IHP GDSFactory PDK¶

A GDSFactory-based Process Design Kit for the IHP SG13G2 130nm BiCMOS open-source technology. It provides parametric layout cells, design rule constants, simulation models, and example designs for tape-out-ready integrated circuits.

Quick start¶

Use Python 3.11, 3.12 or 3.13. We recommend VSCode as an IDE.

uv pip install ihp-gdsfactory --upgrade

Then you need to restart Klayout to make sure the new technology installed appears and start generating IHP-SG13G2 GDSII immediately!

import gdsfactory as gf
from ihp import PDK
from ihp.cells import nmos, rfnmos, npn13G2, rsil, cmim

PDK.activate()

# Create a parametric NMOS transistor
c = nmos(width=1.0, length=0.13, nf=4)
c.write_gds("my_nmos.gds")
c.show()  # opens in KLayout

Available devices¶

cells/ functions are fully parametric pure-Python implementations. cells2/ functions wrap the original IHP PyCell library (requires CNI runtime). Entries marked -- are only available in one implementation.

Project structure¶

ihp/
├── cells/                  # Pure GDSFactory parametric layout cells
│   ├── fet_transistors.py  #   NMOS/PMOS (standard & HV)
│   ├── rf_transistors.py   #   RF-MOSFETs with guard/gate rings
│   ├── bjt_transistors.py   #   SiGe HBTs (npn13G2, npn13G2L, npn13G2V, pnpMPA)
│   ├── resistors.py        #   Polysilicon & metal resistors
│   ├── capacitors.py       #   MIM & MOS capacitors
│   ├── inductors.py        #   Spiral inductors
│   ├── passives.py         #   Diodes, varactors, guard rings
│   ├── via_stacks.py       #   Via stack generators
│   └── fixed.py            #   Pre-built GDS imports (SiGe HBTs, etc.)
├── cells2/                 # Legacy PyCell reference implementations (CNI-based)
├── models/                 # Compact models & SAX S-parameter models
├── gds/                    # Pre-built GDS files for complex cells
├── tech.py                 # Layer map, design rules, technology parameters
├── layers.yaml             # Layer definitions
└── config.py               # Paths and PDK configuration
tests/
├── test_cells.py           # GDS regression & settings tests for all cells
├── test_xor_transistors.py # Polygon-exact XOR tests (GDSFactory vs PyCell)
└── gds_ref/                # Golden reference GDS files
docs/                       # Jupyter Book documentation (Sphinx)

Key architectural concepts:

• cells/ contains pure GDSFactory implementations — each @gf.cell function builds geometry from scratch using add_polygon and design rule constants from tech.py. No external layout tools are needed at runtime.
• cells2/ holds the original IHP PyCell implementations (CNI/OpenAccess-based). These serve as the reference for XOR verification — every polygon in cells/ is tested to match cells2/ exactly.
• models/ provides SAX-compatible S-parameter models for circuit simulation.

Installation¶

We recommend uv

# On macOS and Linux.
curl -LsSf https://astral.sh/uv/install.sh | sh

# On Windows.
powershell -ExecutionPolicy ByPass -c "irm https://astral.sh/uv/install.ps1 | iex"

Installation for contributors¶

git clone https://github.com/gdsfactory/ihp.git
cd ihp
uv venv --python 3.12
uv sync --extra docs --extra dev
python install_tech.py

Running tests¶

make test              # run full test suite
make test-force        # run tests and regenerate reference GDS files

To run specific test subsets:

uv run pytest tests/test_cells.py -v          # GDS regression + settings tests
uv run pytest tests/test_xor_transistors.py -v # polygon-exact XOR vs PyCell

Documentation¶

• gdsfactory docs
• IHP docs from GDSFactory and code
• IHP documentation
• IHP component diagrams

License¶

Apache 2.0

Pre-commit¶

make pre-commit

Release¶

1. Bump the version:

   tbump 0.0.1
   

2. Push the tag:

   git push --tags
   

   This triggers the release workflow that builds wheels and uploads them.

3. Create a pull request with the updated changelog since last release.