# [Greenline] Open projects TABLE

This table got out of hand and is too wide, any ideas on how to improve it?

| Nr | Name | Priority | Status | Who | Depends on | PR | Shape doc | Notes |
| --- | ----------------------------- | -------- | ------ | -------------- | -------------------- | ---------- | ----------------------- | ------------------------------------------------------------------------------------------ |
| 1 | Driver configuration | Medium | WIP | Enrique | | [PR936] | | [Config notes] |
| | | | | | | | | |
| 2 | I/O decisions | Low | | | | | | IO monitor or [HIOPY] or ADIOS (Jean Favre) + restartIO [IO notes] |
| 3 | I/O config | Low | | | 1,2 | | | |
| 4 | I/O coding | Low | WIP | | 2,3 | | | Either continue Magdas IO monitor or interface [HIOPY] |
| | | | | | | | | |
| 5 | CI for standalone driver v1 | High | | | | | | Hardcoded config (ideally the same used for warm bubble) -> ser-data |
| 6 | CI for standalone driver v2 | Medium | | Y | 1,5 | | | Proper config -> ser-data |
| 7 | CI for standalone driver v3 | Low | | N+CR | 1,6 | | | Proper config -> scientific validation (JW) [Vali notes] |
| 8 | Delete old driver + rename | Low | | | 1,7 | | | |
| | | | | | | | | |
| 9 | MPI on CI | High | WIP | N+Mikael | 12b | [PR692], [PR1012] | | [MPI on CI notes], [MPI CI project] |
| 9b | MPI on CI verify with serial | Low | | | | | | Get the tests to verify with mpitask1 instead of mpitask2/4 |
| 10 | MPI min reduction | High+ | Done | N | "9",12b | [PR966] | | |
| 11 | MPI sum,mean,max reduction | High+ | Done | N+J+Mikael | "9",12b | [PR988] | | [Global reduction notes] |
| 12 | Domain decomposition | High+ | WIP | J+N+Mikael | 9 ,12b | [PR540] | [Decomp shape-up] | [Domain dec notes] |
| 12b | MPI-serialize missing exps | High+ | WIP | J | | [PR988] | | Gauss, JABW, WK are missing mpitask2/4 (new polybox needed) torus cartesian coord |
| 13 | MPI communicator(s) | | | N+J | | | | [Communicator notes] |
| | | | | | | | | |
| 14 | Restart | | WIP | J | "1",5,2 | | | Half implemented in IBM branch |
| | | | | | | | | |
| 15 | Tracer adv | High | | CR+N+Rico | 16,17,18 | | | Check what is missing [Tracer adv notes] |
| 16 | Tracer adv port LS coeffs | High | Done | N+Rico | | | [LS coeff. shape-up] | We need to port least square coeffs which are used in Miura 2nd-order scheme |
| 17 | Tracer adv needs exchange | | WIP | N | | | | Add halo exchanges to tracer adv (seems that exchange has been added in advection granule) |
| 18 | Tracer adv identical loop | High | WIP | CR+Rico | | | [tracer impl. shape-up] | Add a loop over N tracers, same scheme |
| 18b | Tracer adv proper loop | Low | | CR+Rico | | | | Modify the loop so that you can choose the individual adv schemes |
| 19 | Tracer adv GT4Py container? | | | CR+Enrique | | | | To avoid manual loop/impl for multiple tracers |
| | | | | | | | | |
| 20 | Microphysics | High | | CR | 18 | | | Add the proper calls to diagnostics and sat-adj + tracer update |
| | | | | | | | | |
| 21 | Turbulence decisions | High | Done | A+Y | | | | 1D or 3D turbulence? 1D multi GPU |
| 22 | Turbulence 1D CPU single node | High | WIP | Y | | | | Implement ICON4Py granule, code is in [Turbulence icon-exclaim] |
| 23 | Turbulence 1D GPU multi node | High | WIP | Y+Will+Andreas | 22 | | | F2Py bindings with OpenACC works or not? [Turbulence notes] |
| 24 | Turbulence 3D | | | Y | | | | |
| | | | | | | | | |
| 25 | Clean physics interface | | | | | | | Clean interface (look at the Component) |
| | | | | | | | | |
| 26 | Land | High | | | | | | Implement simple land (constant, uniform fluxes) |
| | | | | | | | | |
| 86 | Add unit tests for geo2cart | Low | | | | | | math/helpers.py geo2cart_onX don't have unit tests |
| 87 | Proper implementation KHalf | Medium | WIP | Till+Christoph | | [GTPR2339] | | Properly differentiate KDim from KHalfDim |
| 88 | Cleanup of the entire repo | Low | | | | | | Remove old, unused code; refactor/organize stencils |
| 89 | Write some kind of docs | Low | | | | | | Possibly start from the IBM branch docu |
| | | | | | | | | |
| 99 | Warm bubble | | | | 15-18,20,21-23,25,26 | | | Align setup with MPI-M CI test [MR919], [Warm bubble notes] |

[PR540]: https://github.com/C2SM/icon4py/pull/540
[PR692]: https://github.com/C2SM/icon4py/pull/692
[PR1012]: https://github.com/C2SM/icon4py/pull/1012
[PR936]: https://github.com/C2SM/icon4py/pull/936
[PR966]: https://github.com/C2SM/icon4py/pull/966
[MR919]: https://gitlab.dkrz.de/icon/icon-mpim/-/merge_requests/919
[GTPR2339]: https://github.com/GridTools/gt4py/pull/2339
[PR988]: https://github.com/C2SM/icon4py/pull/988

[tracer impl. shape-up]: https://hackmd.io/YwB5wpaARZOo22loClkMUw?both
[LS coeff. shape-up]: https://hackmd.io/z84FEYwkToCiqNicr3K5jQ
[Decomp shape-up]: https://hackmd.io/3aTBq9-6QbKZrHOtsA-GpA

[Config notes]: https://hackmd.io/@gridtools/HyygMYpGbx#2-Plugin-configuration-in-driver
[IO notes]: https://hackmd.io/@gridtools/HyygMYpGbx#4-IO
[Vali notes]: https://hackmd.io/@gridtools/HyygMYpGbx#3-Think-of-automated-scientific-validation
[MPI on CI notes]: https://hackmd.io/@gridtools/HyygMYpGbx#1-MPI-tests-on-CI
[Global reduction notes]: https://hackmd.io/@gridtools/HyygMYpGbx#2-Global-reduction-for-mean-values
[Domain dec notes]: https://hackmd.io/@gridtools/HyygMYpGbx#3-domain-decomposition-itself
[Communicator notes]: https://hackmd.io/@gridtools/HyygMYpGbx#5-Communicator-structure
[Warm bubble notes]: https://hackmd.io/@gridtools/HyygMYpGbx#C-Warm-bubble
[Turbulence notes]: https://hackmd.io/@gridtools/HyygMYpGbx#3-Turbulence
[Tracer adv notes]: https://hackmd.io/@gridtools/HyygMYpGbx#2-Tracer-Advection

[Turbulence icon-exclaim]: https://github.com/C2SM/icon-exclaim/tree/turbulence_1d

[HIOPY]: https://ican.pages.gwdg.de/hiopy/

[MPI CI project]: https://hackmd.io/O4Fymu1dTxqTZSC8rdiVVw