The cloud offers many IT resources such as low-cost and flexibility of products across the Internet. Cloud became the most demanding infrastructure as many cloud providers seek greater business outcomes and environment. Cloud became more difficult as a result. It amply demonstrates that the intelligent cloud's new beginning is currently taking place. Optimizing the cost-effective cloud service configuration and resource allocation adaptability are just a few drawbacks that the cloud is currently experiencing. We can say that using machine learning to boost cloud management intelligence is becoming more and more popular. This article focuses on the architecture of deep reinforcement learning-based intelligent cloud resource management. Cloud will work efficiently and automatically negotiate the best configuration, starting from complex cloud environments by using Deep Reinforcement Learning. At last, We provide a concrete example to illustrate the remarkable power of deep reinforcement learning in the intelligent cloud.