# Hahn T2, wider: ibm_boston, ibm_pittsburgh, ibm_phoenix (Darkwave Disco, paid) and ibm_fez on a fresh record (open instance)
Written and pushed 22 Oct 2026 night BEFORE these jobs. hahn.py unchanged (CODE_HAHN.sha256); the Darkwave chips run through
fire_hahn_disco.py, which executes the unchanged hahn.py source with only the instance name swapped (same pattern as the phoenix direct-ZZ run).
Why: the open instance reaches three processors; the claim "stated T2 differs from delivered" must be read on every processor we can reach.
Stated T2 median age at planning (h): boston 19.6, pittsburgh 6.8, phoenix 21.3, fez 0.9.
Reading per chip: hahn.py --analyze, the same rule (ABOVE / BELOW / CONSISTENT / CANNOT_TELL by copy gap), with the stated T2 median age printed.
Bill rule (same as PLAN_HAHN_FRESH.md): a chip's line goes on the bill only if it reads BELOW while its stated T2 median age at firing is
under 3 h, and every run of that chip combines (chi2/dof < 2) BELOW. Readings on older records go on the site, all runs side by side.
fez here is its second run (run 1: 1.051 [1.003, 1.085] ABOVE, stated T2 0.0 h old); combined with run 1 by the same rule.
KNOWN-GOOD: median P(1) at tau = 0 at least 0.90 per chip. Synthetic test: synth_hahn.text (unchanged).