Illustrative eight-request KV handoff, rack versus local domain
370 ms / 210 msderived
| Value kind | derived — Derived values are guide calculations; their result is only as fixed as the stated inputs, scope, and method. |
|---|---|
| Scope | Joint-tail teaching bound: 180 ms other work, 1 GiB/request, eight requests, 45/300 GB/s service, 250 ms SLO. Rounded to 10 ms; no measured p99. |
| As of | 2026-09 |
| Source | Guide illustrative inputs and derivation; primary link supplies the method, not measurements or market prices. |
| Review | checking…review by 2026-12-05 · standard cadence |
| Recorded changes | none recorded |
| Claim id | part8-inference-request-tail-illustrative |
Where the guide uses it
Not quoted in a chapter yet; it is kept in the curated register.
← Full numbers register — every date-stamped figure in the guide, with revision history.