Abstract: In the paper, we propose a virtual machine (VM) consolidation and placement strategy, named segmentation iteration correlation combination (SICC). The SICC algorithm integrates several algorithms, such as time series analysis, linear correlation analysis and traditional first fit decreasing (FFD). A new dynamic resource consolidation theory is then established based on the virtual machine minimum resource utilization parameter. The simulation results indicate that the novel SICC framework can improve the physical resource utilization by 3% to 20% in the VM granularity and by up 5% in the server granularity.