Repository files navigation

🚀 A High-Performance FPGA Placement and Routing Framework Powered byPyTorch

State-of-the-Art  | GPU-Accelerated  | Production-Ready

• Overview
• Key Features
• Latest Updates
• Visual Demo
• Getting Started
  • Prerequisites
  • Installation
  • Running Benchmarks
• Documentation
• Resources
• Team & Contributors
• Publications
• License

OpenPARF is a cutting-edge FPGA placement and routing framework that delivers superior performance through innovative algorithms and GPU acceleration. Our framework achieves up to 12.7% improvement in routed wirelength and 2X+ speedup in placement compared to state-of-the-art solutions.

• 🔋Advanced Placement Technology

  • Multi-Electrostatic-based placement engine
  • Multi-die FPGA support with SLL count optimization
  • Comprehensive unit support (LUT, FF, DSP, BRAM, IO, etc.)

• ⚡High Performance

  • GPU acceleration via CUDA
  • 0.4-12.7% better routed wirelength
  • 2X+ faster placement speed

• 🛠Production Ready

  • Industry-standard architecture support
  • Extensive benchmarking suite
  • Well-documented API and examples

🎉OpenPARF 2.0 Release (2024/02/02)

• Multi-die FPGA placement support
• Flexshelf architecture format
• Performance improvements and bug fixes
• Release Notes

Experience the power of our electrostatic field-based placement through these visualizations from benchmarkISPD2016/FPGA06:

LUT	FF	DSP	BRAM

• Python 3.7+
• C++14 compatible compiler
• PyTorch 1.7.1
• CUDA 11.0 (for GPU acceleration)
• Gurobi 9.5 (optional)

# Create conda environmentmamba create --name openparf python=3.7mamba activate openparf# Install dependenciesmamba install cmake boost bisonmamba install pytorch==1.7.1 torchvision==0.8.2 cudatoolkit=11.0 -c pytorchpip install hummingbird-ml pyyaml networkx tqdm# Build OpenPARFgit clone --recursive https://github.com/PKU-IDEA/OpenPARF.gitmkdir build&&cd buildcmake .. -DCMAKE_PREFIX_PATH=$CONDA_PREFIX -DPYTHON_EXECUTABLE=$(which python) -DCMAKE_INSTALL_PREFIX=../install -DPython3_EXECUTABLE=$(which python)make -j8&& make install

For detailed installation instructions, seeINSTALL.md.

Download the benchmarks according toBENCHMARKS.md, then run the following command:

cd ../installpython openparf.py --config unittest/regression/ispd2016/FPGA01.json

For more benchmark information, seeBENCHMARKS.md.

• 📘Installation Guide
• 📗Benchmark Guide
• 📙Architecture Customization

• ISPD 2016 Contest
• ISPD 2017 Contest

OpenPARF is developed and maintained byPKU-IDEA Lab at Peking University, under the supervision ofProf. Yibo Lin.

Core Contributors:

• Jing Mai,Jiarui Wang andYibai Meng - Initial release
• Runzhe Tao - Multi-die placement integration
• Jing Mai,Jiarui Wang, Yifan Chen,Zizheng Guo, Xun Jiang,Yun Liang,Yibo Lin - OpenPARF 3.0 with macro placement

For academic papers and citations, please see ourpublications list.

OpenPARF is released under the BSD 3-Clause License. SeeLICENSE for details.