Metric Evaluation

Performance-metric evaluation from a solved sizing.

Small-signal parameters come through a DeviceModel evaluated at the solved geometry, so the same code is exact for both Level-1 (geometry-free λ·Id) and gm/Id (LUT). Shared by both sizers.

circuitgenome.sizer.shared.metrics._evaluate_metrics(transistor_sizing, slot_transistors, cc_pf, tech, spec, model, cc2_pf=None, gd_load_r=0.0, rout1_override=None, rout2_override=None, rout3_override=None, gm1_factor=1.0, gd_tail_override=None, gd_out_extra=0.0)

Compute performance metrics and safety margins from the solution.

Small-signal parameters come through model evaluated at the solved geometry — exact for both Level-1 (geometry-free λ·Id) and gm/Id (LUT).

rout{1,2,3}_override let a caller (the gm/Id pipeline) supply cascode-aware stage output resistances; when None (the Level-1 default) the single-device-gds estimate is used unchanged.

Parameters:

• transistor_sizing (dict[str, TransistorSizing])

• slot_transistors (dict[str, list[Device]])

• cc_pf (float | None)

• tech (TechParams)

• spec (SizingSpec)

• model (DeviceModel)

• cc2_pf (float | None)

• gd_load_r (float)

• rout1_override (float | None)

• rout2_override (float | None)

• rout3_override (float | None)

• gm1_factor (float)

• gd_tail_override (float | None)

• gd_out_extra (float)

Return type:

tuple[dict[str, float], dict[str, float]]