IHP GDSFactory PDK

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#

Category	Device	`cells/` (pure GDSFactory)	`cells2/` (PyCell)
FET	nmos	`nmos`	`nmos`
	pmos	`pmos`	`pmos`
	nmos_hv	`nmos_hv`	`nmosHV`
	pmos_hv	`pmos_hv`	`pmosHV`
RF FET	rfnmos	`rfnmos`	`rfnmos`
	rfpmos	`rfpmos`	`rfpmos`
	rfnmos_hv	`rfnmos_hv`	`rfnmosHV`
	rfpmos_hv	`rfpmos_hv`	`rfpmosHV`
Bipolar	npn13G2	`npn13G2`	`npn13G2`
	npn13G2L	`npn13G2L`	`npn13G2L`
	npn13G2V	`npn13G2V`	`npn13G2V`
	pnpMPA	`pnpMPA`	`pnpMPA`
Resistor	rsil (silicided poly)	`rsil`	`rsil`
	rppd (p-poly)	`rppd`	`rppd`
	rhigh (high-R)	`rhigh`	`rhigh`
Capacitor	cmim (MIM)	`cmim`	`cmim`
	rfcmim (RF MIM)	`rfcmim`	`rfcmim`
	cmom (MOM)	`cmom`	–
Inductor	inductor2	`inductor2`	`inductor2`
	inductor3	`inductor3`	`inductor3`
Passive	svaricap (MOS varicap)	`svaricap`	`svaricap`
	ESD protection	`esd_nmos`	`esd`
	ntap1 / ptap1	`ntap1` / `ptap1`	`ntap1` / `ptap1`
	guard_ring	`guard_ring`	–
	sealring	`sealring`	`sealring`
Diode	diodevdd 2kV/4kV	`diodevdd_2kv` / `diodevdd_4kv`	–
	diodevss 2kV/4kV	`diodevss_2kv` / `diodevss_4kv`	–
	schottky_nbl1	`schottky_nbl1`	–
Antenna	dantenna / dpantenna	`dantenna` / `dpantenna`	`dantenna` / `dpantenna`
Bondpad	bondpad	`bondpad`	`bondpad`

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
│   ├── primitives.py       #   Schematic-only VLSIR elements (R, L, C, sources)
│   ├── 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 and a VLSIR bridge (to_vlsir.py) for SPICE netlist export.
primitives.py provides schematic-only elements (ideal R, L, C, voltage/current sources) with VLSIR metadata for circuit simulation — these have no physical GDS geometry.

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#

License#

Apache 2.0