optimize-ppa

What problem does it solve?

This Skill helps engineers wring out higher utilization, faster Fmax, and cleaner DRC-free flows from HighTide benchmark builds without ever editing upstream RTL by focusing on stats-driven flow, timing, IO, and PDN tuning.

Core Features & Use Cases

• Baseline Measurement: Capture die area, utilization, power, timing, and runtime from bazel-bin logs or legacy logs and reference config.mk, constraint.sdc, BUILD.bazel, io.tcl, and pdn.tcl files before changing anything.
• Utilization Optimization: Incrementally raise CORE_UTILIZATION and PLACE_DENSITY, mitigate congestion with IO pin scripts, macro halos, pin spacing, and ABC area controls, and watch for runtime blowup or routing failure heatmaps.
• Timing and Power Tightening: Read timing reports, adjust clock periods, IO delays, and uncertainty settings, rerun flows, monitor CTS/routing stages, and optionally tighten power through IR-drop-aware PDN scripts and ABC_AREA settings.
• Flow Diagnostics: Generate placement, routing, RUDY, and IR drop images via Docker+OpenROAD, compare metrics across runs in tables, and iterate until additional changes trigger unacceptable congestion or negative margins.

Quick Start

Ask the skill to review the last build metrics for designs/asap7/lfsr, adjust config.mk/constraint.sdc/io.tcl/pdn.tcl to increase utilization and tighten timing, then rerun the flow and compare the new die area, Fmax, and DRC status against the baseline.