Dear lammps users and developers,
I compiled lammps on our GPU machine( TESLA 2050) and I tried to run a test case ( LJ fluid)
using the following input script…
Dear lammps users and developers,
I compiled lammps on our GPU machine( TESLA 2050) and I tried to run a test case ( LJ fluid)
using the following input script…
Maybe Paul or Mike has an idea.
Steve
Bugs have been introduced into the lj/gpu potential in the main LAMMPS distribution; this needs to be updated. For GPU acceleration, please use the code at:
http://code.google.com/p/gpulammps/source/checkout
and read the new README in lib/gpu and documentation in doc/Section_start.html under running on GPUs as this has changed. If you still have problems, I can address them for this code.
Thanks. - Mike
Maybe Paul or Mike has an idea.
Steve
Dear all,
I compiled the latest version(lammps-12Oct10) and still the problem persits.
P.S: I built gpu library with the following option
CUDA_PREC = -D_DOUBLE_DOUBLE
with regards
srinivas
Did you download the code with the svn checkout instructions from the site I provided? Is the root directory name for the distribution gpulammps? The code is not available from the main LAMMPS site.
- Mike
Srinivas,
From your message, it looks like you didn’t try using the code at:
http://code.google.com/p/gpulammps/source/checkout
as Mike suggested. Would you mind trying this and letting us know?
Thanks,
Paul