Cloud computing – more that just marketing hype?

Many of the technologies and techniques used in cloud computing are not new. Virtualization which has been around for some years is an integral part of all cloud solutions. Many have argued that there is nothing more to cloud computing than a re-packaging of existing technologies.

There is no doubt that there is a great deal of hype surrounding the term. However there are a few crucial aspects associated with cloud computing that could mean the technology represents a paradigm shift in how we use computational resources.

First among these is scalability. In particular the ‘elastic’ nature of cloud scalability – the ability to effectively scale up infinitely but also to scale down as demand decreases.

Elastic scalability is closely linked to another important aspect of cloud computing – the payment model. In contrast to existing models the cloud provider charges based on resource use rather than any flat fee or contract.
Users pay only for what they use – so they pay more when demand is high but also pay less when demand is low. This payment model bears a close resemblance to that used by utility companies – for example, electricity customers pay only for the amount of electricity used.

It is reasonable to suggest that cloud computing may signal the beginning of a move towards ‘computation as a utility’.

The advantages of cloud computing are obvious and manifold. Firstly initial outlay for computation costs is minimal.

For example, a new website which traditionally would be forced to invest heavily in infrastructure in preparation for a ‘hoped for’ audience of users could instead use a cloud provider and pay nothing until (and unless) the ‘hoped for’ audience materializes.

A similar attraction is presented to existing businesses. If said business expects a large increase in demand for it’s services at a particular time each year (the holiday season perhaps) then traditionally the only option would have been to purchase more infrastructure to cope with the expected spike in demand.

It would be impractical to scale back this infrastructure after demand had decreased. With cloud hosting this would happen automatically. Indeed the cloud would be capable of handling any potential load. Businesses no longer need to plan infrastructure changes based on what “might” happen.

The nature of cloud computing means that cloud providers will necessarily be large companies. The scale of infrastructure will require a huge amount of investment. It is likely that many smaller companies will resell services hosted on cloud infrastructure provided by these large providers. There are obvious privacy concerns with a future dominated by a few large companies.

The idea that a handful of companies will dominate the worlds computation and by extension information is indeed a troubling one. The analogy with other utility providers falls down when the value of information is considered.

It is likely that cloud providers will insist their systems are data agnostic but the theoretical possibility of intercepting data will likely worry many potential customers.

In conclusion it seems that cloud computing is more that just a marketing term. It may indeed signal the next evolution in computer systems.

Even a cursory examination of the Amazon AWS products (for example) provides an insight into the enormous potential and power of the cloud. The prospect of effectively infinite computer power and scalability with remarkable ease of use and a pay-as-you-go payment model is certainly an attractive one.

One obvious caveat is the necessity for a fast and reliable internet connection. Assuming such a connection exists it would almost seem irrational if in future computation was not handed over to the cloud.

Visit Amazons AWS site for lots of info on their services