[OMPI users] OPENMPI only recognize 4 cores of AWS EC2 machine

Hi, OPENMPI users, I have met some problem while doing experiments on Amazon EC2. I have chosen instance type r3.2xlarge. As amazon claimed, it has 8 vcpu. Each vCPU is a hyperthread of an Intel Xeon core for M3, C4, C3, R3, HS1, G2, I2, and D2. (from *https://aws.amazon.com/ec2/instance-types/

Re: [OMPI users] OPENMPI only recognize 4 cores of AWS EC2 machine

Hi Xing, iirc, open MPI default behavior is to bind to cores (vs hyperthreads), hence the error message. I cannot remember the option to bind to threads, but you can mpirun --oversubscribe if you are currently stuck Cheers, Gilles On Sunday, May 24, 2015, XingFENG wrote: > Hi, OPENMPI users,

Re: [OMPI users] OPENMPI only recognize 4 cores of AWS EC2 machine

Hi, Gilles, Thanks very much for your respond. But if I oversubscribe, would the performance be influenced? On Sun, May 24, 2015 at 7:24 PM, Gilles Gouaillardet < gilles.gouaillar...@gmail.com> wrote: > Hi Xing, > > iirc, open MPI default behavior is to bind to cores (vs hyperthreads), > hence t

Re: [OMPI users] OPENMPI only recognize 4 cores of AWS EC2 machine

yes, that is clearly suboptimal. the optimal option is something like -bind-to hwthreads (I am afk so I cannot confirm this) Cheers, Gilles On Sunday, May 24, 2015, XingFENG wrote: > Hi, Gilles, > > Thanks very much for your respond. But if I oversubscribe, would the > performance be influenc

Re: [OMPI users] OPENMPI only recognize 4 cores of AWS EC2 machine

Thanks, Gilles On Sun, May 24, 2015 at 7:41 PM, Gilles Gouaillardet < gilles.gouaillar...@gmail.com> wrote: > yes, that is clearly suboptimal. > > the optimal option is something like -bind-to hwthreads > (I am afk so I cannot confirm this) > > Cheers, > > Gilles > > On Sunday, May 24, 2015, Xing

[OMPI users] Fwd: Default value of btl_openib_memalign_threshold

Dear OpenMPI developers / users, This is much more a comment than a question since I believe I have already solved my issue. But I would like to report it. I have noticed my code performed very badly with OpenMPI when Infinand is enabled, sometime +50% or even +100% overhead. I also have this slo

Re: [OMPI users] Default value of btl_openib_memalign_threshold

Looking at the code, we do in fact set the memalign_threshold = eager_limit by default, but only if you configured with —enable-btl-openib-malloc-alignment AND/OR we found the malloc hook functions were available. You might check config.log to see if the openib malloc hooks were enabled. My gue