Go to the Indico Home Page Oldest event Older event Up to category Newer event Newest event Printable version
  • Compact style
  • Indico style
  • Indico style - inline minutes
  • Indico style - numbered
  • Indico style - numbered + minutes
  • Indico Weeks View
    • Switch to the management area of this event
    Login

    EP R&D Software Working Group Meeting

    → Europe/Zurich
    Vidyo

    Vidyo
    Live Notes
    Hide

    Software R&D Working Meeting Minutes
    ====

    Apologies: Jakob (GDB meeting)

    Introduction
    ---

    No additional points raised.

    Tracking
    ---

    Technical implementation of 'toolbox' - how to be flexible and still performant?

    There are some 'assumptions' made, e.g. timing, but these are not too restrictive and fit quite generically.

    Moritz re-integrated all of the examples back into the new GitHub mono-repo.

    There is a 'mini-framework' for testing, but not bound to this.

    Rely on compiler a lot - heavy compilation, fast runtime.

    External geometry package used as "source", then building ACTS geometry (flexible on source).

    Next hit can be framed as a nearest neighbour problem, links to many generic ML/hashing libraries. Search time is ~linear (cf. combinatorial factorial explosion).

    Thinking about GPUs lights up many places where CPU code can be improved. Using C++17 library features is in tension with using accelerators where these features are not available. Lesson - always think of *containers* of objects, not just objects.

    Are there common topic with HGCAL? Templated detector is a great idea - puts algorithms on a common footing that is very helpful. DD4hep is the geometry description for this; there are a few parameters. Currently a Si detector - there are thoughts to add a calorimeter around the tracker.

    Valentin - lots of development happening, LCG release is quite far behind.

    Is there going to be a v1.0 release, stabalising the EDM/interfaces? Trying not to change these, but there are still some simplifications still to be done. Adhere to semantic versioing, so 1.0 should be stable. Don't want to lock this in right now.

    Calculation of hit distances? Navigation knows candidate surfaces (caches carried). Checked in propagation step (it's quite fast). Can have certain tolerances.

    Erica - is there use of acts by other communities, e..g neutrionos. Active particpation from Belle II, CEPC, ATLAS, sPhenix, FASER. Not really neutrinios yet.

    Adaption to lockdown - no negative effects observed. Daily fast team meetings and live logbooks to help with communications.

    There are minutes attached to this event. Show them.