[Home]
[Feed]
[Lectures]
[Exercises]
[Projects]
[Login]
Domain-Specific Programming Systems
Previous
|
Next
--- Slide 21 of 69
Back to
Lecture Thumbnails
nemo
Two methods discussed during the lecture to parallelize the for loop -
Store the flux on edges at first and in another pass update the vertex fields (both can happen in parallel).
Find disjoint set of edges for which the for loop can run in parallel.
Two methods discussed during the lecture to parallelize the for loop -