CUDA and PyTorch Operations

This repository contains CUDA and PyTorch-based implementations for matrix operations and performance comparisons. It includes:

1. CUDA Matrix Multiplication (C++)

• Implements matrix multiplication using CUDA cores.
• Compares GPU performance with CPU performance.
• Demonstrates the use of CUDA kernels for parallel computation.

2. PyTorch Matrix Multiplication (Python)

• Compares matrix multiplication using CUDA cores (FP32) and Tensor cores (FP16).
• Highlights the performance benefits of Tensor cores for half-precision computations.
• Includes result validation to ensure accuracy between FP32 and FP16 computations.

---

Requirements

For the CUDA C++ Code:

• NVIDIA GPU with CUDA support.
• CUDA Toolkit installed.
• C++ compiler (e.g., nvcc).

For the PyTorch Python Code:

• PyTorch installed with CUDA support.
• NVIDIA GPU with Tensor Core support (for FP16 acceleration).

---

How to Run

1. CUDA C++ Code:

• Compile and run using nvcc:

  nvcc -o MatMul MatMul.cu
  ./MatMul

2. PyTorch Python Code:

• Run the Python script:

  python matmul_pytorch.py

---

Features

• Demonstrates GPU acceleration for matrix operations.
• Compares CPU and GPU performance for matrix multiplication.
• Highlights the use of Tensor cores for FP16 computations in PyTorch.

---

Notes

• Ensure your system has the required hardware and software for CUDA and PyTorch.
• The code includes result validation to ensure correctness of GPU computations.