Follow up on discussions on computational needs for conventional accelerators

[Jean-Luc Vay]

Thanks to everyone who participated in today’s discussions and to the presenters.

Based on today’s discussions and on Ji Qiang’s feedback on the guidance from the Computational Frontier, one could use the 2013 subgroup report as guidance (https://www.slac.stanford.edu/econf/C1307292/docs/ComputingFrontier/CforAccel-43.pdf) for one joint white paper on accelerator modeling needs, shooting for ~10 pages limit for the full paper. 

Within this constrain, we propose to adjust the target size of each subsection to be roughly proportional to the number of LOIs that it covers, with an average of about a page per subtopic (as a reminder, the list of subtopics and corresponding LOIs can be found here: https://snowmass-compf2-accbeammodel.github.io/loi/submitted.html).

One potential difficulty could arise from some co-authors not agreeing to support all or part of one subsection. We will thus need to find a way to indicate that. We can cross that bridge when we get to it.

Concerning the writing of a common contributed paper on machine learning for accelerators, as was discussed last week, it would then become a one page summary in the new paper.

Let us know by responding to this email if there is any question, comment or concern.

The tentative plan for October 28 is to discuss the LOIs on Advanced Accelerator Concepts.

Best,

Jean-Luc, Ji and Axel