This paper use deep reinforcement learning to complish placement optimization. It is opportunity and challenge to physical design engineer. So I will do it from a interpreter.