CUDA Sessions

This repository contains CUDA learning materials and example kernels for getting started with NVIDIA CUDA programming. It provides a progression from basic CUDA concepts to optimized implementations.

Repository Structure

• cuda-intro/: Introductory CUDA examples covering basic kernels, memory management, and simple algorithms.
• matmul-basic/: matrix multiplication implementations demonstrating optimization techniques like tiling.

Prerequisites

• NVIDIA GPU with CUDA support
• NVCC compiler installed (part of CUDA Toolkit)

How to Compile and Run CUDA Files

Compile a CUDA file using NVCC:

nvcc filename.cu -o outputfile
./outputfile

For example:

nvcc SimpleMatrixMultiplication.cu -o SimpleMatrixMultiplication
./SimpleMatrixMultiplication