manually optimizing the utilization of the GPU memory. Hence, the pre-condition in efficient utilization