Vectorized Image Filter (C++ - SIMD Accelerated)

This project demonstrates the application of a 3x3 blur filter using both scalar (classic loop) and SIMD (vectorized with AVX2) techniques on grayscale-converted images.

🧠 Purpose

The aim is to compare traditional loop-based filtering with vectorized processing using AVX2 intrinsics to showcase the performance benefits of SIMD in system-level software development.

⚙️ Technologies Used

• C++17
• GCC with -mavx -mavx2 flags
• AVX2 intrinsics via <immintrin.h>
• PPM image format (raw RGB)
• Linux command line (make, g++)

📁 Project Structure