Re: cuda error

[Jean-Noël Grad]

Hi Le,

I've never seen that error message before. Do you get a backtrace? Is 
the error reproducible when running the script without mpirun/mpiexec?

Best,
JN

On 10/31/19 3:37 PM, Le Qiao wrote:
> Hi all,
>
> I'm having a Cuda error when running LB simulations on the ubuntu 18.04 
> LTS with Espresso dev, Cuda 10.1 and OpenMPI 2.1.1. This error will be 
> gone every time when I reboot the computer but will still appear 
> randomly later. The script is working fine on another machine. I'm not 
> sure if it's a hardware or software issue now, Does anyone has any 
> suggestions on this issue?
>
>
> CUDA error: unknown error
> --------------------------------------------------------------------------
> MPI_ABORT was invoked on rank 0 in communicator MPI_COMMUNICATOR 4
> with errorcode 1.
>
> NOTE: invoking MPI_ABORT causes Open MPI to kill all MPI processes.
> You may or may not see output from other processes, depending on
> exactly when Open MPI kills them.
>
> Cheers
> Le