feat[fastlanes]: add optimized 1024-bit transpose implementations#6135

Draft

joseph-isaacs wants to merge 10 commits intodevelopfrom

claude/bitpacking-transpose-optimization-tM1U4

Commits on Jan 24, 2026

feat[fastlanes]: add optimized 1024-bit transpose implementations
claude
committed
chore[fastlanes]: add perf measurement examples for transpose
claude
committed
feat[fastlanes]: add BMI2 PEXT/PDEP transpose and fix GFNI implementations
claude
committed
feat[fastlanes]: add scalar_fast and ARM64 NEON transpose implementations
claude
committed
perf[fastlanes]: fully unroll BMI2 transpose for 12% performance gain
claude
committed
test[fastlanes]: add verification against fastlanes crate transpose
claude
committed
feat[fastlanes]: add AVX-512 VBMI transpose with 7.5x speedup
claude
committed
feat[fastlanes]: add dual-block VBMI transpose for 10% more throughput
claude
committed
feat[fastlanes]: add 4-block VBMI transpose for 7% additional speedup
claude
committed

Commits on Jan 26, 2026

docs[fastlanes]: add transpose optimization plan and results
claude
committed