Method and apparatus for dynamically allocating power in a data center

A non-transitory computer-readable storage medium having computer executable instructions stored thereon, the instructions when executed cause a computer to perform a method, the method comprising:

• measuring power consumption of a computer system having one or more servers;
• estimating probability distribution of power demand for each of the one or more servers, the estimation based on the measured power consumption;
• estimating performance loss via the estimated probability distribution;
• computing power capping limits for each of the one or more servers, the computation based on the estimated probability distribution and the performance loss, wherein the power capping limits indicate power consumption limits of the one or more servers;
• and
• dynamically allocating the power capping limits to each of the one or more servers,
• wherein computing power capping limits is based on a difference between a first performing loss of a first power capping limit and a second performing loss of a second power capping limit, the second power capping limit is set by increasing the first capping limit by a power unit.

An apparatus comprising:

• a server rack having one or more servers;
• a processor coupled with the server rack via a network interface, the processor having logic to perform a method comprising:
• measuring power consumption of the server rack;
• estimating probability distribution of power demand for each of the one or more servers, the estimation based on the measured power consumption;
• estimating performance loss via the estimated probability distribution;
• computing power capping limits for each of the one or more servers, the computation based on the estimated probability distribution and the performance loss, wherein the power capping limits indicate power consumption limits of the one or more servers;
• and
• dynamically allocating the power capping limits to each of the one or more servers of the server rack,
• wherein computing power capping limits is based on a difference between a first performing loss of a first power capping limit and a second performing loss of a second power capping limit, the second power capping limit is set by increasing the first capping limit by a power unit.