The invention relates to a management method for a hardware performance and a cloud computing system. The cloud computing system comprises a plurality of node devices, and a management node for realizing a management method, wherein the node devices are configured in a plurality of node resource pools. The management method for the hardware performance comprises the following steps of detecting loads of the node resource pools to judge bottlenecks and corresponding bottleneck resource pools with the bottlenecks, evaluating and selecting at least one switching node from the node devices of the other node resource pools than the bottleneck resource pools, and changing the switching node to distribute the switching node into the bottleneck resource pool from the original node resource pool.