eng-0177
window recorded
citable URL: https://halobench.com/records/eng-0177/ — this address never moves; the anchor /records/#eng-0177 keeps resolving
73.35 Whmean 154 W − 10.1 W idle floor (box-idle-all-empty) = 143.9 W attributable
window 2026-08-11T20:08:30Z → 2026-08-11T20:37:05Z · 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 73.3464 Wh · 2.2224 p at 30.3 p/kWh
covers run-0111
aihydra qwen35-122b task-9 regression arm window (2026-08-11T20:08:30Z..2026-08-11T20:37:05Z, 1715s), matching run-0111's recorded run-meta window. Counter interpolated from raw ~10s-resolution HA history samples pulled fresh 2026-08-18 as part of the run-0101/run-0102/run-0111 batch: start 11.444064340 kWh, end 11.517410772 kWh. wh_per_task is Wh-per-CORRECT-ANSWER: wh_total / tasks_passed (1 of 8 trials, run-0111.metrics) = 73.3464 Wh, 2.2224 pence/correct at the standing 30.3 p/kWh tariff. This is a narrow task-9 regression run with the pinned Haiku simulator, not a full field-comparison arm.
Wh, never joules — read as a cumulative-counter difference across the window; delta over the idle baseline is the only figure that means anything