HSF Weekly Meeting #155, 17 January, 2019
Present/Contributors: Graeme Stewart, Dario Menasce, Sudhir Malik,
Andrei Gheata, Danilo Piparo, Giulio Eulisse, Tommaso Boccali, Serhan
Mate, Witek Pokorski, Stefan Roiser, Concezio Bozzi, Andrea Valassi,
Javier Cervantes Villanueva, Marco Clemencic, Torre Wenaus, Eduardo
Rodrigues, Agnieszka Dziurda, Antoine Perus, Caterina Doglioni, Chris
Tunnell, Davide Costanzo, Liz Sexton-Kennedy, Heather Gray, Josh
McFadyen, Martin Ritter, Michael Kirby, Paul Laycock, Daniel Elvira,
David Lange
Google Summer of Code 2019
- See slides attached to the
agenda.
- Do students come to your institute?
- No, they are all remote. Has a very big pool of candidates for
this reason.
- Payment is linked to continued success at checkpoints.
- Level of students?
- Proposals are filtered by organisation first, if more proposals
than awarded slots.
- Having a good group of mentors (>2) really helps.
- Activity now is important to have a good allocation of slots from
Google.
HSF/WLCG/OSG Workshop News and Planning
- Indico
- registration is open, please sign up.
- N.B. The JLab rate at the Marriott is a good one.
- There is a Google Sheet linked from the
main
page to help
coordinate car sharing from Dulles (you take your own
responsibility for this!).
- Please advertise again in your community.
- There is a
poster
that can be printed and put on your local notice boards.
- Reminder: David Lange, Michel Jouvin and Graeme Stewart organising
for HSF.
- General Session News:
- New Workflows, New Facilities (Tuesday AM).
- Benedikt Hegner and Maria Girone are convening.
- Closing Session, visionary talk and discussion on the future
(Friday).
- Amber Boehnlein has agreed to do the final talk.
- HSF Sessions:
- Software on Accelerators (2 slots).
- Graeme, Michel, David L will convene; send suggestions
(generators, tracking, etc. style projects)
- Various invitations issued now, but further
suggestions welcome.
- Talks that have generic conclusions evidently
favoured.
- Simulation (2 slots, with one in parallel with other HSF
sessions; idea is to use first slot for topics of general
interest, second for items that are more technical or narrow
in focus).
- Heather: It would be useful to get feedback about which
topics should be plenary vs more technical (I suspect that
most of the topics that we have would be of quite general
interest).
- fast simulation.
- physics models.
- common geometry tools.
- pile-up.
- vectorisation (Andrei).
- common geometry tools.
- simulation code on new architectures (particularly g4).
- Reconstruction (2 slots, 1 “parallel”)
- Brainstorming on topics, converging on Tuesday.
- Question on whether there could be more non-HEP-specific
talks (we have contacts from recent ITNs on interesting
topics).
- Initial topics (we will trim down/add, this is to see
overlap with other sessions):
- Hybrid architectures and accelerators for triggers.
- ML for reconstruction.
- ML for trigger.
- Connect to DarkMachines effort (some people from
Unsupervised Learning group based in the US).
- LHC: Run-3 preparations and expectations.
- HL-LHC: Updates on TDRs.
- Beyond colliders: what are smaller experiments
planning / doing (use Physics Beyond Colliders
platform to advertise a call for talks).
- Analysis (2 slots, 1 “parallel”)
- Using events at CERN to help plan contribution (see below).
- Software Tools (1 slot)
- Martin, Giulio and Serhan will discuss next week.
- Overlaps with analysis are interesting.
- Education and Training (1 slot)
- Initial thoughts from Sudhir/Dario:
- Highlight ramping of US centric training efforts
FIRST-HEP/IRIS-HEP.
- How can these efforts be expanded to Europe and beyond.
- Invite speakers/European counterparts (people who might
lead efforts there - NOT adhoc basis).
- Now that ITN deadline for European graduate schools
has passed maybe we can talk about them as a source of
coordinated effort (and maybe a step towards bigger
efforts)?
- PyHEP (1 slot)
- We met this week and converged on a draft agenda for the
HSF session at JLab.
- Basically a 30-minute part with lightning talks and the
rest for 2 topical presentations, leaving the last 20
minutes for an open discussion (e.g.: activities in 2019,
community interests).
- One spillover BoF slot.
Activity and Working Group Updates
General points:
- There is now a mailing list for all working group convenors
- hsf-wg-convenors (at) googlegroups.com.
- This is postable by all (does anyone know how to restrict
posting to members of other groups, like hsf-forum?).
- If you are involved in a WG and did not get an email then
please let Graeme know.
- We will allow this group to be able to edit the
HSF community
calendar
to add their meetings.
- Please set up mailing lists and advertise them on your group page.
- HSF website “Mailing Lists and Discussion Forums” links to all
WGs’ and activities’ pages.
- We can help with website
practicalities.
- For the first announcement of new group’s activities you can send
an email to the whole HSF using the hsf-forum list (any subscriber
can post).
- Activities can also be added to a banner announcement on the HSF
frontpage (see, e.g., this
commit).
Data Analysis
- Preparing 2 events at CERN:
- 23 Jan - what analysers distill from their experience.
- 13 Feb - technology used for analysis.
Detector Simulation
- We’ve been working on the agenda for the fast simulation session
at the workshop (see above).
- Planning the discussion on Fast Simulation for 30 January
- We plan to have 3 overview talks: ATLAS, CMS and LHCb and each
talk would be 20+10’.
- Potential topic for meeting on 6 March: Applications of machine
learning for fast simulation.
Reconstruction and Software Triggers
- We aim to have the first meeting: February 6th.
- Changed name of the mailing list:
https://groups.google.com/forum/#!forum/hsf-recotrigger
- We discussed among ourselves potential topics for the HSF/WCLG
workshop, will give more specific proposal next week.
- We aim to discuss with experiment representatives in the coming
weeks.
Event Generators
- Getting organized (discussing mailing lists, meetings…).
- Follow up of the November workshop.
- Contacted the Madgraph team about their GPU code - Stefan
Roiser interested in doing some concrete work and tests on
that.
- Not much progress on the write up of the workshop.
- Andy Buckley will give a theory plenary at ACAT also covering
the workshop.
- No specific plans for the JLAB workshop, some generator speakers
have been or may be contacted for other sessions
(GPU/accelerators, HPC).
Licensing
- Discussion about licensing in the AF meeting - how to relicense
Gaudi code (some parts currently (c) CERN and GPLv3).
- HSF will help to carry this discussion with CERN to a
conclusion.
- Still need to discuss FastJet license.
Training
- Got great feedback (~350 people) in response to training needs in
terms of software and computing for HEP.
Packaging
PyHEP
- Graeme is stepping down since over-committed. But will keep
contributing when possible - thank you so much, Graeme!
- Welcome to Jim Pivarski to the conveners team!
CWP
Publication status for Individual WG Papers
AOB
- We will do a clean-up of the HSF Coordinators list.
- Active coordinators and ex-officio members stay on the list.
- New proposals for active coordinators are always welcome.
- Reminder - if anyone gives a talk that is of relevance to HSF
activities we can add it to our
presentations
list.