SI/PI local evidence and remaining release blockers

Status: draft_local_evidence Release use: prohibited_until_external_review

This report records what can be proven from local repository artifacts as of 2026-05-18. It does not claim vendor, board-house, package-vendor, or foundry approval.

Local evidence inputs

• package/e1-demo-pinout.yaml: placeholder package pinout.
• docs/board/kicad/e1-demo/fab-notes.md: KiCad planning notes that mark the board package as non-release demo planning evidence.
• pd/openlane/runs/RUN_2026-05-18_05-41-42/signoff-run.yaml: selected local OpenLane run record.
• pd/openlane/runs/RUN_2026-05-18_05-41-42/final/metrics.json: routed digital metrics for local context.

Findings

The local tree has enough information to define SI/PI work items, but not enough to close SI/PI release. Package electrical models, released board stackup, impedance targets, and board-level simulation outputs are absent.

Clock, reset, debug, GPIO, IRQ, and JTAG nets therefore remain unsimulated at the board/package boundary. Power integrity also remains open because the repository does not contain rail impedance targets, decoupling impedance analysis, return-path review, or first-article measurements.

Release blockers

• Archive a package model: IBIS, SPICE, S-parameter, or extracted parasitics.
• Release board stackup and impedance rules.
• Run or review board SI for clock, reset, debug, GPIO, IRQ, and JTAG nets.
• Archive PI impedance, decoupling, return-current, and plane-continuity review.
• Cross-reference the released package model and board stackup from pd/signoff/manifest.yaml.