matrix_mul

Different implementations of matrix multiplication optimized for x86 machines using AVX-256 instructions.

Prerequisites

• CMake (version 3.10 or higher)
• A C++ compiler with AVX2 support
• Windows: Visual Studio with C++ workload or MinGW-w64
• Linux/MacOS: GCC or Clang

Building the Project

1. Create a build directory:

mkdir build
cd build

2. Configure with CMake:

cmake ..

3. Build the project:

• On Windows with Visual Studio:

cmake --build . --config Release

• On Linux/MacOS:

make

The executable will be created in:

• Windows: build/bin/Release/dgemm_exe.exe
• Linux/MacOS: build/dgemm_exe

Running the Program

Simply run the executable:

• Windows: .\bin\Release\dgemm_exe.exe
• Linux/MacOS: ./dgemm_exe

The program will perform matrix multiplication performance tests using different implementations and display the results.

Features

• Optimized matrix multiplication using AVX-256 instructions
• Performance comparison between different implementations
• Automatic matrix generation for testing
• Performance metrics calculation

Technical Details

• Uses C++11 standard
• Implements AVX2 and FMA instructions for optimal performance
• Includes both debug and release configurations
• Employs Link Time Optimization (LTO) when supported