optivlsi 0.1.2

High-performance VLSI CAD algorithms with Numba-accelerated implementations

optiVLSI

optiVLSI is a Python package that implements a collection of classic graph and circuit algorithms, including:

• Bellman‑Ford
• Dijkstra
• Prim
• Kruskal
• Lee (maze solver)
• BDD (binary decision diagrams)
• Simulation engines (compiled‑code and event‑driven)

The project has been refactored into a proper Python package with a modern pyproject.toml, type hints, comprehensive tests, documentation, and CI/CD pipelines.

Quick Start

# Install the package
pip install optivlsi

# Run a quick demo
python -m optivlsi.lee.algorithms.lee_algorithm

Documentation

• API Reference
• Algorithms Overview
• Benchmarks

Contributing

See the CONTRIBUTING.md file for guidelines.

Implemented Algorithms

Graph Algorithms

Algorithm	Package	Variants
Bellman-Ford Shortest Path	optivlsi.bellman_ford	Pythonic, NetworkX, Numba
Dijkstra Shortest Path	optivlsi.dijkstra	Pythonic, NetworkX, Numba
Kruskal Minimum Spanning Tree	optivlsi.kruskal	Pythonic (DSU), NetworkX, Numba
Prim Minimum Spanning Tree	optivlsi.prim	Pythonic, NetworkX, Numba

Routing

Algorithm	Package	Variants
Lee Maze Routing	optivlsi.lee	Pythonic BFS, NetworkX, Numba

Digital Circuit Simulation

Algorithm	Package	Variants
Compiled-Code Simulator	optivlsi.simulation.compiled_code	Gate classes, Numba
Event-Driven Simulator	optivlsi.simulation.event_driven	Event propagation, Numba

Binary Decision Diagrams

Algorithm	Package	Variants
ROBDD	optivlsi.bdd	Python, Numba

Running Tests

# Run all tests
pytest

# Run with coverage
pytest --cov=optivlsi --cov-report=term

# Run benchmarks
pytest tests/test_benchmarks.py --benchmark-only

Documentation

Full documentation is available in the docs/ directory:

• API Reference
• Algorithm Details

Detailed research paper: OptiVLSI.pdf

Benchmarking

Each algorithm module includes an automate.py file for automan-based benchmarking across various problem sizes. The package also provides pytest-benchmark integration for performance regression detection.

A live benchmark comparison report is automatically generated on every push:
👉 View Benchmark Report

Optimization Tools Used

• Numba: All algorithms have Numba-accelerated variants with JIT compilation
• Automan: Automated simulation and benchmarking infrastructure
• NetworkX: Reference implementations using standard graph library

Project Structure

optivlsi/                   # Main package
├── bellman_ford/           # Bellman-Ford algorithm
├── dijkstra/               # Dijkstra's algorithm
├── kruskal/                # Kruskal's MST
├── prim/                   # Prim's MST
├── lee/                    # Lee maze routing
├── simulation/
│   ├── compiled_code/      # Compiled-code simulator
│   ├── compiled_code_numba/# Numba-accelerated variant
│   └── event_driven/       # Event-driven simulator
├── bdd/                    # ROBDD
└── utils/                  # Shared utilities
bellman-ford/               # Original standalone modules
dijkstra/                   # (preserved for reproducibility)
kruskal/                    #
prim/                       #
lee-algorithm/              #
compiled-code-simulator/    #
event-driven-sim/           #
ROBDD/                      #

Original Research

This open-source codebase started as a course project for AE6102 - Parallel Scientific Computing and Visualization at IIT Bombay. The original standalone research modules are preserved in their respective directories.

Collaborators

• Rohan Rajesh Kalbag
• Neeraj Prabhu

License

MIT