cumatfuse 0.1.0

A high-performance CUDA library for fused matrix multiplication with ReLU

🚀 cuMatfuse

cuMatfuse is a minimal CUDA library for fused matrix multiplication with ReLU activation: C = ReLU(A × B)

✨ Features

• CUDA-accelerated fused GEMM with ReLU
• Lightweight and portable
• Easy to extend and integrate into ML inference pipelines

🛠 Requirements

• NVIDIA GPU (Compute Capability 7.0+)
• CUDA Toolkit ≥ 11.8
• CMake ≥ 3.18
• Linux or Windows

🔧 Build Instructions

git clone https://github.com/divakar-yadav/cuMatfuse.git
cd cuMatfuse
mkdir build && cd build
cmake ..
make

🧪 Run Example

./cuMatfuse_demo

Expected Output:

C = ReLU(A × B):
22 28
49 64
76 100
6 8

📦 License

MIT