twogranule

confirmedshipC · Apple M4Systems & data structuresJuly 14, 2026

the question

Is the effective false-sharing granule on Apple M4 64 bytes within a core cluster and 128 bytes across clusters?

what came out

The false-sharing recovery edge sits at 64 bytes within a core cluster (1.84 ns/op) but at 128 bytes across clusters, where the same 64-byte-separated layout still shares (9.15 ns). A read-only load control stays flat (~0.8-1.5 ns), so the penalty is write-ownership-driven. This inverts the pad-to-128B rule.

method & receipts

Result: confirmed
Pre-registered — the prediction was written down and committed before the run.
Reproducible — one script re-runs the whole thing from scratch.
Tested — a correctness/benchmark suite ships alongside the code.

→ read the code and re-run it

github.com/v-code01/twogranule