The hype surrounding Cloud Computing has grown exponentially in recent months, but what is it?  Is it the biggest breakthrough since the integrated circuit, or is this just another flash in the technology pan?

What is it?

Could Computing refers to dynamically scalable computer resources which can be drawn upon to perform tasks as required. The cloud metaphor refers to the abstract and dynamic nature of the physical computer hardware infrastructure in place.

A History

The first example of Cloud Computing to be used on a massive scale was the SETI@home project. This project was publicly launched on May 17, 1999 and currently draws on the computer resources of over 3 million volunteers. When each volunteer's computer rests idle, it is treated as part of a vast cloud of computer resources. This cloud is then called upon to analyze large volumes of radio telescope data. Any promising results are highlighted and returned to SETI (Search for Extra-Terrestrial Intelligence) to be verified.

Even without an army of volunteers, the benefits of this Cloud Computing concept are immediately evident. By abstracting or disconnecting an application from actual computer running it, the application can be scaled up, or scaled down as the application's requirements change over time.

The heavy hitters of the internet realized long ago that this type of scalability was essential to their success. As such many the most popular applications running today are hosted by multiple computers across multiple locations. For example, Twitter, YouTube, Facebook, Google Maps, Microsoft Hotmail, Amazon.com all use scalable computer resources to ensure that they can meet the demands of their users.

Today

So, if Cloud Computing has been in use for such a long time, why are we hearing so much about it today? The answer lies in the implementation. While this type of scalability is great, actually implementing it can be hugely complex and expensive, requiring large risky up front capital investments.

New developments have surfaced in recent years from providers who have entered this space and begun offering up their Cloud Computing infrastructures for rent by the hour, Gigabyte or CPU cycle. This trend is breaking down barriers, and allowing more and more companies to leverage the benefits of Cloud Computing.

Benefits

Think of the "Cloud" in Cloud Computing as a large pool of shared computers that can be drawn from, as needed, to support a web application. This large pool of computers must exist at all times, whether you use them or not.  The significant development with the recent Cloud Computing service offerings is that you only pay for the cloud when you use it.

As consumers of these Cloud Computing services, organizations can leverage the reserve computing capacity of an IT elephant, without paying for it. In addition to heavy computing requirements, high bandwidth usage applications can leverage file replication services and reduced bandwidth rates, by hosting their files "in the Cloud".

By leveraging these Cloud Computing services new applications can be rapidly scaled up as demands increase, and should computing needs be overestimated Cloud Computing resources can instantly be scaled back, slashing sunk costs and eliminating this as a source of project risk.

Limitations

Cloud Computing is not a magic bullet, and some applications will not benefit from Cloud Computing. For example small applications which can easily run on a single server with low bandwidth requirements will not see cost benefits from Cloud Computing services. See: Cloud Computing Costs for Small Applications

Another barrier slowing down the adoption of Cloud Computing is the fact that existing applications will need to be modified or re-written to leverage the benefits of Cloud Computing. The potential benefits and modifications are best analyzed and performed by talented people with Cloud Computing expertise, but these people can be difficult to find.

Cloud Computing is not a giant leap forward, but rather the natural evolution of computers and the internet.  The benefits of adopting this new technology can be great, but as with any business decision, careful analysis of both the costs and benefits should be performed prior to adoption.

Key Cloud Computing Service Offerings

Amazon S3 (Simple Storage Service) which was launched in March 2006 allows users to serve up fast replicated content from Amazon's servers for low bandwidth charges. Part of Amazon's lineup of Amazon Web Services (AWS)

Google App Engine launched in April 2008 allows developers to upload Python or Java applications to Google servers for execution. This allows scalable application execution without knowledge of the underlying server.

Amazon EC2 (Elastic Compute Cloud) became generally available on October 23, 2008. This service offers Linux and Windows based virtual servers, for rent by the hour without setup charges. The user is provided with full root/administrative rights to what appears to be a distinct server. The server can be configured through remote desktop administration as though it was a physical server. Part of Amazon's lineup of Amazon Web Services (AWS)

Windows Azure is currently in beta, but is expected to be released by the end of this year. The Microsoft publicity machine is generating significant amounts of the current Cloud Computing buzz. Windows Azure will be similar in features and price to the existing Amazon EC2 service. This service should be publicly available by the end of 2009 as part of Microsoft's Microsoft Azure lineup.

Microsoft announced accelerated development of SQL Azure Database in May 2009, but no release date has yet been scheduled. This service will provide scalable relational database solutions, presumably with full MS-SQL feature support. Part of Microsoft's Microsoft Azure lineup.