The box pruning series
Links for the “box pruning revisited” series:
- Part 1 - the setup
- Part 2 - compiler options
- Part 3 - don’t trust the compiler
- Part 4 - sentinels
- Part 5 - hardcoding
- Part 6a - data-oriented design
- Part 6b - less cache misses
- Part 7 - integer comparisons
- Part 8 - branchless overlap test
- Part 9a - SIMD
- Part 9b - better SIMD
- Part 9c - data alignment
- Part 10 - integer SIMD
- Part 11 - the last branch
- Part 12 - ASM FTW
- Part 13 - code alignment
- Preview of Part 14 -loop unrolling
- Part 14a - loop unrolling
- Part 14b - Ryg rolling
- Part 14c - that’s how I roll
- Part 14d - integer comparisons redux
- Part 14e - a bugfix and a reboot
- Part 15 - AVX
- Part 16 - improved pair reporting
- Part 17 - PhysX lessons
- Part 18 - Recursive box pruning
- Part 19 / 20 / etc - multi threading
To be continued…