4 Improving the Efficiency of AI Applications Using In-Memory Computation