Improvements in Graupel code

[iomaganaris]

• Added an if-statement to make sure that we only enter the update functions only if there is some mask or previous level activated
• Restructure precip_qx_level_update for better if-statement handling
• Added gpu_maxnreg = 128 option to the optimizer to improve the scan kernel performance
  • Needs feat[next-dace]: Enable setting gpu_maxnreg attribute in maps GridTools/gt4py#2464

[havogt]

Can we swap the conditionals to avoid the code duplication? I wouldn't like to take this code, because it's clearly less readable.

What's the impact of this change?

[iomaganaris]

I reverted this change since I didn't see any significant speed up in the end with my changes

[havogt]

I would like to refactor to the following. Do you think it has performance implications?

    if current_level_activated:
        vt = previous_level_q.vc * prefactor * power((rhox_prev + offset), exponent) if previous_level_q.activated else 0.0
        x = (zeta * (flx_eff - flx_partial)) / ((1.0 + zeta * vt) * rho)  # q update
        p = (x * rho * vt + flx_partial) * 0.5  # flux
    else:
        x = q
        p = 0.0

[havogt]

TODO

(moving kmins inside didn't make a difference?)

[iomaganaris]

It made the code much slower. I didn't look into why though

[havogt]

I think it's slightly nicer like this anyway.

[github-actions]

Mandatory Tests

Please make sure you run these tests via comment before you merge!

• cscs-ci run default
• cscs-ci run distributed

Optional Tests

To run benchmarks you can use:

• cscs-ci run benchmark-bencher

To run tests and benchmarks with the DaCe backend you can use:

• cscs-ci run dace

To run test levels ignored by the default test suite (mostly simple datatest for static fields computations) you can use:

• cscs-ci run extra

For more detailed information please look at CI in the EXCLAIM universe.