PARAGON Documentation

Parallel Graph Engine for High Performance Computing

Overview

PARAGON is a high performance graph processing engine built using a hybrid C++ + Python architecture. The core engine is implemented in C++ with parallel execution capabilities, enabling efficient processing of large scale graphs using multicore CPUs. Python bindings powered by pybind11 provide a clean and user friendly interface, making PARAGON both powerful and accessible.

The project is designed with a focus on performance, scalability, and modularity. It leverages thread-level parallelism, efficient memory usage, and optimized graph traversal techniques to accelerate computation-heavy workloads. PARAGON serves as a foundation for building advanced graph analytics systems and research-oriented applications.

Currently, PARAGON supports a range of essential graph algorithms, including:

  • Breadth-First Search (BFS)

  • Depth-First Search (DFS)

  • PageRank (push and pull variants)

  • Connected Components

  • Single Source Shortest Path (SSSP)

  • Triangle Counting

With a growing ecosystem of features, PARAGON aims to evolve into a complete parallel graph processing framework.

Getting Started

To begin using PARAGON, follow the installation instructions and run your first graph program:

Contents: