HSF Weekly Meeting #178, 23 January, 2019 - Lund Workshop Planning Meeting

Present: Graeme Stewart, Pere Mato, Agnieszka Dziurda, Philippe Canal, Sam Meehan, Stefan Roiser, TJ Khoo, Andrea Valassi, Serhan Mete, Witek Pokorski, Paul Laycock, Mason Proffitt, Ben Morgan, Markus Schulz, Liz Sexton-Kennedy, David Lange, Josh McFayden, Kyle Knoepfel, Mark Neubauer, Efe Yazgan, Daniel Elvira, Caterina Doglioni, Chris Jones

Organising team for HSF are: Graeme, Heather, David and Michel - please get in touch with any suggestions you have.

Plenary

Parallel

Working Group Thoughts and Feedback

Reconstruction and Software Triggers

See above for an interdisciplinary thought from Caterina. copied from above:

Accelerator crossover with simulation, generators and frameworks… identify generic work where possible.

Detector Simulation

We think that one full day of parallel session on simulation is a very good idea, taking into account how critical simulation CPU performance and physics precision is becoming. We would probably like to have 4 different sessions along the general axis of improvements/modernization of Geant4, fast simulation and new hardware technologies (accelerators, HPC, etc). Physics precision is also a very important topic that should not be ignored, but this should in principle be covered by the LPCC workshop that we are planning to have in the second half of 2020. We would be very pleased to have more Geant4 community involved. We would also be interested to hear about successful attempts of porting simulation software to accelerators from other communities (outside HEP).

Data Analysis

Topics for the start of year are Analysis languages and Analysis meets DOMA, our first WG meeting is tomorrow :) Cross-over topics in other WG parallels might be of interest, seeing as we needn’t have a dedicated session:

Frameworks

Not certain who will be at Lund yet from FNAL. Cross cutting topics would be really interesting - scheduling accelerator jobs, portability issues. How to make use of an HPC as an application. Exploring different frameworks and examining complications… once finished

PyHEP

Tools and Packaging

Serhan : Will probably be able to discuss tomorrow. Last two HOW workshops, we had 1h30min parallel sessions with approximately 50% talk

Training

Sam M. : I’m sorry that I could not attend the meeting [ref. Coordination meeting 16 Jan]. I do think that having a session would be good and I think that if we want to consider outcomes, then having the critical mass of people to brainstorm what types of bootcamps we need would be helpful. We can then discuss this with the experience of the last year in mind [specifically how we held the workshops] and that would provide the concrete list of events that we need to plan in the coming year - spanning ~July 2020 - July 2021. Caterina: another “crazy host” idea: would you be willing (and would there be interest) to organize a mini-bootcamp before/after the workshop? Students may be encouraged to fly on Sunday night and in the morning we have the rooms. In Lund we have huge interest from all faculties on our Jupyter course and if we do something generic we can “expand our horizons”. → Caterina will follow up with Sam towards the end of next week.

Generators

We have a generator meeting at 4pm after this one, where we will also discuss the LUnd workshop more in detail. Lund is sort of the ‘home’ of MC generators and there are people there working in Pythia, MadGraph, Herwig, possibly more. It would certainly be nice to organise one parallel session to meet and engage with some of those experts, who are not normally involved in our WG activities. We have not yet made contacts but will do soon. So in principle we request one parallel slot (90 minutes?). We could have a plenary talk if people find this useful [Andrea more generally suggested that each WG gives one, but some people feel otherwise]

AOB

We are asked for a letter of collaboration for an AI Institute proposal going to the HSF, “AI Institute: Community Laboratory for AI Research at the Intersection with Physics (CLARIPHY)” (from Pete Elmer). The HSF’s collaboration would be in training and workshops: https://docs.google.com/document/d/13YHqtL3UI0QV5KQ8fUQ0ab0jiDzs7Nw089lbzjW-rS4/edit?usp=sharing Can we approve this? Comments until tomorrow morning CERN time as the proposal deadline is very close. We would like a ~1 page summary so that we can check we approve the idea. Mark: there will likely be other asks for this call. Graeme: we can sign off without a meeting as long as the letter and project description are circulated.