EU Einstein Toolkit Workshop, Lisbon, 2018

Workshop page: https://centra.tecnico.ulisboa.pt/network/grit/einsteintoolkit2018/

Discussion session

Proposed working groups review

WG3: maintainers club

• Ian to leave
• Helvi to join

WG4: matter codes

• google doc exists (Bruno)
• no more updates since
• suggestion: to coordinate / merge with IllinoisGRMHD

WG5: Initial data

• low priority

WG6: outreach & education

• already exists

WG7: Datavault

• hosted at NCSA
• Qian has finished postdoc and moved to University of Waterloo
• repository exits
• Sascha suggests to use public LIGO format https://dcc.ligo.org/LIGO-T1500606-v5/public, as is currently done already
• planning to get a student to continue working on it

Working groups update

Cosmology and particles (Helvi and Eloisa)

• Eloisa has left; volunteers welcome to join lead the group
• Goals: Modified theories, High-energy, cosmology, ...
• diverse team, different expertise. has people that work with other codes (not only ET based)
• idea is to match people doing theory and numerical development
• regular calls every first monday of month at 16h30 (UK, PT), 10h30 (CDT)
• idea is to move forward with specific projects
• everyone welcomed to join

Task scheduler (Erik, Steve, Sam, Roland)

• essentially Steve and Sam
• see Steve and Sam's talks in the workshop
• backwards compatibility
• figure out difficulties with AMR and then update ET thorns to use it
• future: automatic allocation of memory (in storage)

Education and outreach (Matt, Gab, Helvi)

• one call early this summer
• Steve wants to join and contribute jupyter notebooks with tutorial and such
• Gab working with undergraduate students
• for outreach: use ET logo in all movies done with ET for visibility
• generic.cfg should work out of the box in any computer outside of Macs.
• problem with Macs that it called "gcc" what is actually clang (which doesn't understand OpenMP, for instance).

IllinoisGRMHD (Zach)

• Zach has major contribution to upcoming release (five new thorns, more thorns proposed)
• very active team

Spherical coordinates

• not officially approved
• RIT internal
• should contact RIT folks

Performance and optimization (Erik, Roland, Helvi, Eric West)

• idea: make the code faster
• build infrastructure to easily check speed on new clusters and machines
• task based parallelism in Cactus
• someone needs to merge master with ianhinder/rkprol, benchmark, and take over
• many bits and pieces have already been implemented and pull requests exist to merge with ET master
• NRPy: see Zach's presentation
• lots of work being done, but mostly behind the scenes
• Helvi to do scaling and performance comparison with BAM, etc
• anyone interested welcome to join

Other items

• memory leak: Wolfgang to open ticket.
• list of papers to cite after running ET with some list of thorns.
• use something like joss for citations and credit?
• ETK waveform catalog