GPU package can't proceed properly

Hi all,

I am using gpu package to minimize and run nve dynamics of a system of 67000 atoms. But after successfully minimization, the nve dynamics can not proceed properly. Even after I had deleted the nve dynamics, it stuck before been able to write a restart file as :

Reading data file …
orthogonal box = (0 0 0) to (57.328 72.4913 190.329)
1 by 2 by 4 MPI processor grid
67200 atoms
WARNING: Resetting reneighboring criteria during minimization (min.cpp:167)

Hi all,

I am using gpu package to minimize and run nve dynamics of a system of 67000
atoms. But after successfully minimization, the nve dynamics can not proceed
properly. Even after I had deleted the nve dynamics, it stuck before been
able to write a restart file as :

Reading data file ...
orthogonal box = (0 0 0) to (57.328 72.4913 190.329)
1 by 2 by 4 MPI processor grid
67200 atoms
WARNING: Resetting reneighboring criteria during minimization (min.cpp:167)

--------------------------------------------------------------------------
- Using GPGPU acceleration for eam:
- with 8 proc(s) per device.
--------------------------------------------------------------------------
GPU 0: GeForce GTX 560, 224 cores, 0.58/1 GB, 1.7 GHZ (Double Precision)
--------------------------------------------------------------------------

Initializing GPU and compiling on process 0...Done.
Initializing GPU 0 on core 0...Done.
Initializing GPU 0 on core 1...Done.
Initializing GPU 0 on core 2...Done.
Initializing GPU 0 on core 3...Done.
Initializing GPU 0 on core 4...Done.
Initializing GPU 0 on core 5...Done.
Initializing GPU 0 on core 6...Done.
Initializing GPU 0 on core 7...Done.

apart from your obvious problem with a (newly added)
accelerated pair styles, i would like to remark that with
compiling for double precision and 8-fold oversubscription,
you are not likely to see much acceleration from your
GPU. ... or from your CPU, as it looks like you have
a hyperthreading enabled CPU.

axel.

Hi all,

I am using gpu package to minimize and run nve dynamics of a system of 67000
atoms. But after successfully minimization, the nve dynamics can not proceed
properly. Even after I had deleted the nve dynamics, it stuck before been
able to write a restart file as :

quick question:

does your input run on the CPU?

axel.

And what does "stuck" mean?

Steve

Hi Jing,

I couldn’t reproduce the error you observed with (minimization + eam/alloy/gpu + write_restart) on my local machine.

Note that since write_restart command evokes initialization of the gpu library, you would see a series of "Initializing GPU # on core #… " like in the log file you showed. The simulation in fact did not stuck, it just finishes normally.

Cheers,
-Trung

Hi Trung,

Thanks you for clarifying and testing. Did you run a nve dynamics after minimization? I can never produce restart file if I use any two of minimization or fix nve before write_restart. Was the restart file generated normally on your computer?

Appreciated your help!

Best,
Jing