Optimizing Dot Product Calculation for Large Matrices: A Comparison of Two Approaches