CUDA

Practical GPU Computing & CUDA Projects

A collection of practical, ready-to-run CUDA projects focused on GPU computing, parallel programming, performance acceleration, and real-world computational workloads.

---

Projects

#	Project	Focus
01	Vector Addition	Parallel vector computation
02	Vector Multiplication	Element-wise GPU operations
03	Matrix Addition	2D GPU computation
04	Matrix Multiplication	Parallel matrix computation
05	Parallel Reduction	Parallel aggregation
06	Image Blur	GPU image processing
07	Edge Detection	Computer vision acceleration
08	Convolution	GPU convolution
09	Matrix Transpose	GPU memory optimization
10	CPU vs GPU Benchmark	Performance comparison
11	Device Query	GPU device information
12	GPU Memory Bandwidth	GPU memory performance
13	Parallel Histogram	Parallel histogram computation
14	Image Histogram	GPU image histogram
15	Monte Carlo Simulation	GPU-based simulation
16	cuBLAS Matrix Multiplication	GPU linear algebra
17	FFT Processing	Frequency-domain processing
18	GPU Random Numbers	GPU random number generation
19	CUDA Streams	Asynchronous GPU execution
20	CUDA Graphs	Graph-based execution
21	Unified Memory	Unified memory management
22	Pinned Memory	Fast host-device transfers
23	Sparse Matrix Operations	Sparse linear algebra
24	Linear System Solver	GPU linear system solving
25	GPU Image Processing Pipeline	Multi-stage GPU image processing

---

Requirements

• NVIDIA GPU with CUDA support
• NVIDIA GPU Driver
• CUDA Toolkit
• nvcc compiler
• C++ compiler

Check your CUDA installation:

nvcc --version

Check your NVIDIA GPU:

nvidia-smi

---

Getting Started

Clone the repository:

git clone https://github.com/hasheramin5-cyber/CUDA.git
cd CUDA

Navigate to a project:

cd "01. Vector Addition"

Compile:

nvcc VectorAddition.cu -o VectorAddition

Run on Windows:

.\VectorAddition.exe

Run on Linux:

./VectorAddition

Each project contains its own README.md with project-specific requirements, compilation instructions, usage, and expected output.

---

Focus Areas

• CUDA Programming
• GPU Computing
• Parallel Computing
• GPU Acceleration
• CUDA C/C++
• Computer Vision
• Performance Optimization
• High-Performance Computing

---

Repository Structure

CUDA/
│
├── Assets/
│   └── cuda-logo.svg
│
├── 01. Vector Addition/
├── 02. Vector Multiplication/
├── 03. Matrix Addition/
├── 04. Matrix Multiplication/
├── 05. Parallel Reduction/
├── 06. Image Blur/
├── 07. Edge Detection/
├── 08. Convolution/
├── 09. Matrix Transpose/
├── 10. CPU vs GPU Benchmark/
├── 11. Device Query/
├── 12. GPU Memory Bandwidth/
├── 13. Parallel Histogram/
├── 14. Image Histogram/
├── 15. Monte Carlo Simulation/
├── 16. cuBLAS Matrix Multiplication/
├── 17. FFT Processing/
├── 18. GPU Random Numbers/
├── 19. CUDA Streams/
├── 20. CUDA Graphs/
├── 21. Unified Memory/
├── 22. Pinned Memory/
├── 23. Sparse Matrix Operations/
├── 24. Linear System Solver/
├── 25. GPU Image Processing Pipeline/
│
├── .github/
│   ├── workflows/
│   │   └── ci.yml
│   ├── ISSUE_TEMPLATE/
│   │   ├── bug_report.md
│   │   ├── cuda_error.md
│   │   └── feature_request.md
│   └── pull_request_template.md
│
├── .gitignore
├── LICENSE
├── CONTRIBUTING.md
├── CODE_OF_CONDUCT.md
└── README.md

---

About

This repository focuses on practical CUDA implementations designed to demonstrate how NVIDIA GPUs can accelerate computationally intensive workloads through parallel execution.

The projects are kept independent, reproducible, and focused on real GPU-computing use cases.

---

License

This repository is licensed under the MIT License.