docs: measured ARM numbers, and a correction

On a Raspberry Pi 5 at N = 100 000 and equal recall (0.9940 vs 0.9945),
medians of three runs:

  f32             33 413 ms build   6 164 QPS
  int8 scalar     18 950 ms         ~6 190 QPS   (what v2.7.0 shipped)
  int8 NEON      ~17 000 ms          6 640 QPS
  int8 SDOT       14 464 ms          7 267 QPS   1.18x f32, 2.3x build

The docs said quantised search stayed off by default because aarch64
"falls back to the scalar loop, where the original trade still
applies" — that it was ~13% slower than f32 there, as on x86. That was
extrapolated rather than measured, and it was wrong: x86's portable
baseline is SSE2 against hand-written AVX2 f32 kernels, but on aarch64
NEON is the baseline and the scalar loop vectorises well, so it already
matched f32. Corrected in BENCHMARKS.md, README.md and CLAUDE.md; the
released v2.7.0 changelog entry is left as it was and the correction is
recorded in a new one.

Labelled as Pi 5 figures throughout — a Pi's memory bandwidth and cache
are far below an M-series or flagship phone, so the ratios will move.

Co-Authored-By: Claude Opus 5 (1M context) <[email protected]>

BENCHMARKS.md

@@ -108,11 +108,37 @@ second**, builds **1.8x faster**, and holds a quarter of the vectors. (Compare
 only at equal `ef`: with re-scoring the harness raises `ef` to at least the
 candidate pool, so the `ef = 16` and `ef = 32` rows are not like-for-like.)
 
-It is still **off by default**, for portability rather than performance: the
-int8 kernel is AVX2-only, and on aarch64 — including `clawhdf5-android` — it
-falls back to the scalar loop, where the original trade still applies. A NEON
-kernel would remove that caveat. On an x86-64 deployment, turning it on is a
-win on every axis measured.
+#### On ARM (Raspberry Pi 5, Cortex-A76)
+
+`dot_i8` has two aarch64 kernels: `SDOT` for CPUs with the ARMv8.2
+dot-product extension (Cortex-A76 and later, Neoverse-N1, all Apple Silicon)
+and plain NEON (`vmull_s8` + `vpadalq_s16`) otherwise. Medians of three runs
+at N = 100 000, ef = 64, recall@10 0.9940 in every int8 row against f32's
+0.9945:
+
+| int8 kernel | build | QPS | vs f32 |
+|---|---:|---:|---:|
+| *(f32 baseline)* | 33 413 ms | 6 164 | 1.00x |
+| scalar (what v2.7.0 shipped) | 18 950 ms | ~6 190 | 1.00x |
+| plain NEON | ~17 000 ms | 6 640 | 1.08x |
+| **SDOT** | **14 464 ms** | **7 267** | **1.18x** |
+
+These are Pi 5 numbers, not "ARM" numbers: a Pi has far less memory bandwidth
+and cache than an Apple M-series or a flagship phone, so the ratios will move
+on other hardware. The plain-NEON row is that code on an A76 with `SDOT`
+disabled, not a measurement of a pre-A76 core.
+
+**A correction.** Until this was measured, this section said aarch64 "falls
+back to the scalar loop, where the original trade still applies" — that is,
+that quantised search was ~13% slower than f32 on ARM. That was extrapolated
+from x86 and it was wrong. On x86-64 the portable baseline is SSE2 while the
+f32 kernels are hand-written AVX2, so scalar int8 lost; on aarch64 NEON *is*
+the baseline, the compiler vectorises the scalar loop well, and scalar int8
+already matched f32 for search while building 1.76x faster.
+
+So on every configuration measured — x86-64 AVX2, and Pi 5 with each of the
+three int8 kernels — the quantised index is at least as fast as f32 at equal
+recall, builds faster, and holds a quarter of the vectors.
 
 A measurement trap worth recording: the synthetic `clustered` generator in the
 `clawhdf5-ann` tests draws clusters far tighter than any real embedding, so

CHANGELOG.md

@@ -1,5 +1,30 @@
 # Changelog
 
+## Unreleased
+
+### Performance
+- `clawhdf5-accel`: **`dot_i8` has aarch64 kernels** — `SDOT` for CPUs with
+  the ARMv8.2 dot-product extension (Cortex-A76 and later, Neoverse-N1, every
+  Apple Silicon generation) and plain NEON (`vmull_s8` + `vpadalq_s16`) for
+  the rest, selected at runtime. `SDOT` is issued through inline assembly,
+  because the `vdotq_s32` intrinsic is still behind the unstable
+  `stdarch_neon_dotprod` feature. On a Raspberry Pi 5 at N = 100 000 and
+  equal recall, the quantised index answers **1.18x the queries per second**
+  of f32 (7 267 vs 6 164) and builds **2.3x faster** (14 464 vs 33 413 ms).
+  Both kernels are tested bit-for-bit against scalar on real hardware, each
+  explicitly — dispatch only ever takes one path on a given CPU, so testing
+  through it alone would have left the plain-NEON fallback unexercised on any
+  machine with `SDOT`.
+
+### Corrections
+- The v2.7.0 entry for `dot_i8` said `quantized_index` stayed off by default
+  because "aarch64 falls back to the scalar loop", implying the ~13% search
+  penalty measured on x86 applied on ARM too. It did not. That figure came
+  from scalar int8 against hand-written AVX2 f32 kernels on x86, whose
+  portable baseline is SSE2; on aarch64 NEON is the baseline, and measured on
+  a Pi 5 the scalar int8 loop already matched f32 for search while building
+  1.76x faster. The claim was extrapolated rather than measured.
+
 ## v2.7.0 (2026-09-20)
 
 ### Upgrade Notes

CLAUDE.md

@@ -44,10 +44,12 @@ Cargo workspace with 16 crates under `crates/` (plus `libaec-sys`, an internal F
   which roughly halves a loaded store's memory (2.72x -> 1.74x the raw vectors
   at 100K); because quantised distances are approximate and `ef` cannot
   compensate, the query path then re-scores the candidate pool against the
-  exact embeddings, which holds recall at the f32 index's level. On AVX2 it is
-  also 1.63x the QPS and 1.8x the build speed (`clawhdf5_accel::dot_i8`); it
-  stays off by default only because that kernel is AVX2-only and aarch64 falls
-  back to scalar. `hybrid_search` keeps one incremental BM25
+  exact embeddings, which holds recall at the f32 index's level. It is also
+  faster at equal recall: 1.63x the QPS on x86-64 (AVX2) and 1.18x on a
+  Raspberry Pi 5 (`clawhdf5_accel::dot_i8`, NEON `SDOT` via inline asm since
+  the intrinsic is unstable; plain NEON on pre-dotprod cores). The aarch64
+  code is `cfg`'d out on x86, so x86 CI never compiles or lints it — test it
+  on real ARM (`rpivision02`, 10.0.2.3, is a Pi 5). `hybrid_search` keeps one incremental BM25
   index for the life of the store and never writes the store: Hebbian
   activation boosts are persisted by the next checkpoint (or on drop), not per
   query. Measure any search-path change with

README.md

@@ -442,10 +442,9 @@ copy of the embeddings as `i8`, roughly halving a loaded store's memory
 (2.72x -> 1.74x the raw vectors at 100k x 384). Quantised distances are
 approximate, so the query path re-scores the candidate pool against the exact
 embeddings the store already holds, which keeps recall at the `f32` index's
-level. On AVX2 it is also **faster** — 1.63x the queries per second and 1.8x
-the build speed at equal recall — because the int8 kernel is SIMD too. It
-stays off by default only because that kernel is AVX2-only and aarch64 falls
-back to a scalar loop. See `BENCHMARKS.md`, "Quantising the index copy".
+level. It is also **faster**: 1.63x the queries per second at equal recall on
+x86-64 (AVX2) and 1.18x on a Raspberry Pi 5 (NEON `SDOT`), with index builds
+1.8x and 2.3x faster respectively. See `BENCHMARKS.md`, "Quantising the index copy".
 | `parallel` | no | Rayon parallel search |
 | `fast-math` | no | BLAS matrix-vector multiply |
 | `accelerate` | no | Apple Accelerate / AMX (macOS) |