A benchmark comparing CPU vs GPU performance using $1024 \times 1024$ Matrix Multiplication ($O(N^3)$ complexity).

This project demonstrates the honest benefit of GPGPU. Unlike synthetic tests with artificial loops, Matrix Multiplication is a foundational operation in AI and Physics that naturally benefits from massive parallelism.

• Task: Calculate $C = A \times B$ for $1024 \times 1024$ matrices.
• Complexity: Requires approx. 2 Billion floating-point operations.
• CPU Implementation: Standard naive triple-loop ($O(N^3)$).
• GPU Implementation: OpenGL Compute Shader (4.6) dispatching 1 thread per element.

• CPU Time: ~1850ms
• GPU Time: ~100ms (includes memory upload/download overhead)
• Speedup: ~18x

• Scoop (Windows Package Manager)
• Visual Studio (C++ Compiler)
• OpenGL 4.3+ capable GPU

1. Install Dependencies

   scoop install cmake vcpkg vcpkg integrate install vcpkg install glfw3 glad --triplet=x64-windows

2. Configure & Build

   cmake -B build -S . -DCMAKE_TOOLCHAIN_FILE="$env:USERPROFILE/scoop/apps/vcpkg/current/scripts/buildsystems/vcpkg.cmake" cmake --build build --config Release

3. Run

   .\build\Release\gpgpu_mm.exe