meteharun: Implemented optimized matrix multiplication

[meteharun]

Summary

Implemented and benchmarked three matrix multiplication methods:

• Naive
• Blocked (cache-optimized, block_size = 32)
• Parallel (using OpenMP)

Optimizations

• Used tiling in blocked version to improve cache locality
• Used #pragma omp parallel for to parallelize row-wise computation in the parallel version
• Rounded float values to 2 decimals to ensure correct validation

Performance Results

Included detailed benchmark table in README.md.

Challenges

• For smaller matrix sizes, blocked matmul had overhead and performed worse
• Had to adjust file paths due to CLion running from the cmake-build-debug directory
• Minor issues with OpenMP scoping resolved with default(none) and shared(...)

All tests pass. Speedup for large matrices especially in parallel option is significant.