1 of 13

The Software Working Group collected information on the community's specific software tools and practices during the Yellow Report Initiative.

Q7. Do you have any comments on your current experience with EIC Software?

There are too many generators and simulation tools used at the moment.

Recall...

2 of 13

The ESC / EIC SWG has agreed for years, and has looked for ways to consolidate.

Towards which we’ve developed a proposal

For a simulation tool that is:

  • supported by the whole community,
  • developed and maintained together.

3 of 13

Project EAST (eA Simulation Toolkit)

“Fast and full simulations in Geant4 for large-scale detector systems with a plug and play modular approach”

Of course too long as a project name

but we now have an acronym:

4 of 13

This is not about a framework…

5 of 13

… but about a collection of tools…

6 of 13

… and their common development...

7 of 13

Towards a detector simulation tool

The point of commonality before collaborations branch to their specialized software.

Event

Generator

Detector

Simulation

Digitization

Reconstruction

  • Comprehensive, centrally maintained
  • Based on Geant4 for fast and full simulations
  • With library of potential detector options

8 of 13

Of course in 2021, for the proto-collaborations developing the detector proposals, using some newfangled non-existent common simulation tool is not an option.

But if we develop this tool in 2021, with its interfaces, the proto-collaborations have their convergence target.

So post-proposals in 2022, we can come together.

9 of 13

Detector

Simulation

  • comprehensive, centrally maintained application
  • based on Geant4 for fast and full simulations
  • with library of potential detector options

Requirements

  • ability to reuse existing simulation work
  • ease of switching detector options with comparable levels of detail
  • ease of switching between detailed and coarse detector descriptions
  • ease of leveraging new and rapidly evolving
    • technologies, e.g., AI/ML
    • computing hardware, e.g., heterogeneous architectures

Initial Focus

  • interface to MCEG output (but no further work on MCEGs)
  • clear separation of detector effects and responses (digitization)
  • common geometry interface between detector simulation tool and reconstruction tools

10 of 13

… in a project led by a renowned Geant4

leader and expert ...

11 of 13

Makoto Asai (SLAC)

  • Known to most/all of you.
  • Geant4 project leader and deep technical expert for >20yrs.
  • Has collaborated with EIC and NP community for many years on G4 simu needs.
  • Designer and principal developer of Geant4's capability to support both fast and full simulation.
  • Led Geant4’s multi-threaded reengineering to support high concurrency heterogeneous architectures, with excellent results in the memory economizations achieved.
    • Together with the integrated fast simulation support, opens the door to leveraging AI/ML in the unified simulation.
  • Now leading the next phase in concurrent Geant4, sub-event parallelism.
  • Expert in the geometry and modular detector description tools that will be the basis of unified geometry.
  • Expert in the Geant4 physics models that have to be tuned for NP experiments, having presided over their development and integration for much of the past 20 years.

12 of 13

Detector

Simulation

  • comprehensive, centrally maintained application
  • based on Geant4 for fast and full simulations
  • with library of potential detector options

Deliverables

  • Create CAD interface to the simulation (a distinct project that will occupy the next ~month).

  • Create an initial version of the fast and full simulation tool.
  • Produce a documented prototype of an existing simulation integrated in the new tool.
  • Develop and deliver a documented common physics list for EIC detectors.
  • Deliver a framework extensible to heterogeneous architectures using Geant4’s new task-based concurrency and sub-event level parallelism

13 of 13

… with a planned T0 a ~month from now, during which time we’d like your input

To tailor the project to priority needs, and a plan and outcome that serves all the (proto-)collaborations.