This study proposes the use of low-quantization techniques, hexagonal sampling, and implicit convolution to enable lighter and more memory-efficient diffraction calculations. The combination of these methods reduces memory usage and speeds up diffraction calculations.