In recent years, the considerable growth rate in the utilization of computing devices increases the voluminous data. It necessitates the transferring of the data from the hardware to virtualized infrastructure for the computation such as cloud computing technology, termed as Software-Defined Computing (SDC). Instead of computing the functions in a specific hardware device, the computations of the software-defined functions happen in any number of devices in the cloud. In addition to the networking, the software-defined process is applicable for both storage and computation. The SDC is a new paradigm, that manages the complexities integrated with the cloud computing system through the various software-defined cloud components. By incorporating an abstraction layer comprising both physical and virtual resources, the software-defined architecture assists the user to perform dynamic computation. It alleviates the complexities in the cloud infrastructure including minimization of manual assembly and configuration of physical resources. Moreover, it performs auto-scaling of resources based on the user requirement. In SDC, the service provider provides the Application Programming Interface (API) to control the resource configuration. The Software-Defined Computing provides numerous opportunities for its development, which are as follows: • In SDC, it is crucial to make the decision regarding the execution of mobile applications. Moreover, optimization of the programming model for mobile applications helps in partitioning the applications efficiently. • The development of SDC needs to focus on self-management, self-healing, self-optimization. • The SDC often meets the incurred overhead during the workload variation, which leads the controller to obtain negative performance. • Addressing the response latency in the SDC is crucial for ensuring computing optimization. • To avoid the interaction in the monolithic model, SDC is decomposed into some interacting sub-models resolving the scalability problem. However, the SDC decomposition introduces the cyclic dependency, which requires the mathematical theories are needed to solve the cyclic dependency, while minimizing the errors introduced by the sub-models interaction.