🧠 Pi-HPC-Lab

High-Performance Computing on Raspberry Pi Cluster
A lightweight distributed computing lab for mathematics, physics, and AI simulation — built entirely on Raspberry Pi devices.

---

🌍 Overview

Pi-HPC-Lab is a collection of distributed computation experiments designed for small ARM clusters
(such as Raspberry Pi 5 or Jetson Orin Nano).
It demonstrates how fundamental mathematical, physical, and optimization problems can be solved in parallel across multiple nodes — using pure Python and lightweight frameworks (e.g., sockets, Dask, NumPy).

This project aims to:

• Build a mini supercomputer environment with low-cost hardware.
• Study parallel algorithms for mathematical & physical simulations.
• Provide reusable code templates for distributed computing education and experimentation.

---

⚙️ Cluster Setup

Hardware Example

Role	Device	Notes
Master	Raspberry Pi 5 (8GB)	Runs Dask Scheduler or custom socket server
Worker 1–3	Raspberry Pi 5 / Jetson Nano	Execute compute tasks
Network	Gigabit Ethernet Switch	Optional PoE power
Storage	NVMe SSD / USB3 External	For caching & logs

Software Stack

• Python 3.10+
• NumPy, Dask, Matplotlib
• SSH access between nodes
• Optional: Ray, MPI4py, or FastAPI for extended experiments

Install Dependencies

sudo apt update && sudo apt install -y python3-pip
pip install numpy dask distributed matplotlib

🧩 Project Structure

pi-hpc-lab/
├── matmul/              # Distributed matrix multiplication
├── linear_regression/   # Distributed gradient descent
├── heat_equation/       # PDE simulation (2D heat diffusion)
├── montecarlo_pi/       # Monte Carlo π estimation
├── fft_image/           # Distributed FFT and image filtering
├── benchmark/           # Performance comparison utilities
└── README.md