I’m trying to run with the GPU package but vnc_get_devices is failing (lammps then also fails). Can anyone advise? Details are as follows:
$ mpirun -n 1 ./nvc_get_devices
Cuda driver error 101 in call at file ‘./geryon/nvd_device.h’ in line 273.
*** The MPI_Abort() function was called before MPI_INIT was invoked.
*** This is disallowed by the MPI standard.
*** Your MPI job will now abort.
[mycomp:17159] Abort before MPI_INIT completed successfully; not able to guarantee that all other processes were killed!
I also tried without mpirun, with the same result.
I'm trying to run with the GPU package but vnc_get_devices is failing
(lammps then also fails). Can anyone advise? Details are as follows:
$ mpirun -n 1 ./nvc_get_devices
Cuda driver error 101 in call at file './geryon/nvd_device.h' in line 273.
this means that it is not even possible to open the GPU device. this
hints at a problem with your GPU setup.
can you use nvidia-smi? and query the GPU status?
*** The MPI_Abort() function was called before MPI_INIT was invoked.
*** This is disallowed by the MPI standard.
*** Your MPI job will now abort.
[mycomp:17159] Abort before MPI_INIT completed successfully; not able to
guarantee that all other processes were killed!
I also tried without mpirun, with the same result.
This occurs with lammps versions 1Feb14, 5Sep14 and 26Nov14.
GPU is a Tesla C2075.
Compilation was with double-precision (Makefile.linux.double).
"make -f Makefile.linux clean" was run beforehand.
nvcc --version gives
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2012 NVIDIA Corporation
Built on Thu_Apr__5_00:24:31_PDT_2012
Cuda compilation tools, release 4.2, V0.2.1221
cuda 4.2 is *very* old. have you considered upgrading?