box3d/benchmark
Repository files (latest commit first)
Filename Latest commit message Latest commit date
Erin Catto 8f3b2e12eb
Manifold allocation cap (#227)
Degenerate scenarios can lead to a huge number of manifolds per contact.
This creates memory and performance problems. So I'm capping them to 8
using cheap heuristics for culling.
2026-11-17 13:51:59 -08:00
..
amd7950x_avx2 Tree perf (#179) 2026-10-06 09:51:48 -07:00
amd7950x_scalar SIMD hull collision (#93) 2026-07-18 22:43:03 -07:00
amd7950x_sse2 SAT optimizations (#170) 2026-09-24 21:16:19 -07:00
m2air_neon SIMD hull collision (#93) 2026-07-18 22:43:03 -07:00
CMakeLists.txt Reduce ghosts (#212) 2026-11-10 21:53:10 -08:00
ghost_baseline.csv Static query (#214) 2026-11-10 23:05:04 -08:00
ghost_baseline_hf.csv Static query (#214) 2026-11-10 23:05:04 -08:00
ghost_scan.c Static query (#214) 2026-11-10 23:05:04 -08:00
main.c More static queries (#216) 2026-11-11 17:41:42 -08:00
mesh_benchmark.c Reduce ghosts (#212) 2026-11-10 21:53:10 -08:00
mesh_benchmark.h Reduce ghosts (#212) 2026-11-10 21:53:10 -08:00
mesh_soak.c Manifold allocation cap (#227) 2026-11-17 13:51:59 -08:00
soak_baseline.csv Manifold allocation cap (#227) 2026-11-17 13:51:59 -08:00