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
@joe61vette has some table-lookups that are very expensive and so we don't want to keep re-evaluating for every time we call a functor property at a given location
Design
very similar to what we do for the finite element arguments
Impact
speed up some finite volume computations at the expense of memory
The text was updated successfully, but these errors were encountered:
Got to use structured bindings and `try_emplace` for the first time.
I'm a little concerned about the cost of binary trees but I feel like
hashing of the finite volume functor arguments could be even more expensive.
I think if this shows up as an important piece of a profile, then we
can investigate an unordered_map implementation
Closesidaholab#20470
Reason
@joe61vette has some table-lookups that are very expensive and so we don't want to keep re-evaluating for every time we call a functor property at a given location
Design
very similar to what we do for the finite element arguments
Impact
speed up some finite volume computations at the expense of memory
The text was updated successfully, but these errors were encountered: