---

HSF Weekly Meeting #157, 31 January, 2019

Present/Contributors: Graeme Stewart, Danilo Piparo, Jim Amundson, Agnieszka Dziurda, Jim Pivarski, David Lange, Witek Pokorski, Serhan Mete, Pere Mato, Liz Sexton-Kennedy, Stefan Roiser, Caterina Doglioni, Michel Jouvin, Guilherme Amadio, Daniel Elvira, Gloria Corti, Davide Costanzo

Apologies: Heather Gray

Google Summer of Code 2019

• Lots of PRs coming in now.
• Remember the HSF deadline: February 4.
  • But better if you can submit this week.
• This presentation has all the details.

HSF/WLCG/OSG Workshop News and Planning

• Indico - registration is open.
  • Early registration ends on 1 February; 1 week extension probable - please register right now.
• There is a Google Sheet linked from the main page to help coordinate car sharing from Dulles (you take your own responsibility for this!).
  • If you are arriving/travelling from another airport you can also use that spreadsheet (just make it clear which airport you will be at).

Session Planning (see block timetable)

• Reminder: David Lange, Michel Jouvin and Graeme Stewart organising for HSF.
• General Sessions:
  • Opening plenary session from HEP experiments and other communities (Monday)
    • Confirmed talks from LHC Experiments, DUNE, Belle II, Dark Matter, Electron-Ion Consortium, LIGO, IceCube, Lightsources; yet to confirm LSST (speaker invited).
    • Plus discussion on WLCG future evolution.
  • Technology watch and networking (Tuesday AM).
  • New Workflows, New Facilities (Tuesday AM).
    • Benedikt Hegner and Maria Girone are convening.
  • Closing Session, visionary talk and discussion on the future (Friday).
    • Funding updates:
      • DE (Thomas Kuhr)
      • UK, IRIS (Pete Clarke)
      • US, IRIS-HEP (Pete Elmer)
      • If you are aware of any other projects that have secured funding for software efforts can you let us know?
    • Amber Boehnlein will do the final talk.
• HSF Sessions:
  • Please fill in a first version of your detailed timetable in the coming week, even if not all slots are confirmed
    • All session convenors have management rights, let us know if you have any problems.
    • And do remember it’s a workshop - schedule generous discussion time!
  • Software on Accelerators (2 slots).
    • Graeme, Michel, David L are convening.
      • Confirmed talks on Accelerator Technologies, Framework Integration, Patatrack, ALICE GPU Algorithms, MadGraph on GPUs
      • One last speaker to confirm, will leave plenty of time for discussion.
  • For other HSF Parallel Sessions, see WG activity reports below.
  • Due to the logistics of rooms at JLab we propose that:
    • We only have one session on Wednesday 16h-17h30
    • We move the other two WG’s second sessions to Thursday 9h-10h30
    • We move Software Tools to Thursday 16h-17h30
      • N.B. Still likely that small rooms would be available for any BoFs

Activity and Working Group Updates

General points:

• Reminder to make sure your WG page has
  • Brief description of the interests of the group
  • Information on how to join the group’s mailing list
  • Information on how to contact the convenors (convenor list or individual emails)
  • A link to the group’s Indico category

Data Analysis

• Second meeting (afternoon of Feb 13th) of the WG in preparation: https://indico.cern.ch/event/789007/
• Refined focus of the meeting to survey innovation and best practice from the community
• Meeting will be workshop in style, shorter talks, focus on the ideas, “what is the ground-breaking, earth-shattering idea(s) / functionality that make your framework/tool so useful?”

Detector Simulation

• First topical meeting yesterday on FastSim.
  • Nice overview from LHC experiments.
  • In future address non-LHC experiments as well.
  • Daniel can help with contact to IF (division between fast/full sim for IF less clear, but simulation very relevant for them too).
• Next meeting on ML, 6 March.
• Invitations being sent for the workshop, some speakers might have to be remote.

Reconstruction and Software Triggers

• Indico updated (15+5) min with 10 min discussion.
• Talks accepted:
  • Machine Learning for the Primary Vertex reconstruction.
  • Using FPGAs to accelerate machine learning inference.
  • A comprehensive real-time analysis model at the LHCb experiment.
  • [still waiting for speaker, but Speakers Committees have said yes]: real-time analysis in CMS/ATLAS/ALICE.
• We are waiting still to get feedback from:
  • DUNE/Belle II/LSST (machine learning).
• We planned to have a first meeting next week, but due to people’s unavailability we decided to postpone - more news next week.

Software Tools

• Invitations for JLab session to be sent tomorrow.

PyHEP

• Plan to divide 1.5 hours (14:00‒15:30 on Thursday) into
  • two 15+5 minute long talks
    • considering LPC Coffea group’s Numpy-based CMS analyses
    • invited Stan Seibert to present Numba (Python JIT-compiler)
  • three 5+5 minute lightning talks
    • confirmed talk on zfit by Albert Puig, Jonas Eschle, Rafael Silva Coutinho
    • invited ROOT Team to present ROOT packaging in Conda
    • looking for a final talk on Python in neutrinos, dark matter, or BELLE…
  • 20 minutes of free-form discussion

Licensing

• Still chasing up CERN KT and FastJet authors.

Packaging

• Meeting yesterday. Group plans to track work on various projects. Main topic was an update on Conda - very interesting to see how this tool has hugely progressed since the first meetings of the WG. New ROOT recipe seems to be very nicely integrated now.
• Next meeting in 1 month, 27 February.

AOB

• URL checks on website are unreliable; another URL check code could be tried to see if that improves things.
• Working on an update to the instructions for converting Google Doc -> Word -> Markdown.

---