Files
ResInsight/ApplicationLibCode
Magne Sjaastad b367a65062 #15110: Parallelize duplicate grid node removal
Insert nodes concurrently into a lock-free hash table keeping the lowest index per coordinate, then number unique nodes per fixed block, so the result is independent of thread count. Reduces removal time from 1.8 s to about 0.45 s for 78.4M nodes. Add a disabled performance test and docs/remove-duplicate-grid-nodes.md describing the algorithm.
2026-11-05 18:33:07 +01:00
..
2026-11-02 11:34:30 +01:00
2025-10-31 13:21:19 +01:00
2025-11-19 10:25:32 +01:00
2024-07-01 14:09:56 +02:00