Optimization of Power and Computational Density of a Data Center

Abstract

Techniques for optimizing power and computational density of data centers are described. According to various embodiments, a benchmark test is performed by a computer data center system. Thereafter, transaction information and power consumption information associated with the performance of the benchmark test are accessed. A service efficiency metric value is then generated based on the transaction information and the power consumption information, the service efficiency metric value indicating a number of transactions executed via the computer data center system during a specific time period per unit of power consumed in executing the transactions during the specific time period. The generated service efficiency metric value is then compared to a target threshold value. Thereafter, a performance summary report indicating the generated service efficiency metric value, and indicating a result of the comparison of the generated service efficiency metric value to the target value, is generated.

1 . A method implemented by at least one computing device, the method comprising:
 determining a benchmark value of an efficiency metric for a hardware component executing as part of a data center;   generating a target value of the service efficiency metric for the hardware component;   determining a difference between the benchmark value of the service efficiency metric and the target value of the service efficiency metric; and   causing output of a notification based on the difference between the benchmark value with the target value.   
     
     
         2 . The method of  claim 1 , wherein the notification includes a recommendation for improving the service efficiency metric. 
     
     
         3 . The method of  claim 1 , wherein the service efficiency metric comprises power consumption data for the hardware component. 
     
     
         4 . The method of  claim 1 , wherein the target value of the service efficiency metric comprises a maximum amount of power to be consumed by the hardware component. 
     
     
         5 . The method of  claim 1 , wherein causing output of the notification comprises transmitting the notification to a client device. 
     
     
         6 . The method of  claim 1 , wherein determining the benchmark value of the efficiency metric for the hardware component is performed based on historical power consumption information for the hardware component. 
     
     
         7 . The method of  claim 1 , wherein the data center comprises at least one server device and the hardware component comprises a processer of the at least one server device. 
     
     
         8 . A method implemented by at least one computing device, the method comprising:
 determining, for each of a plurality of hardware components executing as part of a data center, a benchmark value of an efficiency metric for the hardware component;   ascertaining, for each of the plurality of hardware components, a target value of the service efficiency metric for the hardware component;   computing a difference between the benchmark value of the service efficiency metric and the target value of the service efficiency metric for each of the plurality of hardware components;   generating a report that summarizes the differences between the benchmark values of the service efficiency metric and the target values of the service efficiency metric for the plurality of hardware components; and   causing output of the report at a display of a client device.   
     
     
         9 . The method of  claim 8 , wherein the service efficiency metric comprises power consumption data for the plurality of hardware components. 
     
     
         10 . The method of  claim 8 , wherein the service efficiency metric comprises temperature data for the plurality of hardware components. 
     
     
         11 . The method of  claim 8 , wherein the service efficiency metric comprises a number of users accessing the data center during a time period. 
     
     
         12 . The method of  claim 8 , wherein the service efficiency metric comprises a cost incurred from operating the plurality of hardware components during a time period. 
     
     
         13 . The method of  claim 8 , wherein the plurality of hardware components comprise a plurality of data servers of the data center. 
     
     
         14 . A system comprising:
 one or more processors; and   a computer-readable storage medium having instructions stored thereon that are executable by the one or more processors to perform operations comprising:
 determining a benchmark value of an efficiency metric for a hardware component executing as part of a data center; 
 generating a target value of the service efficiency metric for the hardware component; 
 determining a difference between the benchmark value of the service efficiency metric and the target value of the service efficiency metric; and 
 causing output of a notification based on the difference between the benchmark value with the target value. 
   
     
     
         15 . The system of  claim 14 , wherein the notification includes a recommendation for improving the service efficiency metric. 
     
     
         16 . The system of  claim 14 , wherein the service efficiency metric comprises power consumption data for the hardware component. 
     
     
         17 . The system of  claim 14 , wherein the target value of the service efficiency metric comprises a maximum amount of power to be consumed by the hardware component. 
     
     
         18 . The system of  claim 14 , wherein causing output of the notification comprises transmitting the notification to a client device. 
     
     
         19 . The system of  claim 14 , wherein determining the benchmark value of the efficiency metric for the hardware component is performed based on historical power consumption information for the hardware component. 
     
     
         20 . The system of  claim 14 , wherein the data center comprises at least one server device and the hardware component comprises a processer of the at least one server device.