This directory contains an example cuda program named "helloCuda.cu"
Program Compilation:
nvcc -o helloCuda helloCuda.cu -arch sm_60
Program Execution:
CUDA programs must be executed on a GPU-enabled machine.
To request a node with GPUs, execute the following commands at the SCC prompt: qrsh -V -l gpus=1 -l gpu_c=3.5
module load cuda/10.0
Users who are members of any Medical Campus projects must also include their project group name: qrsh -P groupname -V -l gpus=1 -l gpu_c=3.5
module load cuda/10.0
When the terminal window appears, to execute the program, type: ./helloCuda
Contact Information
Katia Oleinik: help@scc.bu.edu
Operating System Requirements
The examples presented in this directory were written for Linux OS, but should
work on any OS where
- c and nvcc compilers available
- NVIDIA's GPU(s) with compute capability 2.0 and higher