Microelectronic Design, Emulation, Fabrication & Validation

Microelectronic Design, Emulation, Fabrication & Validation

Executive Summary

This facility supports an end-to-end microelectronic workflow—from IC/RF front-end design and pre-silicon emulation, through fabrication handoff, to post-silicon validation and correlation.

Tool pillars

  • EM & high-frequency verification: ANSYS Electronics Desktop / HFSS
  • Custom IC schematic-to-layout: Cadence Virtuoso (PDK-based)
  • System/channel realism for wireless experiments: Remcom Wireless InSite

New project? Start with Quick Start, then follow the Recommended end-to-end workflow.

Navigation

Capabilities

Design

  • Analog / mixed-signal / RF IC schematic design and simulation
  • RF/microwave modeling (antennas, passives, packages, interconnects)
  • Co-design of circuits + EM structures + system constraints

Emulation

  • Pre-silicon exploration using behavioral/compact models
  • Parameter sweeps and sensitivity studies to de-risk tapeout
  • “What-if” studies on corners, parasitics, and layout-dependent effects

Fabrication handoff

  • Layout readiness checks (DRC/LVS/PEX workflows via PDK)
  • Tapeout package preparation (GDS + documentation)
  • Coordination for MPW/shuttle or foundry pathways (availability depends on project and partner access)

Validation

  • Post-silicon comparison: measured vs simulated
  • S-parameter correlation, de-embedding planning, and model updates
  • Reproducible reporting for publications and future tapeouts

Quick Start

  1. Pick your entry point

  2. Build a minimal, reviewable baseline (aim for 1–2 days)

    • One schematic or EM model that reproduces a known reference
    • One plot that becomes your golden regression (S-parameters, gain, NF, phase noise, etc.)
  3. Decide your validation target early

    • What will be measured? what fixtures? what calibration/de-embedding approach?

Tool map

ToolBest forTypical outputs
ANSYS AEDT / HFSS3D EM simulation of antennas, passives, packages, interconnectsS-parameters, radiation patterns, fields, loss/Q, EM co-sim models
Cadence VirtuosoIC design from schematic → layout → verification → sign-offSchematics, simulations, layout, DRC/LVS/PEX reports, GDS
Remcom Wireless InSiteSite-specific channel realism via ray-tracing / empirical modelsCoverage maps, channel impulse response, path loss, multipath statistics

ANSYS Electronics Desktop & HFSS

Best for: EM simulation and validation of RF/microwave components, antennas, and high-frequency structures using 3D FEM workflows.

Self-paced resources

Structured / premium resources

Research-oriented learning path

  1. Install AEDT (student version where appropriate).
  2. Complete the Intro HFSS course.
  3. Reproduce at least one reference antenna/passive example.
  4. Modify: substrate, port type, mesh/convergence settings.
  5. Capture convergence evidence and solver settings.
  6. Add parametric sweeps and HPC workflows as needed.

Practical tips

  • Ports/boundaries + mesh/convergence dominate result quality—treat them as first-class design artifacts.
  • Save convergence plots and solver settings for paper-quality reproducibility.
  • Consider related AEDT tools (Q3D, SIwave, Icepak) when SI/PI/thermal coupling matters.

Cadence Virtuoso

Best for: custom IC design (analog, mixed-signal, RF) from schematic capture to layout, verification, and simulation.

Official training

University and open tutorials

Academic/community access

From zero to tapeout-ready

  1. Reproduce a “hello world” design (e.g., inverter/op-amp) in a standard PDK.
  2. Learn simulation flows: DC/AC/transient, corners, and basic Monte Carlo.
  3. Transition to layout and run DRC/LVS.
  4. Add PEX, then compare pre/post-layout results.
  5. Automate repetitive tasks with SKILL only after the manual flow is stable.

Practical tips

  • Define sign-off checks early (DRC/LVS/PEX/corners/Monte Carlo) and keep them consistent.
  • Maintain a versioned tapeout checklist (schematic, layout, verification reports, notes).

Remcom Wireless InSite

Best for: RF propagation modeling via 3D ray-tracing and empirical models to predict coverage, channel characteristics, and system performance in realistic environments. We often use it for testbed planning and RIS-assisted communications studies.

Official tutorials and videos

Documentation and support

Channel realism learning path

  1. Run the indoor tutorial to understand the ray-tracing workflow.
  2. Build/import a simple environment and assign materials.
  3. Generate coverage maps and channel outputs; sanity-check with simplified baselines.
  4. Add mobility/time variation if your experiment needs it.
  5. Export results for MATLAB/Python post-processing (plots, statistics, comparisons).

Practical tips

  • Treat geometry + material definitions as the model—document them like a circuit schematic.
  • Record frequency bands, antenna patterns, and material properties for reproducibility.

Recommended end-to-end workflow

  1. Concept + requirements
    • Target frequency, bandwidth, power, sensitivity, interfaces, and measurement plan
  2. Pre-silicon modeling / emulation
    • Behavioral models → schematic simulations → early EM checks
  3. Layout + verification
    • DRC/LVS/PEX + correlation against schematic intent
  4. EM co-simulation (as needed)
    • Packages/interconnects/antennas/passives + parasitics
  5. Fabrication handoff (project-dependent)
    • GDS + documentation + sign-off checklist + versioned deliverables
  6. Post-silicon validation
    • Calibration/de-embedding + measured vs simulated correlation
  7. Documentation
    • Store tool versions, scripts, and “golden” plots to ensure publishable, repeatable results

Checklists & best practices

Tapeout-readiness (typical)

  • ✅ DRC clean + reports archived
  • ✅ LVS clean + connectivity assumptions documented
  • ✅ PEX completed + pre/post-layout deltas reviewed
  • ✅ Corner coverage defined (PVT, mismatch/Monte Carlo as appropriate)
  • ✅ Foundry deliverables packaged (GDS + runsets + README + version tags)

Measurement-correlation (typical)

  • ✅ Fixture and calibration plan (SOLT/TRL/etc.) selected before the first measurement run
  • ✅ De-embedding approach defined (structures, reference planes, uncertainty notes)
  • ✅ Measured vs simulated plots use the same reference planes and same conditions
  • ✅ Model updates are traced to measurements (revision history + rationale)

Reproducibility basics

  • Keep a lab notebook + version control for projects, scripts, and plots.
  • Save tool versions, PDK versions, and solver settings with every “golden” plot.

Notes for students and new lab members

  • Start small: one reproducible result beats ten half-working models.
  • Ask for help with: tool version, screenshots, the smallest reproducing case, and expected vs observed behavior.
  • Treat setup details (ports, boundaries, runsets, fixtures) as part of the design—not an afterthought.

Access & support

Who this is for

  • Students and researchers doing IC/RF design, EM validation, and post-silicon correlation
  • Projects that need a documented path from simulation assumptions → fabrication deliverables → measured results

When requesting help, include

  • Tool + version (and PDK version if using Virtuoso)
  • The smallest reproducing case (project archive or screenshot series)
  • What you expected vs what you observed
  • Any “golden” plot you are trying to match (and how it was generated)

Recommended project folder skeleton

  • 00_requirements/ (spec, interfaces, measurement plan)
  • 10_models/ (behavioral/compact models)
  • 20_schematic/ (schematic + sims)
  • 30_layout/ (layout + DRC/LVS/PEX reports)
  • 40_em/ (HFSS projects, ports/boundaries notes, convergence evidence)
  • 50_fab/ (GDS, runsets, README, sign-off checklist)
  • 60_measurement/ (fixtures, calibration notes, raw data)
  • 70_correlation/ (measured vs simulated plots, model updates)

Collaboration norms

  • Assume every result should be reproducible by someone else in 6 months.
  • Treat runsets, ports/boundaries, and calibration steps as design artifacts.
  • If a project is headed toward tapeout, plan a review gate for: spec freeze → pre-layout sign-off → post-layout sign-off → handoff package.