Result
jtl
jtl_out_of_envelope
jtl
Refused
This design sits outside the box the heads were measured on. The refusal names the field, the value and the range, so it is repairable.
The same cell at 3x its nominal bias. Outside the box the heads were measured on, so it returns no number and says which field, what value, and what range was allowed.
sfq_outside_envelope
bias_ua=450.0 is outside the trained range [64.0, 266.0] for cell 'jtl'; the surrogate has no data there and would be extrapolating
6 quantities refused
the surrogate's margin labels for this cell were produced on a testbench whose input is an ideal one-flux-quantum voltage source; it forces the phase advance regardless of bias, so the labels describe the stimulus, not the device. Measured under a current-driven input the bands are much narrower: jtl bias 0.8-1.2x under the reference drive, against a labelled 0.2-2.0x. The width is feeder-dependent -- across nine defensible feeders it ranged 0.30-0.65 (2.17x) -- so any published margin must name its feeder and carry that spread. Refused rather than served.
bias_low · bias_high · bias_width · ic_low · ic_high · ic_width
fluxus-s · captured from the same service the API serves · reproduces the JoSIM oracle, not silicon; nothing here is fab-calibrated. The cell library targets MIT-LL SFQ5ee+ and is literature-reconstructed, not foundry GDS. Margins are served for 3 of 8 library cells.