The cloud is defined as a service delivered over a network in this post. Cloud computing is different but related, see here. This post takes a look at why the Cloud computing is receiving so much attention and hype, and explains it and the motivation behind it.

In the News and Recent Cloud (hype) Headlines:

“Using 'Clouds' for computing tasks promises a revolution in IT similar to the birth of the web and e-commerce” -- McKinsey March, 2009

"The shift to delivering IT through a utility model is poised to change the business computing landscape as we know it." -- Information Week Oct 25, 2008

"We believe cloud computing will be a hugely disruptive trend for the entire enterprise information technology market." -- Morningstar June 29, 2009

“The current U.S. economic woes will only drive more enterprises to consider and adopt cloud offerings. Spending on IT Cloud services will hit $43 billion by 2012.” IDC

"Why go to the trouble and expense of building and managing complex systems to handle your spiraling data-crunching needs when another company can do it for you?" -- Business Week Nov 21, 2008

Gartner describes the level of hype within the IT industry surrounding cloud computing as “deafening”. However Jackie Fenn, a Gartner VP, commented “Looking at real benefit, rather than the hyped expectations, we see a number of potentially transformational technologies that will hit the mainstream in less than five years, including cloud computing.”

Technically, cloud computing is not new. The concept is widely attributed to John McCarthy; in 1961, McCarthy was a pretty smart computer fellow – he coined the term “Artificial Intelligence” and contributed significantly to that area of study. But in 1961 he predicted that computing power would someday be priced like a utility. This utility model of computing was progressing very quickly through the 70s and into the 80s – at that time it was called “time-sharing”.

Time-sharing, as in the condos, is a pretty simple concept. The cost of the resources are divided among a large number of users. This lowers the cost and increases the availability of the resources to a larger population. With computing, it works because each user’s demand averages out the load. Time-sharing sped up the development of multi-user operating systems. Several companies offered time sharing services in the 70s; including IBM and GE. Users generally connected over telephone modems at “superfast” speeds of 120 bits per second.

But timesharing lost favor as the PC gained popularity. Mainframe computers focused on big tasks instead of lots of little ones, and the smaller tasks found themselves quite comfortable on the PC. The PC became more powerful, inter-networked, and soon the concept of PC Severs came about (Banyan and Novell). PC Servers grew in power and numbers – and like Tribbles they outgrew their welcome. The past decade has been full of server consolidation efforts.

Meanwhile several new technologies built momentum that enable cloud computing. The biggest being virtualization. But in terms of telecom other important technologies include VoIP, SIP, and Web interfaces. Plus telecom, like computing in general, was also been subject to server multiplication to support the increasing quantity of voice related applications.

Timesharing and the utility model are old concepts, what makes cloud computing "new" is the concept of elasticity through the not so new concept of virtualization.  Cloud computing capacity can be scaled up or down dynamically and immediately. Combine this with time-sharing and a utility based financial model and the cloud grows some legs.

Recently Larry Ellison, CEO of Oracle/Sun commented on the hype surrounding cloud computing. He made some very good points: “We’ve redefined ‘cloud computing’ to include everything we currently do. So it has already achieved dominance in the industry. I can’t think of anything that isn’t cloud computing.” He then compared the current predictions of the end of the traditional data center to similar past hype predictions such as the PC would kill mainframes, and that Open Source would kill licensed software.

There is merit to Ellison’s points; mainframes and licensed software are alive and well as is the hype around the cloud. But he isn’t exactly a neutral party either. Oracle/Sun does well with the traditional data center model and isn't too eager to legitimize the attack upon it. But Ellison is ignoring a principle that contributed significantly to his company’s past success; Creative Destruction (Schumpeter). This principle describes how innovation can erode the profits and position of firms (Microsoft, Oracle, Sun, etc.) by changing/improving the model. This is expanded by Nicholas Carr in The Big Switch; stating that IT transitioned from competitive advantage to an also-ran burden - IT is a candidate for creative destruction.

When used properly, IT should deliver competitive advantage. But increasingly this is hard to do. It takes a lot of resources just to keep even. A textbook example of innovation and advantage through IT is the frequent flyer program introduced by American Airlines. This program simultaneously identified and rewarded their best customers while delivering AA information about them. The program created strong loyalty and impressive financial results. Competitive loyalty programs were quickly introduced as defensive measures, but it took years for them to truly catch-up.

Fast forward to present day, and frequent flyer programs are no longer competitive differentiators, but rather a minimum requirement for airlines. Worse, the “rewards” have diminished value (miles are hard to use – thank you IRS), and the costs associated with maintaining this “miles currency” actually hurts the airlines in today's price sensitive economy. This “IT inflation” is the real driver for creative destruction. What if general business eliminated the data center with a lower cost alternative? Mitch Kapor, co-founder of Lotus Software, recently remarked that we are approaching a day when an organization may no longer need IT staff.

Nicholas Carr wrote “...the biggest upheaval since the invention of the PC....IT departments will have little left to do once the bulk of business computing shifts ...into the cloud.” Over the long run, economics always wins. Basically, the hype, reality, and popularity of cloud computing can be summed up in one word; “cost”.

Consider Amazon's storage cloud (Amazon S3) which currently charges $0.15/GB of managed storage. An SMB can store 100 GBs of data for $15/month with no commitment, and no upfront costs. There are additional charges, but the overall point is economics will increasingly drive public cloud computing. Cloud computing is the remedy for IT inflation.

Cloud computing offers several key benefits that effectively relate to economics. Applications can be deployed faster and simpler without upfront capital costs for servers and storage. No net new operational expenses for the data center. Plus simplified remote access. There is little doubt that cloud computing will not have a huge impact to IT.

Telecom though isn't as clear. Telecom is a real time application and that creates a challenge for virtualization. Telecom latency (host and/or network) can be mili-seconds away from being non suitable. Plus few telecom applications are available for the cloud, particularly since so many are tightly integrated to hardware. Also voice applications tend to require strict SLAs – and the true cost of outages can be significant.

Cloud computing will indeed profoundly impact telecom. Additionally, there will be plenty of mistakes made and reasons not to embrace the cloud. But voice is in the process of integrating with many other IT applications that may indeed be destined for the cloud - this includes email, presence, collaboration, and other many more. These are the topics of future posts.

Dave Michels regularly blogs at www.pindropsoup.com and is principal of Verge1.com. You can follow Dave on Twitter at @DaveMichels.