Start with the tightest deadline

• How many weeks before your planned tapeout did the place-and-route flow start missing timing on a critical block in your most recent project? Options: 0-2 weeks, 3-4 weeks, 5-8 weeks, More than 8 weeks
• Tell me about the full P&R flow you run for representative blocks, from floorplan through signoff, including where you insert manual fixes or custom scripts.
• Which categories best describe the primary tools and custom layers in your current flow? Options: Sequential optimization P&R, Concurrent multi-objective P&R, Custom scripts built around incumbent tool, In-house placement and routing engines, Hybrid flow with manual handoffs
• Who on your team owns first-response triage when timing first regresses, and what role does each person play? Options: Cad manager, Block lead, Timing lead, Physical design engineers, Dedicated closure squad, Other
• Estimate how often a representative congested block requires more than two manual ECO iterations to reach signoff. Options: Almost always, Often, Occasionally, Rarely, Never

Where the tapeout actually stalls

• If a single congested partition adds four weeks, what part of your flow most amplifies that delay? Options: Placement quality under congestion, CTSI and clock tree adjustments, Routing closure and congestion hot spots, Manual ECO propagation and constraints tuning, Signoff regressions and DRC rework
• Walk me through the last block that forced a tapeout slip or near-slip, including the block type, size, switching activity, and where timing failed.
• Which specific block characteristics tend to correlate with repeat failures in your experience? Options: High pin density, High switching activity, Heavy macros or memories, Clock domain crossings, Severe local congestion
• Who notices the issue first, and how long does it take from first detection to the start of a focused fix campaign? Options: Design lead within 24 hours, Timing engineer within 48 hours, CAD manager within a week, Only during signoff runs, Other
• Name the block or partition that you would immediately delay tapeout for if it missed timing.

Where manual work eats your schedule

• When manual ECOs are your path to closure, what predictable costs are you accepting? Options: Extra engineering headcount, Extended tapeout schedule, Higher risk of regressions, Increased signoff iterations, All of the above
• How many person-weeks did manual ECOs and routing rework consume in your last tapeout? Options: 0-2 person-weeks, 3-5 person-weeks, 6-10 person-weeks, More than 10 person-weeks
• Describe the common manual steps your engineers take during an ECO cycle, and which steps are hardest to automate.
• Identify the team owning manual fixes and estimate the percent of their time consumed during crunch periods. Options: Physical design team 0-10%, Physical design team 10-25%, Physical design team 25-50%, Dedicated closure team 50%+, Shared across teams
• If you could reduce manual ECO iterations by half, what would that save you in schedule or cost, in concrete terms?

Hidden assumptions that break the run

• What permission, data, or compute gaps would stop an evaluation before it starts?
• List the specific design artifacts we would need access to within two weeks to run a representative evaluation, and who controls each artifact. Options: Gate-level netlist, SDC and constraints, Floorplan and macro placements, RC/extraction decks, Custom scripts and tool wrappers
• Identify any build or environment dependencies that require coordination across your IT or network teams.
• Estimate available bare-metal or cloud compute in core-hours the buyer can commit to the pilot, per week. Options: <1,000 core-hours, 1,000-5,000 core-hours, 5,000-20,000 core-hours, >20,000 core-hours
• Would you pause or cancel the evaluation if legal or data-authority approval to share constraint files cannot be obtained within 14 days? Options: Proceed with reduced dataset, Pause until approval, Cancel evaluation, Unsure

Other paths you're weighing

• What would have to be true for you to keep the current P&R approach rather than change course?
• Select from the following alternatives you are evaluating or have evaluated recently. Options: Tune incumbent P&R tool and scripts, Build internal optimizer or automation, Evaluate another vendor tool, Use external closure services, Delay tapeout to invest engineering time, Other
• Has anyone on your team proposed solving closure internally without a vendor, and if so, what resources and timeline did they estimate? Options: Yes, <3 months, Yes, 3-6 months, Yes, 6-12 months, No internal proposal, Plan exists but no estimate
• Name a measurable condition in your incumbent flow that would keep you from switching this cycle.
• List the internal stakeholders who must approve a vendor-led pilot, and note typical objections. Options: Physical design director, CAD manager, IT/security, Procurement, Legal, Architecture/SoC lead

Define measurable success

• Assuming the evaluation meets your targets, what would make you sign within one week?
• What single acceptance metric would you use to decide go no-go for migration after the pilot? Options: Block-level timing margin in ps, Percent reduction in ECO iterations, Full-run wallclock time, DRC violations threshold, Power or area delta limit
• Specify your target timing margin at block-level that constitutes acceptable closure, in picoseconds, for representative blocks. Options: <5 ps, 5-10 ps, 10-20 ps, >20 ps
• Choose the maximum runtime per full P&R iteration you will accept during evaluation. Options: Less than 12 hours, 12 to 24 hours, 24 to 48 hours, 48 to 72 hours, More than 72 hours
• State the maximum routing-related DRC violations acceptable post-optimization before you consider a run clean. Options: 0, 1-5, 6-20, More than 20
• Describe acceptable power or area tradeoffs and how you expect them reported in the evaluation deliverable.

Decision mechanics and next steps

• Suppose the pilot delivers promised convergence in two to three iterations, what internal process would accelerate adoption? Options: Fast-track procurement, Pilot-to-production migration plan, Executive signoff after pilot, Layered rollout by partition
• Provide the names or roles that must sign the purchase and the timeline procurement and legal need to approve after acceptance.
• Outline the top three migration responsibilities you expect to retain, and the top three you expect the seller to own.
• In the event the pilot proves results but integrating your custom scripts requires more effort than planned, would you pause migration or accept a staged rollout? Options: Pause migration and reassess, Proceed with staged rollout, Require seller to deliver full integration, Unsure
• How soon could your team start a pilot, given a signed NDA and data authorization? Options: Immediately, Within 2 weeks, 2 to 6 weeks, Later than 6 weeks
• What single unresolved technical or organizational risk would stop you from starting a pilot within your target window?

• Solution Experience Session — Place-and-Route Convergence

• Confirm the current state and its cost
• You confirm the described current state and accept the quantified schedule and engineering cost as accurate.
• Run a candidate P&R benchmark on one buyer-provided representative block and deliver a QoR comparison report (timing, iterations, runtime, power, area, DRC) before the next session.
• Map failure modes to the representative scenario
• You confirm the demonstrated workflow materially reduces iterations and manual ECOs on the representative scenario.
• Provide the representative block netlist, constraint files, current run logs, and the list of failure modes to reproduce.
• Walk through the seller's concurrent optimization on a representative run
• Confirm the acceptance metrics (timing slack targets, maximum acceptable runtime, allowed manual ECO hours, and DRC tolerance) and sign off on the benchmark schedule.
• You and the seller agree on the benchmark scope, acceptance metrics, required artifacts, and the schedule for the follow-up run.
• Prepare the execution checklist and mapping of compute and environment requirements needed to run the benchmark on the provided data.
• Compare outcomes and translate to schedule impact
• Validate this matches your needs
• Agree next evaluation scope and data handoff

• Solution Experience Session — Place-and-Route Convergence
• Solution Experience Deck
• Solution Brief

• meeting
• slides
• document

• PDK and technology-file integration
• Port physical-constraint scripts to new flow
• Floorplan and macro/block placement
• Global placement with congestion-aware optimization
• Detailed placement and legalization
• Clock tree synthesis and skew optimization
• Concurrent timing–power–routability optimization runs
• Signoff-quality routing and shielding
• DRC/LVS closure and fix generation
• Post-route extraction and STA signoff
• Automated ECO insertion and iterative fixes
• Runtime and convergence tuning for target frequency

PDK and technology-file integration

• Which PDK version and technology node (for example 3nm / 5nm / 7nm) must be integrated for this engagement? Options: 3nm, 5nm, 7nm, Other
• Provide the list of technology decks you will supply (LEF, tech LEF, .lef macro files, liberty .lib, DRC deck, LVS deck, parasitic extraction rules)
• Who will host the read-access for the PDK and where (internal artifact repository, secure FTP, vendor portal)? Options: Internal artifact repository, Secure FTP, Vendor portal, Other
• Attach or state any required license servers or feature flags needed to enable the process-specific models (for example, parasitic extraction-enabled license)
• Indicate any custom process rule exceptions or proprietary layer mappings (for example custom via stacks, reserved routing layers) that differ from the base PDK
• Specify which deliverable we should produce to confirm integration (LEF/DEF parse report, .lib timing sanity report, DRC parse log) Options: LEF/DEF parse report, .lib timing sanity report, DRC parse log, All of the above
• List any out-of-scope PDK activities we should exclude from the fixed work (for example full-cell characterization, transistor-level library re-characterization)

Port physical-constraint scripts to new flow

• Which constraint artifacts do you currently use that must be ported (SDC timing constraints, IO placement constraints, floorplan constraint files, power intent CPF/UPF)? Options: SDC timing constraints, IO placement constraints, Floorplan constraint files, CPF/UPF power intent, Other
• List the custom SDC constructs or scripts the team relies on (false path patterns, multi-cycle path definitions, multi-corner definitions)
• Who on your team owns constraint validation and what contact should we use for SDC clarifications?
• Estimate how many constraint files and total lines of SDC we should expect to convert Options: <1k lines, 1k-5k lines, 5k-20k lines, >20k lines
• Indicate whether your power intent is in CPF, UPF, or a mix and whether retention or power-switch designs require special handling Options: CPF, UPF, Mixed, No power intent provided
• Provide example SDC or power-intent checks you require after porting (for example: preserved false paths, consistent clock group definitions)
• Which test harness or regression harness should we use to validate translated constraints (internal regression, provided testbench, platform CI)? Options: Internal regression, Provided testbench, Platform CI, Other

Floorplan and macro/block placement

• Which tapeout-critical blocks or macro names should be included in the scoped floorplan (provide block names or GDS grouping identifiers)?
• Specify the target die size, target core utilization, and any preallocated macro keep-out regions from the GDS or floorplan DEF
• Who will provide the initial floorplan sources (existing DEF, placement constraints, floorplan script) and in what path or repository?
• Indicate whether hard macros require re-placement, legalizing, or only boundary checks against the new LEF Options: Re-placement required, Legalize only, Boundary checks only
• List any thermal, power-grid, or IO adjacency constraints that must be enforced during macro placement
• Provide the expected deliverable for this module (annotated DEF, updated floorplan script, placement constraint diff) Options: Annotated DEF, Updated floorplan script, Placement constraint diff, All of the above
• Estimate the expected number of hard macros and soft blocks in scope for placement validation Options: <10 macros, 10-50 macros, 50-200 macros, >200 macros

Global placement with congestion-aware optimization

• Which representative blocks should be run through global placement to measure congestion behavior (provide block IDs or netlist names)?
• Describe any congestion metrics you currently track that we should report (e.g., routing demand per micron, overflow histogram)
• Identify whether you require layer-aware congestion targets for specific metals or via stacks Options: Yes, layer-aware targets, No, global targets only, Unsure — consult required
• Which placement constraints (pin access rules, keepouts, fence regions) must be preserved during global placement?
• Specify the acceptable routing resource utilization threshold after placement (for example maximum channel utilization percent) Options: <60%, 60-80%, 80-95%, >95%
• Who is the owner for signoff of placement results in your team and what artifact do they expect (placement DEF, congestion heatmap)?

Detailed placement and legalization

• Which pin-access or cell-shift policies must be enforced during legalization (for example no cell rotation, keep IP macro pins fixed)? Options: No rotation, Pins fixed, Allow limited shifts, Custom policy provided
• Provide the set of standard-cell rows and row-spacing rules from the LEF that must be respected
• List any custom legalization checks you require (multi-height cell handling, IP site compatibility)
• Who will verify the post-legalization DRC-free run locally before routing starts?
• Estimate the number of cells and density (% utilization) the detailed placement flows will act on Options: Low (<40%), Medium (40-70%), High (>70%)
• Specify deliverables after legalization (legalized DEF, legalization log, cell movement report) Options: Legalized DEF, Legalization log, Cell movement report, All of the above
• Indicate any preferred handling for fixed macros that overlap pre-routed power stripes or retention cells

Clock tree synthesis and skew optimization

• Which clock domains and root names in your SDC should we synthesize CTS for (list clock net names and frequencies)?
• Provide target skew and insertion-delay budgets per domain (for example max skew 50ps, insertion <200ps)
• Identify whether deskew networks, localized buffer trees, or hierarchical CTS are required for any domain Options: Deskew networks, Localized buffer trees, Hierarchical CTS, No special handling
• Who owns the clock gating and clock-disable policies that must be preserved during CTS?
• List any constraints on routing or shielding for clock nets that must be enforced (for example metal layer restrictions, shielding distances)
• Specify the artifact you will accept as CTS deliverable (CTS DEF, clock netlist dump, clock-tree report) Options: CTS DEF, Clock netlist dump, Clock-tree report, All of the above

Concurrent timing–power–routability optimization runs

• Which representative scenarios and corners should we run concurrently (timing corners, PVT corners, IR drop cases)? Options: Setup corner, Hold corner, Worst-case PVT, IR/EM power case, All listed
• Provide the reference .lib views (max-transition, min-delay, etc.) that the multi-objective optimizer must target
• Identify constraints on power vs timing tradeoffs we should not exceed (for example max 5% power increase allowed for timing gain)
• Who will validate optimization strategy and approve changes to cell-sizing or gate-swapping recommendations?
• List the measurable QoR reports you require after each optimization run (timing report, power summary, congestion map, routability metric) Options: Timing report (STA), Power summary, Congestion map, Routability metric, All of the above
• Indicate whether you require automated rollback points and versioned DEF outputs after each optimization iteration Options: Yes, require versioned rollback, No, single final output, Consult required

Signoff-quality routing and shielding

• Which routing layers and via rules from the tech LEF must be used to meet signoff routing rules?
• Specify shielding requirements for sensitive nets (clock, reset, power) and acceptable metal layers for shield insertion
• List any special routing strategies required for high-speed I/O or SerDes lanes in scoped blocks
• Who will run signoff DRC/LVS on routed nets and which rule deck should be used for pre-checks?
• Provide the expected routing deliverables (routed DEF, routing report, via usage summary) Options: Routed DEF, Routing report, Via usage summary, All of the above
• Indicate whether incremental routing handoffs or full-route regeneration is preferred during iterative runs Options: Incremental handoff, Full-route regeneration, Either as needed

DRC/LVS closure and fix generation

• Which DRC and LVS rule decks (file names or deck versions) should we run for closure verification?
• Provide the maximum allowable DRC and LVS counts or categories that you will accept before escalating to manual fixes
• Indicate whether fixes should be automatic edit scripts, flagged fixes in a report, or a mixed approach Options: Automatic edit scripts, Flagged fixes report, Mixed approach
• Who will approve any layout edits that change macro boundaries or IP placement during DRC/LVS fixes?
• List any DRC categories that are informational-only and do not require hard closure for this evaluation
• Specify the format you require for the DRC/LVS acceptance artifact (annotated GDS diff, categorized error list, per-rule histogram) Options: Annotated GDS diff, Categorized error list, Per-rule histogram, All of the above

Post-route extraction and STA signoff

• Which parasitic extraction format do you require (SPEF, RSP, CCS) and which extraction deck version should be applied? Options: SPEF, RSP, CCS, Other
• Provide the STA corners and clock uncertainty values that the post-route STA must use for signoff

• decision_readiness
• stakeholders
• current_state
• desired_state
• success_criteria
• gaps
• decision_readiness
• gaps
• current_state
• desired_state
• stakeholders
• success_criteria
• decision_readiness
• stakeholders
• desired_state
• success_criteria
• current_state
• gaps
• decision_readiness
• decision_readiness
• decision_readiness
• decision_readiness

• Mutual Non-Disclosure Agreement (NDA)
• Order Form / Subscription Agreement
• Master Services Agreement (MSA)
• Statement of Work (SOW) — Evaluation & Migration
• Evaluation Acceptance Agreement
• Migration & Go/No-Go Commitment
• Data Access & Handling Authorization
• Data Processing Agreement (DPA)
• Service Level Agreement (SLA) for Support & Availability
• Change Order Agreement
• Payment Schedule & Invoice Terms

• Go-live Health Check (weeks 1-4)
• First Measurement, Early KPI Review (weeks 4-10)
• Operational Convergence Review (90-120 days post go-live)
• Quarterly Success Review (ongoing)

• Refresh training materials and schedule enablement sessions for any teams identified with proficiency gaps.
• Publish the incumbent decommission or archive plan with evidence of data migration and access changes.
• Aggregate outcome presentation across representative blocks
• Confirm timing closure rate and median iterations to convergence are at acceptable levels or have a funded remediation plan tied to Solution Evaluation targets.
• Agree remediation or escalation plans for all high-severity production issues with target resolution windows.
• Produce a short lessons-learned summary that will be used to update operational runbooks and onboarding materials.
• Publish the consolidated outcomes report with per-block trends and lessons learned.
• Create a prioritized enhancement request list with brief scope and acceptance criteria.
• Schedule targeted technical detailed review sessions for each unresolved high-severity item.
• Quarterly operational metrics and trend review
• Validate quarterly trends for average P&R runtime and open high-severity issues and confirm they are stable or improving versus Solution Evaluation targets.
• Agree the prioritized enhancement list for the next quarter and handoff of execution to the delivery backlog.
• Confirm owners and target dates for remaining high-severity issues to ensure continued progress.
• Update the shared dashboard with current quarter metrics and a brief narrative explaining any deviations from targets.
• Publish the next-quarter prioritized enhancement and support backlog with expected delivery windows.
• Re-confirm scope owners and success criteria custodian
• Confirm the deployment checklist is complete and evidence for each item is available in the shared workspace.
• Identify any critical blockers preventing production runs and agree remediation actions and target dates.
• Ensure every acceptance criterion in Solution Evaluation has an assigned owner for post-deployment tracking.
• Provide the completed deployment validation checklist with links to environment snapshots and run logs.
• Publish a short report of initial run outcomes and any immediate errors observed for async review.
• Create remediation tickets for any critical blockers with target resolution dates.
• Present first measurements vs Solution Evaluation targets
• Determine whether median negative slack and average P&R runtime per representative block are on track to meet Solution Evaluation targets or require remediation.
• Document root causes for the top gaps and a specific remediation plan with milestones and target dates.
• Confirm incumbent tool status is either fully decommissioned or retained read-only with an archive and a plan to eliminate silent fallback work.
• Deliver a per-block timing slack and runtime breakdown with traceable links to the runs used for analysis.
• Produce a remediation plan for the top three root causes with concrete parameter changes or process steps and target completion dates.
• Review manual ECO workload and iteration counts
• Enhancement request backlog review and prioritization
• Deployment and migration validation
• Root-cause diagnosis for any metric gaps
• Surface persistent production issues and escalations
• Early adoption signals and usage patterns
• Open issues burn-down and owner updates
• Incumbent system wind-down status
• Blockers, open issues, and temporary workarounds
• Agree corrective actions and timeline to stable operations
• Adoption and training needs check
• Capture lessons learned and update runbooks
• Agree immediate remediation actions
• Agree next operational adjustments and enhancement requests
• Meeting efficiency check