Definitions of cloud computing are easy to find, but a single, authoritative definition is hard to come by. Perhaps the best work in this area was done by Böhm, et al. By compiling characteristics of 17 different scholarly and industrial definitions, the authors identified five primary characteristics of cloud computing allowing a definition such as: "Cloud computing is a service that delivers scalable hardware and/or software solutions via the Internet or other network on a pay-per-usage basis." (Emphasis indicates essential definition elements).

Cloud computing can further be broken down into three common types: SaaS, PaaS, and IaaS. SaaS (Software as a Service) allows users to log into and utilize preprogrammed software that is owned and maintained by the service provider. PaaS (Platform as a Service) gives users tools and languages owned and maintained by the service provider that can be used to build and deploy customized applications. IaaS (Infrastructure as a Service) provides users with storage and processing, allowing users full control over the use of that infrastructure. There are other divisions of cloud computing, but these are the most common.

Conceptual Origins of Cloud Computing
Looking back, it seems that cloud computing was seen as the end goal of many computer pioneers in the 1960s, or, at least, the goal of the early experiments that would eventually become the Internet.

There are three main figures commonly cited as laying the conceptual framework for cloud computing: John McCarthy, JCR Licklider, and Douglas F. Parkhill.

McCarthy first proposed in 1957 that time sharing of computing resources might allow companies to sell excess computation services for maximum utilization of the resource. He even imagined that computation might be organized as a utility.

Licklider, a programmer at the Advanced Research Projects Agency, highlighted some of the promise and challenges in cloud computing in a 1963 memo to those he described as the "Members and Affiliates of the Intergalactic Computer Network." Specifically, he talked about the ability to send a problem to a network of computers that could then pool their resources to solve it, and the need to establish a shared language to allow the computers to talk to one another.

In 1966 Parkhill published "The Challenge of the Computer Utility," which identified many of the challenges facing cloud computing, such as scalability and the need for large bandwidth connections. He also initiated a comparison with electric utilities.

Why We Are in Cloud Computing Time
If cloud computing has been around for so long conceptually, why does it seem like a revolutionary idea at all? Because only now are we in cloud computing time.

Science fiction scholars commonly use the shorthand "steam engine time" to describe the phenomenon that ideas pop up several times but don't catch on for many years. They point out that the Romans knew what steam engines were and could make them, but it wasn't until 1600 years later that the technology came to fruition. The world just wasn't ready for steam engines. The same is true of cloud computing.

The necessary elements that had to be in place before cloud computing could become a reality were the presence of very large datacenters, high-speed Internet connectivity, and the acceptance of cloud computing as a viable model for supplying IT needs.

The presence of very large datacenters is a crucial piece in the foundation of cloud computing. To be able to offer cloud services at a competitive price, suppliers must have datacenters sufficiently large to take advantage of the economies of scale benefits that can reduce costs 80-86% over the medium-sized datacenters that many companies previously utilized. These very large datacenters were manufactured for their own use by many companies that would later become cloud computing providers, such as Amazon, Google, and Microsoft.

Almost universal access to high-speed Internet connectivity is crucial to cloud computing. If your data is bottlenecked getting to and from the cloud, it simply can't be a practical solution for your IT needs.

Finally, it is important for potential users to see cloud computing as a viable solution for IT needs. People need to be able to trust that some ethereal company is going to be able to provide for your urgent IT needs on a daily basis. This cultural work was done by many disparate influences, from MMOs to Google, which expanded acceptance of online resources beyond the IT community. Another crucial but oft-neglected part of this cultural work was performed by peer-to-peer computing, which introduced many people to the notion that they could utilize the resources of other computers via the Internet.

Cloud Computing Timeline: Who, When, and Why
There are many good timelines about cloud computing available, and several are available in my resources section, but it's still important to give a basic timeline to show the evolution of cloud computing service offerings:

• 1999: Salesforce launches its SaaS enterprise applications
• 2002: Amazon launches Amazon Web Services (AWS), which offer both artificial and human intelligence for problem solving via the Internet
• 2006: Google launches Google Docs, a free, web-based competitor to Microsoft Office
• 2006: Amazon launches Elastic Compute Cloud (EC2) and Simple Storage Service (S3), sometimes described as the first IaaS
• 2007: Salesforce launches Force.com, often described as the first PaaS
• 2008: Google App Engine launched
• 2009: Microsoft launches Windows Azure

Armbrust, et al. note many motives that drive companies to launch cloud computing services, including:

• Profit: By taking advantage of cost savings from very large datacenters, companies can underbid competitors and still make significant profit
• Leverage existing investment: For example, many of the applications in AWS were developed for internal use first, then sold in slightly altered form for additional revenue
• Defend a franchise: Microsoft launched Windows Azure to help maintain competitiveness of the Windows brand
• Attack a competitor: Google Docs was launched partly as an attack on Microsoft's profitable Office products
• Leverage customer relationships: Windows Azure gives existing clients a branded cloud service that plays up perceived reliability of the brand, constantly emphasizing that it is a "rock-solid" cloud service

These are the motives that bring competitors to offer cloud computing services, but what drives companies and individuals to adopt cloud computing, and what barriers still exist to full cloud implementation.

The Cloud Computing Market: Where It's At, and Where It's Going
According to a study by IT trade group CompTIA, up to 80% of businesses use some form of cloud computing, although the degree of use varies widely. IBM's studies show that although only 8% of businesses believe cloud computing currently has a significant impact on their business, it is expected to grow to more than 30% in the next three years.

Cloud computing is often sold on the basis of price, but the primary benefit companies are seeking from cloud computing, according to recent surveys, is flexibility. With the huge swings caused by viral phenomena on the Internet, companies can see demand for their site and services fluctuate wildly in a short period of time. Cloud computing gives companies the flexibility to purchase computing resources on demand. A more conventional benefit of cloud computing's flexibility is the ability to avoid hiring and firing IT personnel for short-term projects.

One of the major obstacles to full adoption of cloud computing services remains security concerns. Although cloud-based security solutions exist, there is still a perception that cloud computing puts data at risk compared to private datacenters and increases the operational impact of denial-of-service attacks.

Despite these concerns, however, all sectors of the cloud computing market are expected to thrive in the near future, with revenue in nearly all sectors doubling within the next 3-5 years.