eng-0167
window recorded
citable URL: https://halobench.com/records/eng-0167/ — this address never moves; the anchor /records/#eng-0167 keeps resolving
9.68 Whmean 119.77 W − 10.1 W idle floor (box-idle-all-empty) = 109.67 W attributable
window 2026-08-17T16:54:29Z → 2026-08-17T16:59:20Z · provenance recorded
method wall-meter · aihydra energy sensor via Home Assistant (counter-difference on sensor.hardware_ai_hydra_energy, raw ~10s-resolution history samples, linearly interpolated to the run-meta window) · no contention
per task 2.4205 Wh · 0.0733 p at 30.3 p/kWh
covers run-0432
aihydra qwen3-coder-next-80b-screen SMOKE window (16:54:29Z-16:59:20Z, 291s), matching run-0432 exactly. Counter interpolated from raw ~10s-resolution HA history samples: start 25.915112 kWh, end 25.924794 kWh. wh_per_task is Wh-per-CORRECT-ANSWER: wh_total / tasks_passed (4 of 5, run-0432.metrics) = 2.4205 Wh, 0.0733 pence at the standing 30.3 p/kWh tariff. WHOLE-SESSION energy (server already warm from the FIT probe moments earlier; includes simulator wait time end to end). A 5-task SMOKE window, not a like-for-like comparator against any full standard-26-task tau2 Wh/correct figure on this board (see eng-0168 for that comparison).
Wh, never joules — read as a cumulative-counter difference across the window; delta over the idle baseline is the only figure that means anything