Compilation and Installation

Procedure

Use PuTTY to log in to the server as the root user.
Go to the main program installation directory.
```
cd /path/to/LAMMPS
```
Decompress the installation package.
```
tar -xvf lammps-10Feb2021.tar.gz
```
Go to the directory generated after the decompression.
```
cd lammps-10Feb21
```

Modify the Makefile.linux configuration file of the GPU module.

Open the Makefile.linux file.
```
cd lib/gpu
vi Makefile.linux
```

Press i to enter the insert mode and modify the file as follows:

CUDA_HOME = /usr/local/cuda-11.4
NVCC=nvcc
CUDA_ARCH=-arch=sm_80  # Graphics card model. The current environment uses A100 (Ampere architecture).
CUDA_PRECISION=-D_DOUBLE_DOUBLE  # calculation precision.

Press Esc, type :wq!, and press Enter to save the settings and exit.

Compile the GPU module of LAMMPS.
```
make -f Makefile.linux -j 16
```

View the GPU information.

./nvc_get_devices

If the following information is displayed, the compilation is successful:

Found 1 platform(s).
Using platform: NVIDIA Corporation NVIDIA CUDA Driver
CUDA Driver Version:                           11.40
Device 0: "NVIDIA A100-PCIE-40GB"
Type of device:                                GPU
Compute capability:                             8
Double precision support:                        Yes
Total amount of global memory:                   39.5861 GB
Number of compute units/multiprocessors:           108
Number of cores:                               20736
Total amount of constant memory:                 65536 bytes
Total amount of local/shared memory per block:      49152 bytes
Total number of registers available per block: 65536
Warp size:                                     32
Maximum number of threads per block:             1024
Maximum group size (# of threads per block)         1024 x 1024 x 64
Maximum item sizes (# threads for each dim)         2147483647 x 65535 x 65535
Maximum memory pitch:                         2147483647 bytes
Texture alignment:                              512 bytes
Clock rate:                                    1.41 GHz
Run time limit on kernels:                        No
Integrated:                                    No
Support host page-locked memory mapping:         Yes
Compute mode:                                Default
Concurrent kernel execution:                     Yes
Device has ECC support enabled:                 Yes

Go to the src directory.
```
cd ../../src
```
Modify the MAKE/OPTIONS/Makefile.g++_openmpi file.
1. Open the MAKE/OPTIONS/Makefile.g++_openmpi file.
```
vi MAKE/OPTIONS/Makefile.g++_openmpi
```
2. Press i to enter the edit mode and modify the content in lines 54, 55, and 56.
```
FFT_INC =    -DFFT_FFTW -I/path/to/FFTW/include
FFT_PATH =  -L/path/to/FFTW/lib
FFT_LIB =    -lfftw3
```
  After performing 9.a, press Esc, type :set nu, and press Enter to display the line number.
3. Press Esc, type :wq!, and press Enter to save the settings and exit.

Perform the compilation.

make yes-std
make no-lib
make yes-gpu
make -j 96 g++_openmpi

Parent topic: LAMMPS 10 Feb 2021 Porting Guide (CentOS 8.2)