Installing LAMMPS with GPU acceleration

LAMMPS LAMMPS Installation
1

Dear experts, I installed LAMMPS with a GPU accelerator on my machine. I used Opencl API and the installation was successful. But when I run my simulation an error occurs and is pasted below.

OpenCL error in file '/home/user/softwares/lammps-23Jun2022/lib/gpu/geryon/ocl_kernel.h' in line 467 : -4.

I am pasting the output for ocl_get_devices and ./lmp -h below:

Found 1 platform(s).

Platform 0:

Device 0: "NVIDIA GeForce GT 730"
  Type of device:                                GPU
  Supported OpenCL Version:                      3.0
  Is a subdevice:                                No
  Double precision support:                      Yes
  Total amount of global memory:                 1.95227 GB
  Number of compute units/multiprocessors:       2
  Total amount of constant memory:               65536 bytes
  Total amount of local/shared memory per block: 49152 bytes
  Maximum group size (# of threads per block)    1024
  Maximum item sizes (# threads for each dim)    1024 x 1024 x 64
  Clock rate:                                    0.705 GHz
  ECC support:                                   No
  Device fission into equal partitions:          No
  Device fission by counts:                      No
  Device fission by affinity:                    No
  Maximum subdevices from fission:               1
  Shared memory system:                          No
  Subgroup support:                              No
  Shuffle support:                               Yes

lmp -h ::

OS: Linux "Ubuntu 20.04.4 LTS" 5.15.0-56-generic x86_64

Compiler: GNU C++ 9.4.0 with OpenMP 4.5
C++ standard: C++11
MPI v3.1: MPICH Version:	3.3.2
MPICH Release date:	Tue Nov 12 21:23:16 CST 2019
MPICH ABI:	13:8:1

Accelerator configuration:

GPU package API: OpenCL
GPU package precision: mixed
OPENMP package API: OpenMP
OPENMP package precision: double

Compatible GPU present: yes
2

You should try with the 28 March 2023 version of LAMMPS that we just released, which has multiple bugfixes and performance improvements versus the GPU code in the stable version.

But don’t get your hopes up too high. You have a rather old and entry level GPU there, the chance to see a significant speedup is very low.

3

Ok, thank you dr. Kohlmeyer, I will try the newest version.

4

Dear Dr. Kohlmeyer, the issue is not solved and the error is persistent. I am pasting the error below:

 UCL Error: Error compiling OpenCL Program (-2) ...
----------------------------------------------------------
<kernel>:902:5: error: expected identifier or '('
    for ( ; nbor<nbor_end; nbor+=n_stride) {
    ^

The outputs for ocl and lmp are as following:

Found 1 platform(s).

Platform 0:

Device 0: "NVIDIA GeForce GT 730"
  Type of device:                                GPU
  Supported OpenCL Version:                      3.0
  Is a subdevice:                                No
  Double precision support:                      Yes
  Total amount of global memory:                 1.95227 GB
  Number of compute units/multiprocessors:       2
  Total amount of constant memory:               65536 bytes
  Total amount of local/shared memory per block: 49152 bytes
  Maximum group size (# of threads per block)    1024
  Maximum item sizes (# threads for each dim)    1024 x 1024 x 64
  Clock rate:                                    0.705 GHz
  ECC support:                                   No
  Device fission into equal partitions:          No
  Device fission by counts:                      No
  Device fission by affinity:                    No
  Maximum subdevices from fission:               1
  Shared memory system:                          No
  Subgroup support:                              No
  Shuffle support:                               Yes


OS: Linux "Ubuntu 20.04.4 LTS" 5.15.0-56-generic x86_64

Compiler: GNU C++ 9.4.0 with OpenMP 4.5
C++ standard: C++11
MPI v3.1: MPICH Version:	3.3.2
MPICH Release date:	Tue Nov 12 21:23:16 CST 2019
MPICH ABI:	13:8:1

Accelerator configuration:

GPU package API: OpenCL
GPU package precision: mixed
OPENMP package API: OpenMP
OPENMP package precision: double

Compatible GPU present: yes
5

Please provide a (simple, short) input deck to reproduce this.

Also, make certain that you have up-to-date GPU drivers and corresponding OpenCL support libraries.