WarpX icon indicating copy to clipboard operation
WarpX copied to clipboard

warpx.roundrobin_sfc: A runtime parameter to control distribution mapping

Open WeiqunZhang opened this issue 9 months ago • 1 comments

The default is false. If it's true, AMReX's RRSFC strategy will be used to override the default SFC strategy used by amrex::AmrCore. The motivation for this is that this might mitigate the load imbalance issue during initialization by avoiding putting neighboring boxes on the same process.

WeiqunZhang avatar May 03 '24 00:05 WeiqunZhang

Looks good. Would his affect the runtime, slowing the MPI communication since neighboring boxes will be far away from each other?

dpgrote avatar May 04 '24 01:05 dpgrote

@WeiqunZhang sorry for the delay, can you please rebase?

ax3l avatar Aug 12 '24 21:08 ax3l

Thank you, @WeiqunZhang! I will test/use this down the line. :)

n01r avatar Aug 13 '24 01:08 n01r

Thank you so much!

After checking back with our collaborators, we currently do not have a reproducer at hand, but I'll include it in the coming release so we can try it when we should find one again.

@n01r might have one we can try this on, too.

ax3l avatar Aug 28 '24 20:08 ax3l

@dpgrote

Looks good. Would his affect the runtime, slowing the MPI communication since neighboring boxes will be far away from each other?

The idea is to overcome out-of-memory in the init routines (e.g., for LPI sims with a lot of particles in close-by ranks) and then apply load balancing based on a user-defined cost (and distribution) model, as usual.

ax3l avatar Aug 28 '24 20:08 ax3l