The simplest definition for Cloud Computing is the one in which data and applications are served to users over the Internet. Users no longer need applications such as Outlook on their desktop, but instead use versions of these, or other applications, through a web browser. Data is, similarly, stored centrally and served up to users as and when they need it, and on any platform they choose.

Originally, Cloud Computing was an indistinct term for a vague and distant future in which computing would occur in a few remote locations without the need for much human intervention. Infinite computing resources would be available for any need at costs approaching zero. Certainly, users would not need to know or care about how the computers, their software, or the network functioned.

Cloud Computing customers do not generally own the physical infrastructure serving as host to the software platform in question. Instead, they avoid capital expenditure by renting usage from a third-party provider. They consume resources as a service and pay only for resources that they use. Many Cloud Computing offerings employ the utility computing model, which is analogous to how traditional utility services (such as electricity) are consumed, while others bill on a subscription basis.

Sharing “perishable and intangible” computing power among multiple tenants can improve utilization rates, as servers are not unnecessarily left idle (which can reduce costs significantly while increasing the speed of application development). A side effect of this approach is that overall computer usage rises dramatically.

We cycled between periods when computing was more centralized (and seemed more remote and less accessible to users) and periods when computing was right on user desktops. No one was ever satisfied. Centralized computing failed to give users enough control and was too inflexible. Distributed computing made every user his own system administrator and was very inefficient.

In the last few years, as the cost of a unit of computing power has continued to decrease – but the cost of humans with the skills to implement and manage computer systems has not – the vision of centralized computing has returned. It has taken several turns. Some computer scientists have suggested (and experimented with) a vast Grid of computers, attached via the Internet, whose power can be combined for large-scale tasks when needed. In some cases, very large computing systems can be part of these grids for specialized tasks. Others have suggested a computing utility which would provide just as much computing power as an organization needed, on an on-demand basis, much like electricity.

Eventually, as large web users such as Google and Amazon built out enormous data centers for their own purposes, they realized they could permit others to access these “clouds” of computing power at relatively attractive prices, and so began the Cloud Computing era. Today, many companies are putting together large data centers, sometimes as extensions of their own needs, sometimes just for customers to use. Originally the idea was these clouds of computing would offer processing power and storage. Anything else would be added by the customer. As the idea became more popular, additional functionality has been added. Some clouds also offer systems management. Others are actually providing a set of applications as part of the cloud.

For most organizations, a decision on Cloud Computing will be a matter of choosing which cloud to use. Many may use several, selecting different clouds for different purposes. For some large enterprises and government organizations, a cloud of their own may be an appropriate solution. This allows the organization to optimize the cloud for its own purposes and to make it available to its own constituency.

*Did You Know — According to a recent Unisys poll, security and privacy concerns are still big barriers to Cloud Computing?  A survey asked, “What do you see as your greatest barrier to moving to the cloud?”, and 51 percent cited security and data privacy. 21 percent cited integration of cloud applications with existing systems as a potential barrier.