Amber18安装

最新推荐文章于 2025-03-22 17:13:56 发布

Jsylar

最新推荐文章于 2025-03-22 17:13:56 发布

阅读量2.4k

点赞数 3

分类专栏： Amber

本文链接：https://blog.youkuaiyun.com/qq_33953882/article/details/113978887

版权

Amber 专栏收录该内容

12 篇文章

订阅专栏

一、依赖安装

1.gcc-4.8.5安装

1.下载
链接：https://pan.baidu.com/s/12yOB-nB3v86KsKjXDz2V9Q
提取码：x6gy
2.解压

tar -jxvf gcc-4.8.5.tar.bz2

3.进入解压后的目录

cd gcc-4.8.5

4.下载和安装依赖库

./contrib/download_prerequisites

7.编译安装

./configure --prefix=/path/gcc -enable-checking=release -enable-languages=c,c++,fortran -disable-multilib

make

make install

8.添加环境变量

2.anaconda3安装

1.下载
链接：https://pan.baidu.com/s/12yOB-nB3v86KsKjXDz2V9Q
提取码：x6gy
2.安装

bash ./Anaconda3-2020.11-Linux-x86_64.sh

3.openmpi-1.6.5安装

1.下载
链接：https://pan.baidu.com/s/12yOB-nB3v86KsKjXDz2V9Q
提取码：x6gy
2.解压

tar jxvf openmpi-1.6.5.tar.bz2

3.进入目录

cd openmpi-1.6.5

4.编译和安装

./configure --prefix=/path/openmpi

5.环境变量

#openmpi
export PATH=/path/openmpi/bin:$PATH
export LD_LIBRARY_PATH=/path/openmpi/lib:$LD_LIBRARY_PATH
export LIBRARY_PATH=/path/openmpi/lib:$LIBRARY_PATH
export LD_LIBRARY_PATH=/path/openmpi/bin:$LD_LIBRARY_PATH
export LIBRARY_PATH=/path/openmpi/bin:$LIBRARY_PATH

4.zlib-1.2.11安装

1.下载
链接：https://pan.baidu.com/s/12yOB-nB3v86KsKjXDz2V9Q
提取码：x6gy
2.解压

tar zxvf zlib-1.2.11.tar.gz

3.进入目录

cd zlib-1.2.11

4.编译和安装

./configure --prefix=/path/zlib

make

make check

make install

5.环境变量

#zlib
export PATH=/path/zlib/lib:$PATH
export LD_LIBRARY_PATH=/path/zlib/lib:$LD_LIBRARY_PATH
export LIBRARY_PATH=/path/zlib/lib:$LIBRARY_PATH
export PATH=/path/zlib/include:$PATH
export LD_LIBRARY_PATH=/path/zlib/include:$LD_LIBRARY_PATH
export LIBRARY_PATH=/path/zlib/include:$LIBRARY_PATH

5.szip-2.1.1安装

1.下载
链接：https://pan.baidu.com/s/12yOB-nB3v86KsKjXDz2V9Q
提取码：x6gy
2.解压

 tar zxvf szip-2.1.1.tar.gz

3.进入目录

cd szip-2.1.1

4.编译安装

./configure --prefix=/path/szip

make

make check

make install

5.环境变量

#szip
export PATH=/path/szip/lib:$PATH
export LD_LIBRARY_PATH=/path/szip/lib:$LD_LIBRARY_PATH
export LIBRARY_PATH=/path/szip/lib:$LIBRARY_PATH
export PATH=/path/szip/include:$PATH
export LD_LIBRARY_PATH=/path/szip/include:$LD_LIBRARY_PATH
export LIBRARY_PATH=/path/szip/include:$LIBRARY_PATH

6.hdf5-1.12.0安装

1.下载
链接：https://pan.baidu.com/s/12yOB-nB3v86KsKjXDz2V9Q
提取码：x6gy
2.解压

 tar zxvf hdf5-1.12.0.tar.gz

3.进入目录

cd hdf5-1.12.0

4.编译安装

./configure --with-zlib=/path/zlib  --with-szlib=/path/szip --prefix=/path/hdf5 --enable-fortran --enable-parallel  --host=x86_64

make

make check

make install

5.环境变量

#hdf5
export PATH=/path/hdf5/bin:$PATH
export LD_LIBRARY_PATH=/path/hdf5/bin:$LD_LIBRARY_PATH
export PATH=/path/hdf5/lib:$PATH
export LD_LIBRARY_PATH=/path/hdf5/lib:$LD_LIBRARY_PATH
export LIBRARY_PATH=/path/hdf5/lib:$LIBRARY_PATH
export PATH=/path/hdf5/include:$PATH
export LD_LIBRARY_PATH=/path/hdf5/include:$LD_LIBRARY_PATH
export LIBRARY_PATH=/path/hdf5/include:$LIBRARY_PATH

7.pnetcdf-1.12.1安装

1.下载
链接：https://pan.baidu.com/s/12yOB-nB3v86KsKjXDz2V9Q
提取码：x6gy
2.解压

 tar zxvf pnetcdf-1.12.1.tar.gz

3.进入目录

cd pnetcdf-1.12.1

4.编译安装

 ./configure --prefix=/path/pnetcdf --with-mpi=/path/openmpi --enable-subfiling --enable-shared --enable-large-file-test --enable-null-byte-header-padding --enable-burst-buffering --enable-profiling --host=x86_64

make

make check

make install

5.环境变量

#pnetcdf
export PATH=/path/pnetcdf/lib:$PATH
export LD_LIBRARY_PATH=/path/pnetcdf/lib:$LD_LIBRARY_PATH
export LIBRARY_PATH=/path/pnetcdf/lib:$LIBRARY_PATH
export PATH=/path/pnetcdf/include:$PATH
export LD_LIBRARY_PATH=/path/pnetcdf/include:$LD_LIBRARY_PATH
export LIBRARY_PATH=/path/pnetcdf/include:$LIBRARY_PATH

二、Amber安装

1.下载

2.解压

 tar xvf Amber18.tar
 tar xvf AmberTools19.tar

3.进入目录

cd amber18

4.添加环境变量

#Amber
test -f /path/amber18.sh  && source /path/amber18.sh

5.编译安装

串行版本

./configure --with-python /path/python gnu

make install

并行版本

./configure --with-python /path/bin/python --with-pnetcdf /path/pnetcdf/ -mpi gnu

make install

./configure --with-python /path/bin/python --with-pnetcdf /path/pnetcdf/ -openmp gnu

make openmp

GPU串行版本

./configure --with-python /path/bin/python -cuda gnu

make install

GPU并行版本

./configure --with-python /path/bin/python --with-pnetcdf /path/pnetcdf/ -cuda -mpi gnu

make install

环境变量

#Amber
export AMBERHOME=$soft/amber18
export PATH=$AMBERHOME/bin:$PATH
export PATH=$AMBERHOME/bat:$PATH
export LD_LIBRARY_PATH=$AMBERHOME/lib:$LD_LIBRARY_PATH
export MPI_HOME=$AMBERHOME/AmberTools
export LD_LIBRARY_PATH=$MPI_HOME/lib:$LD_LIBRARY_PATH
export DO_PARALLEL="mpirun -np 4"
export PYTHONPATH="${AMBERHOME}/lib/python3.8/site-packages:${PYTHONPATH}"
export CUDA_HOME=/usr/local/cuda
export PATH=$CUDA_HOME/bin:$PATH
export LD_LIBRARY_PATH="/usr/local/cuda/lib:${LD_LIBRARY_PATH}"
export LD_LIBRARY_PATH=$CUDA_HOME/lib64:$LD_LIBRARY_PATH