Sun Microsystems have defined the term Data Physics as the consideration of the relationship between the processing elements of an Information System and the data on which these processing elements operate.

The 'Clouds' in Cloud Computing can be divided into storage clouds for data storage and compute clouds which carry out the processing.  Storage clouds complement compute clouds. Since most compute clouds store data in the cloud rather than on a physical computer server it takes time to bring data to a server to be processed.

Sun Microsystems have created a simple equation for data physics. This equation describes how long it takes to move an amount of data from where it is generated, stored, processed and archived. As Sun point out, while Clouds are good at storing data they are not necessarily good at archiving and destroying data on a predefined schedule. In simple terms, large amounts of data or low bandwidth network connections lengthen the time it takes to move data. This can be expressed mathematically as time = (bytes*8) / bandwidth. 

In practical terms, this equation is relevant for both the moment-by-moment processing of data and for long term planning. For example, if the IT infrastructure needs increase for an organisation, that organisation can expand its cloud by temporarily 'renting' resources from a Cloud Provider. This strategy is known as surge computing. However, as this process entails moving data from one cloud to another it can ultimately entail taking more time than if expansion of the cloud did not take place. The data physics equation helps determine whether it makes sense to implement a surge computing strategy where it might take longer to move the data to a public cloud provided by a cloud vendor than it would to process the data within the current IT framework. Data physics can also help determine the cost of moving operations from one cloud provider to another. Whatever data has accumulated in one cloud provider's data centre must be move to another data centre. Such a process takes time.

The cost of moving data can be expressed both in terms of time and in terms of bandwidth charges. Data physics is a reminder to consider the relationship between data and processing and that moving data from storage to processing can cost both time and money. Data stored without computing power nearby has limited value.

Data physics has implications for smart objects and smart networks. The volume of data produced by smart networks will increase exponentially in the coming years and this data will need to be stored by organisations. Given the volume of devices and networks producing this data storage, clouds would appear to be a logical choice for many organisations as the cost of the IT infrastructure otherwise required would be prohibitive. One must factor in, however, that the data produced by smart networks needs to be processed and transformed into meaningful information. Essentially, this means that data that is received by a storage cloud from a smart network will then need to be transferred to a compute cloud. Sun's data physics equation can be used to compute the cost of processing this data within the cloud.

The processing of translating smart object data into meaningful information for an organisation can entail the use of data mining, business analytics and business intelligence techniques. Furthermore, this information will need to be loaded into other software and Information Systems such as Web Content Management Systems, Document Management Systems and Enterprise Resource Planning (ERP) Systems. It is the requirements of the latter two systems that is relevant here as there is currently debate as to whether ERP systems in particular should be maintained on premises or deployed on the cloud. If an organisation's smart object data resides on the cloud while their Enterprise-wide systems are installed on premises this means that data will have to be transferred from the smart network back to the enterprise.

The final issue to consider here relates to data capture. When data is captured it will need to be transferred to the cloud using an Internet or telecommunications connection. While the data may be sent to the storage cloud, it is also possible that the data will need to be transferred from a compute cloud to which the data has been sent to the storage cloud. When architecting an system for capturing and processing smart object data the data physics equation is therefore likely to be very valuable in decision making.

There are two pools of resources to consider here. Smart objects offer a rich pool of data that can be transformed into information for decision making while Cloud Computing offers a rich pool of resources that can enable organisations to process and store huge volumes of data that would otherwise be infeasible. Cloud Computing can act as a complementary enabler for smart object networks but architects and developers need to consider the cost of data transfer when designing a smart object system that is deployed on the cloud.