph Castain > Afraid not. You could start a single process, and then have that process call > MPI_Comm_spawn to launch the rest of them > > Mmmm... sounds good > >

Re: [OMPI users] Set number of processes in another way than -np X

2013/6/6 Ralph Castain > Afraid not. You could start a single process, and then have that process > call MPI_Comm_spawn to launch the rest of them > Mmmm... sounds good I'm writing an example program using ScaLAPACK. I have written the ScaLAPACK code in an independent

Re: [OMPI users] Set number of processes in another way than -np X

Afraid not. You could start a single process, and then have that process call MPI_Comm_spawn to launch the rest of them On Jun 6, 2013, at 7:54 AM, José Luis García Pallero wrote: > Hello: > > I'm newbie in the use of MPI, so probably I ask some stupid question (or >

[OMPI users] Set number of processes in another way than -np X

Hello: I'm newbie in the use of MPI, so probably I ask some stupid question (or previously asked, but in this case I have searched in the archive and I haven't found anything): Exists any other way than -np X in order to pass the number of processes to start for an MPI program? I mean a function

Re: [OMPI users] Oversubcription for mimd/mpmd mpiruns

You could do it by specifying which cores to use - something like mpirun -np 4 --slot-list 0-3 prog_1 : -np 1 prog_2 On Jun 6, 2013, at 1:52 AM, Nima Aghajari wrote: > Dear all, > I am currently using openmpi 1.6.4 and trying to do a parallel performance > analysis for a

Re: [OMPI users] openmpi-lsb_launch failed

Wow - that is ancient! Can you update to something more recent - perhaps something like 1.6.4? I have no idea what the problem might be in something that old, but it certainly was working with LSF back then. Can you run a simple app, perhaps something like "mpirun -n 1 hostname"? On Jun 5,

[OMPI users] Oversubcription for mimd/mpmd mpiruns

Dear all, I am currently using openmpi 1.6.4 and trying to do a parallel performance analysis for a parallel two-program mpirun. So what I have are two programs that are executed like this:        mpirun -np 4 my_prog1 : -np 1 my_prog2   my_prog1 and my_prog2 run sequentially, so when one

Re: [OMPI users] openmpi-lsb_launch failed

Hi , We are using openmpi version 1.4.5. Thanks and Regards, Mahalakshmi Murthy [Maha] TCS | GE Global Research Mail ID: mahalakshmi.mur...@ge.com Mobile No:8147923917 From: users-boun...@open-mpi.org [mailto:users-boun...@open-mpi.org] On Behalf Of Ralph Castain Sent: