1 of 3

Organization of HEP activities

Mia Liu, Mark Neubauer, Phil Harris, Javier Duarte

2 of 3

HEP activities

  • Engagement of postdocs/students from each member institute
  • HEP PIs are already active members of the fast machine learning community
    • Our postdocs/students regularly present on project progress/status in fast ML general/hls4ml/co-processor meetings.
      • On average 2-3 reports for the duration of a project.
      • A good venue to receive feedbacks and motivate further studies
      • → can probably benefit from attendance of experts on algorithm design/hardware in the future?
    • Hosted tutorials on hls4ml and sonic (many ran by our postdocs), in conjunction with fast ml workshops/experiment internal training sessions/events etc -> benefiting HEP/outside-HEP postdocs & students
    • Deployment/integration will offer more opportunities to engage postdocs/students inside/outside the experiments.
      • Also opportunities to engage other stakeholders such as the HEP HW trigger/computing communities.

3 of 3

HEP activities

  • Regular events
    • Since we already present and communicate weekly in the fast ML meetings, we could consider monthly meetings to discuss a3d3 specific items.
    • Is it worth thinking about working group style meetings for more technical discussions?
      • Probably some overlap with HAC for algo/tool development and experiment internal discussions for application focused progress.
  • Deliverables:
    • While we are not required to give a set of deliverables
    • It might be good to have a set of internal deliverables for this project
      • This will help to organize our plans going forward
  • Pre-kick off meeting activity toward development of detailed management plan→ End of October?