Building CUDA enabled Caffe on HGX system notos.sdsu.edu

How to build the Caffe deep learning framework from Berkeley AI Research (BAIR)/The Berkeley Vision and Learning Center (BVLC) on notos.sdsu.edu.

Clone Caffe repository:

$ git clone https://github.com/BVLC/caffe.git

$ cd caffe

$ mkdir build

$ cd build

cmake -DUSE_CUDNN=0 \
-DProtobuf_PROTOC_EXECUTABLE=/usr/bin/protoc \
-DProtobuf_LIBRARY_DEBUG=/usr/lib/x86_64-linux-gnu/libprotobuf.so \
-DProtobuf_LIBRARY_RELEASE=/usr/lib/x86_64-linux-gnu/libprotobuf.so \
-DProtobuf_LITE_LIBRARY_DEBUG=/usr/lib/x86_64-linux-gnu/libprotobuf-lite.so \
-DProtobuf_LITE_LIBRARY_RELEASE=/usr/lib/x86_64-linux-gnu/libprotobuf-lite.so \
-DProtobuf_PROTOC_LIBRARY_DEBUG=/usr/lib/x86_64-linux-gnu/libprotoc.so.17 \
-DProtobuf_PROTOC_LIBRARY_RELEASE=/usr/lib/x86_64-linux-gnu/libprotoc.so.17 \
-DPYTHON_EXECUTABLE:FILEPATH=/usr/bin/python3.8 \
-DPYTHON_INCLUDE_DIR:PATH=/usr/include/python3.8 \
-DPYTHON_LIBRARY:FILEPATH=/usr/lib/x86_64-linux-gnu/libpython3.8.so \
..

$ make VERBOSE=1 -j42 all

$ export PYTHONPATH=/mnt/beegfs/home/username/caffe/python

$ python3.8
Python 3.8.10 (default, Jun 22 2022, 20:18:18)
[GCC 9.4.0] on linux
Type “help”, “copyright”, “credits” or “license” for more information.

import caffe
>>> exit()