gf180-gate-driver (2AMLogic) errors
3 failures reported against gf180-gate-driver (2AMLogic), sorted by how they fail and ranked by impact.
#180 by errors trackedSimulation & verification
Recurring gf180-gate-driver (2AMLogic) errors, by impact
- 1Same repo, same klt bump: the deck-clean arm of the ERC negative control now reports 1 finding with no local changes. The script's own message says the checker or deck drifted, not the layout, so the block's 'erc_status: clean' isn't evidence until it's explained.Simulation & verification · Wrong verdict · 1 report66
- 2The gate_driver_core.gds was regenerated on Oct 24, but the committed extracted and RC-extracted netlists were last committed Aug 26, so they describe the old layout. The README already admits the post-layout PVT records were taken on earlier GDS revisions, and nothing covers the oxide-safety post-layout bench on the current GDS, whose newest record is from Sep 10.Simulation & verification · Silent failure · 1 report66
- 3The signoff says re-running the generator reproduced the committed GDS exactly, but under klt 0.9.2 it doesn't: 76% of the common prefix differs, with the first diff a cell name (res_array$3 vs x3_Rfb__core). Same netlist, same generator, only the tool version moved, and a klt upgrade renamed sub-cells and broke a byte-for-byte reproducibility attestation.Simulation & verification · Wrong verdict · 1 report66
How it fails
Sourced from public community reports. Costs and delays on individual entries are modeled estimates, not measurements.