Opting to follow the computing-design philosophy that the best way to reduce power consumption and increase energy efficiency is to reduce waste,we propose an architecture with a very simple ready-implementation by using an NComputing device that can allow multi-users but only one computer is needed.This intuitively can save energy,space as well as cost.In this paper,we propose a simple and realistic NComputing architecture to study the energy and power-efficient consumption of desktop computer systems by using the NComputing device.We also propose new approaches to estimate the reliability of k-out-of-n systems based on the delta method.The k-out-of-n system consisting of n subsystems works if and only if at least k-of-the-n subsystems work.More specificly,we develop approaches to obtain the reliability estimation for the k-out-of-n systems which is composed of n independent and identically distributed subsystems where each subsystem(or energy-efficient usage application)can be assumed to follow a twoparameter exponential lifetime distribution function.The detailed derivations of reliability estimation of k-out-of-n systems based on the biased-corrected estimator,known as delta method,the uniformly minimum variance unbiased estimate(UMVUE)and maximum likelihood estimate(MLE)are discussed.An energy-management NComputing application is discussed to illustrate the reliability results in terms of the energy consumption usages of a computer system with quad-core,8GB of RAM,and a GeForce 9800GX-2graphics card to perform various complex applications.The estimated reliability values of systems based on the UMVUE and the delta method differ only slightly.Often the UMVUE of reliability for a complex system is a lot more difficult to obtain,if not impossible.The delta method seems to be a simple and better approach to obtain the reliability estimation of complex systems.The results of this study also show that,in practice,the NComputing architecture improves both energy cost saving and energy efficient living spaces.