Christian Svensson 1e0c21d088 lbm-gpu: one atomic per workgroup for the chunk's largest speed ...

Every cell did an atomicMax on the same word; a workgroup reduction first
leaves one per 256 cells. Same result, less contention on a single
address.

Part of #529.

2026-10-17 07:23:40 -04:00

..

decks

…

engine

solve.fluid: backend "cpu" (the default) or "gpu"

2026-10-17 07:23:40 -04:00

explicit

…

fem

…

geometry

…

gpu

…

io

…

lbm

solve.fluid: backend "cpu" (the default) or "gpu"

2026-10-17 07:23:40 -04:00

lbm-gpu

lbm-gpu: one atomic per workgroup for the chunk's largest speed

2026-10-17 07:23:40 -04:00

mcp

…

mesh

mesh: the tet objective sees needles; no tet below scaled Jacobian 0.3 on the skill's tet parts

2026-10-17 00:51:20 -04:00

model

…

parallel

…

render

…

solver

solver: a tie row nearly repeated on a held secondary node is dropped

2026-10-17 00:51:20 -04:00

testkit

mesh: the tet objective sees needles; no tet below scaled Jacobian 0.3 on the skill's tet parts

2026-10-17 00:51:20 -04:00

ui-egui

…

update

…