Conclusion

The patterns illustrated by these Jupyter in HPC use cases are very flexible and can be applied to other science domains as well. We see common themes around parallel execution, reproducible environments and interactive visualization come up repeatedly in our work enabling Jupyter for science on HPC systems.