What is HPC and why should it care about Jupyter, what is NERSC and why does it care about Jupyter.  Why we are not cloud.  Overview our strategy of JupyterHub, JupyterLab, and engagement.  Should have a partial list of HPC efforts we know about and reference them.
In this article we can talk about our Jupyter deployment as one example of many but point out where things are the same across HPC deployments, and where things look like they are headed in the same direction as cloud deployments, it may be good to point out divergences and suggest what we should be doing about them.

JupyterHub

Why are we doing this and not say Open OnDemand

JupyterLab

What we are doing

Engagement

On the HPC side.  On the Jupyter side.

Conclusions

What we talked about.  Also we should talk about future plans and trends we are expecting to see over the next 5 years say.

Acknowledgments

This work was supported by Lawrence Berkeley National Laboratory, through the U.S. Department of Energy under Contract No. DE-AC02-05CH11231. This work used resources of the National Energy Research Scientific Computing Center (NERSC), a U.S. Department of Energy Office of Science User Facility located at Lawrence Berkeley National Laboratory.