You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
In order to facility decomposition of work load for both threading and GPGPU.
The goal really, is that when we get to executing inside a solver, we shouldn't really care what sort of memory or objects we are dealing with, and inputs should be processed essentially as they are now but in potentially different forms, which are not necessarily patch based, or entire particle arrays.
This will also remove the need to use set on patch in a number of places, which is a coupling to SAMRAI
A view hierarchy should be created, and recreated per regrid, so that we have a set of pointers per patch, per level, and it should be easy to access particular members like it currently is on the hybrid state
potentially it might require a hierarchy per solver/state, ie. one for hybrid, one for mhd, or they might share a superclass
The text was updated successfully, but these errors were encountered:
In order to facility decomposition of work load for both threading and GPGPU.
The goal really, is that when we get to executing inside a solver, we shouldn't really care what sort of memory or objects we are dealing with, and inputs should be processed essentially as they are now but in potentially different forms, which are not necessarily patch based, or entire particle arrays.
This will also remove the need to use
set on patch
in a number of places, which is a coupling to SAMRAIA view hierarchy should be created, and recreated per regrid, so that we have a set of pointers per patch, per level, and it should be easy to access particular members like it currently is on the hybrid state
potentially it might require a hierarchy per solver/state, ie. one for hybrid, one for mhd, or they might share a superclass
The text was updated successfully, but these errors were encountered: