Cloud Computing – A Growth Business In The Downturn

There was a great story on the BBC news at the beginning of June 2009 about The Cloud Appreciation Society who wanted to register a new Cloud type – the Asperatus. Much of the initial work to classify real clouds was done by Luke Howard (The Godfather of Clouds) two hundred years ago, but until now in meteorological terms there have been only two basic types of cloud formation, Cumulus and Cirrus, from which derive the 10 main and 27 sub types.
Similarly when it comes to Cloud Computing there’s an urgent need to describe, define and/or – if possible – simplify the term so we can discuss its suppliers and users, as well as measure and forecast its offerings.
It’s also worth asking whether or not it has a special role to play in our current market downturn and recovery, which was deeper than those of 1992 and 2001. The two aims are linked, because if we can’t find an acceptable definition it seems unlikely that its services will help us much. Readers of my reports might also be interested in my presentation on Cloud Computing.

Defining And/Or Describing Cloud Computing

There have been a number of attempts to describe and define the subject. For instance:

• It ‘enables the delivery of personal and business services from remote, centralized servers (the “cloud”) that share computing resources and bandwidth – to any device, anywhere’ (IBM)
• It’s ‘the delivery of IT as a service to companies or end-users over the Internet’ (HP)
• ‘The Network Is The Computer’ (Sun)
• It’s ‘moving the Web forward as a platform for scalable applications’ (Google AppEngine)
• It’s ‘a web service that provides resizable compute capacity in the cloud. It is designed to make web-scale computing easier for developers’ (Amazon EC²)
• It’s ‘Software Plus Services’ (Microsoft)
• ‘When you use any app that runs in the cloud, you just log in, customize it, and start using it’ (Salesforce.com)

There are of course many other more detailed descriptions from these and other players, but the obvious conclusion is that the description and definition is so far a subjective issue. If we turn to Wikipedia we read words ‘style’, ‘concept’ and ‘metaphor’. Again this indicates that we’re dealing with something that can be described, but not yet defined in a logical way.
My view is that it is not yet possible to provide a full logical definition of Cloud Computing. However it is useful to describe it from the point of view of various participants, which I break down roughly into User and Vendor Participants.

Four Types Cloud Computing User

User Participants (Figure 1) view Cloud Computing differently dependent on their role within the organisation. They can be broken into four main types as follows:

• Computer End-Users – it’s the provision of application services over the Internet (or an Intranet) typically through a web-browser. The word ‘Cloud’ implies that the mechanisms of application are hidden – the user supplies the information and receives the solution.
• Software Developers – it’s the development of distributed scalable applications typically through the use of dedicated tools supplied by Infrastructure suppliers. Typically suppliers charge only for deployment on a ‘per seat’ basis. In the case of ‘private’ (rather than ‘public’) Clouds the internal data centre will take the place of the infrastructure supplier.
• Data Centre Managers – it’s the implementation and management of typically large-scale systems using optimised components. Usually deployed in dedicated computer rooms and using rack-mounted components these configurations deploy the software and hardware developed by infrastructure suppliers. Some data centre managers of course work for suppliers (such as Amazon, Google, IBM and Sun). The management of Service Level agreements (SLAs) is of prime importance in Cloud data centres.
• Financial Directors – they need to manage cost, balance capital expenditure against operational expenditure and agree SLAs with data centre managers; as Cloud Computing services are akin to outsourcing there is a potential conflict between financial directors and data centre managers over resources.

Vendor participants can be broken down into a number of different types. For instance:

• Infrastructure Suppliers – it’s the development and deployment optimised software and hardware to drive massive, scalable applications across the Internet. On the software side it includes virtualisation, security, containers, languages, operating systems and browsers. On the hardware side it includes the development of specific grids, clusters, servers, routers, gateways, storage systems and racks to provide massive scale in the data centre. Replacing individual with shared power supplies for multiple servers is an example optimisation.
• Service Providers – it’s virtually any application that involves the provision of services to customers over the network. Telecoms players tend to favour Cloud applications (especially on mobile phones) as they drive data network usage.
• Outsourced Data Centres – is the provision of resources to provide scalable solutions to software developers typically on a ‘per seat’ basis. In this category the needs are identical (although often even large scale) than the user data centre managers.

I believe the Infrastructure Suppliers can also be split into four sub-categories (see Figure 2, where I’ve also given supplier examples).

As the market developed in the first half of 2008 there was some debate about whether High Performance Computing should or should not be included in Cloud Computing. The fact that a number of Universities (North Carolina State, Ontario and Kogakuin University (Japan) with IBM for instance) and at least one supercomputer facility (the NSC in Sweden with HP) have announced Cloud Computing deployments suggests that they want to be included.
Another issue is about scale. Initial Cloud Computer deployments were exclusively in massive data centres – in fact by July 2008 there were probably only around 100 such sites worldwide. It is worth considering how many of the 8k data centres in EMEA (organisational computer systems located in environmentally controlled rooms) will deploy Cloud Computing in future. While I believe developments continue at the higher end, there is no reason why medium and even small businesses shouldn’t deploy cloud servers in the future.

Is Cloud Computing Relevant In A Downturn?

I remember well that the explosion of the Internet bubble pre-dated the 2001 downturn by around a year. That was also associated with the massive over-spend by Telecoms suppliers on 3G licences. It is perhaps a bit ominous when we think of the similarity between Cloud Computing and offerings such as HP’s eServices of 10 years ago.
Earlier this year Sun closed its Network.com to new customers, but maintained the service or the 13 existing clients and their 48 applications – indicating that, for that supplier at least, the early investment hadn’t worked. But (as we know) Sun has had other problems leading to its proposed acquisition by Oracle . On the positive side, however Sun then launched its Open Cloud Platform.
In addition Google (6%), Amazon (18%) and Salesforce.com (23%) are three of the very few major vendors able to post positive revenue growth in 2009. Although Cloud Computing is not the only thing these vendors offer, it is a good indication of the comparative strength of this type of computing.

As for the hardware suppliers, it looks as if optimising devices and systems for Cloud Computing is one way to off-set the huge IT decline between 2008 and 2009 and very relevant to the coming recession in 2011. I also expect that a number of companies will set aside unused equipment to run outsourced applications in a parallel way to the way they developed extremely user-focused offerings after 2001 – any one remember on-off-upgrade-on-demand, where a supplier added extra processors and offered to lease the extra capacity temporarily?
On balance I believe that Cloud Computing is a boon in the downturn and that vendors associated with the subject will do better than those who are not.
My aim in introducing this Cloud Computing theme is to interview and provide an overview of user adoption and vendor approaches with a special emphasis on the way Cloud Computing can help in the downturn. But don’t worry – I’m not aiming to become the Godfather of Cloud Computing!
Please let me know if you have an interesting case to study… and whether you agree or disagree with my description and definitions.
Linkedin users are welcome to download my Cloud Computing presentation.
Please click here for more of my articles on Cloud Computing. We’ve covered many vendor strategies, customer implementations and have applied our own definitions to sizing country markets.