GraphBuilding

Minimal viable example for online graph building on FPGAs for machine learning trigger applications in particle physics.

Prerequisites

This example is optimized for Ubuntu. To run the Graph Building Example the following install the following tools:

Python 3.9:

apt-get install python3.9 python3.9-venv

sbt (Simple Build Tool):

apt-get install -yq curl default-jdk
echo "deb https://repo.scala-sbt.org/scalasbt/debian all main" | sudo tee /etc/apt/sources.list.d/sbt.list
echo "deb https://repo.scala-sbt.org/scalasbt/debian /" | sudo tee /etc/apt/sources.list.d/sbt_old.list
curl -sL "https://keyserver.ubuntu.com/pks/lookup?op=get&search=0x2EE0EA64E40A89B84B2DF73499E82A75642AC823" | sudo apt-key add
apt-get update
apt-get install -yq sbt

Verilator:

apt-get install -yq \
verilator git perl make autoconf g++ flex bison ccache \
libgoogle-perftools-dev numactl perl-doc \
libfl2 libfl-dev zlib1g zlib1g-dev

Install FIRRTL Compiler:

wget -q -O - https://github.com/llvm/circt/releases/download/firtool-1.38.0/firrtl-bin-ubuntu-20.04.tar.gz | tar -zx \
    && sudo mv firtool-1.38.0/bin/firtool /usr/local/bin/

Vivado:

Please install Vivado 2022.1 according to the Xilinx Website

Setup

Run the setup.sh script to generate a new virtual environment and check if all required applications are available.

Running the Example

Configure the environment before running the example

export PYTHONPATH=$PWD/tool/src:$PYTHONPATH
export VIVADO=<path_to_vivado>
source ./graphbuilding_venv/bin/activate

Run the example by calling

python3 ./example.py

Folder Structure

• backend: Contains the currently supported synthesis backend script for vivado
• generators: Includes the current Hardware Generator
• samples: Input files for all three stage described in the example
• tool: Python Code for Generation of Intermediate Graph Representation