Abstract: A method for power management of a network including a plurality of computing devices, comprising: identifying an order of flows between one or more computing devices within a prescribed time period; and reducing a number of wakeups for at least one computing device by increasing a number of adjacent flows that share a computing device.