| 2026-10-10 20:50 | transpose 4x25/8x25 state in vector blocks instead of scalar gather | finwo | 3 | +132 | -18 |
| 2026-10-10 18:05 | perm/unroll fixes for -O2 gcc on avx | finwo | 2 | +96 | -47 |
| 2026-10-10 16:50 | turboshake batching | finwo | 10 | +131 | -29 |
| 2026-10-10 15:28 | Bench all backends | finwo | 6 | +100 | -66 |
| 2026-10-10 15:12 | named-register permutation with fused theta and 4x round unroll | finwo | 3 | +216 | -72 |
| 2026-10-10 14:45 | Hint unroll factor to compiler | finwo | 2 | +20 | -1 |
| 2026-10-10 14:32 | Add turboshake | finwo | 8 | +103 | -50 |
| 2026-10-10 14:16 | Add avx512 backend | finwo | 4 | +127 | -4 |
| 2026-10-10 14:12 | Add scalar+bmi backend | finwo | 6 | +223 | -137 |
| 2026-10-10 13:51 | Repo init | finwo | 17 | +1014 | -0 |
