One of the most obvious manifestations of its cloud DNA is the Kindle’s controversial Silk browser. It’s designed to intermediate all internet activity, routing traffic through Amazon’s EC2 servers - ostensibly to provide a more efficient way to browse. According to Amazon, the number of different hosts being contacted to load a web page adds delay. Its Silk browser goes to Amazon’s servers to retrieve and cache all the necessary page elements.
Not everyone is convinced that this makes for speedier browsing, and some firms may have security concerns about users’ every web request going via Amazon. But it does neatly demonstrate how Amazon’s tablet and its cloud services dovetail beautifully - an appealing proposition for those hooked on Amazon’s AWS infrastructure-as-a-service platform.
Amazon Web Services is now a decade old, having launched as an internal system of distributed computer resources. It wasn’t until 2004 that one of Amazon’s engineers identified the opportunity to sell AWS to other users, and in 2006 the first commercial services became available with its Elastic Compute Cloud (EC2). Once, again, Amazon had identified a massively disruptive business model and set about building its domain expertise.
The low-cost, pay-as-you-go model has appealed to millions of firms across the globe, prompting some market watchers to mull whether Amazon would be best served by spinning off the unit. And not everyone is convinced that AWS is capable of delivering enterprise-grade IT.
Will CIOs learn to love AWS?
Amazon may well be wildly successful selling computing infrastructure in to large companies, but it’s never going to be selling to the CIO or the IT organisation, says Frank Gillett, a principal analyst with Forrester Research.
“Amazon’s offerings are very good for session – web, collaboration – and compute-intensive workloads, but not as much with typical transaction workloads that are typically the core focus of the IT organisation,” he says.
This view will be reinforced by recent service outages, which affected AWS customers including Netflix, Instagram and Pinterest.
To date, functions such as marketing and engineering have embraced Amazon, without necessarily bothering to check with IT, says Gillett.
“CIOs don’t like unpredictable pricing and budget impact, so pay per actual use is not very appealing,” says Gillett.
It is true that long-established large enterprises may be reluctant to bet the business of AWS, says Wang, but for tomorrow’s firms Amazon has become the go-to choice.
“Where you have a lot of legacy infrastructure, it perhaps doesn’t make sense to just throw that out and go to Amazon. But for start-ups, they just fire up an Amazon instance without thinking about it,” he says.
But while AWS may be beloved by the start-up community, it doesn’t mean that the service is just for small fry. Global media streaming firm Netflix has long been an AWS convert. Others, such as photo-sharing service Instagram – now owned by Facebook, but still based on Amazon – have proved that it is possible to build a billion-dollar business on the back of Amazon’s cloud service. And, as we have already seen, an organisation as huge as the UK government has – wisely or not – based part of GOV.UK on Amazon throughout its development and testing stages.
And Amazon has gradually been fleshing out the core virtual infrastructure and storage service to include features that will appeal more to the IT function. These include its AWS Storage Gateway, which provides users with a cloud-based disaster recovery service. More recently, it introduced its high-performance computing services to Europe, with its Cluster Compute Eight Extra Large instance firing up in its Dublin datacentre. This is an instance with more than 60GB of RAM and 3.4TB of attached storage – and is targeted at industries such as financial services, energy and life sciences. As a statement of intent, it’s pretty clear that Amazon doesn’t intend to stay on the enterprise sidelines.
It cannot, however, be assumed that Amazon will get a free run at the enterprise. Google, it seems, has finally awoken to either the opportunities presented by the combination of an infrastructure-as-a-service platform allied to a low-cost tablet – or it has cottoned on to the threat Amazon poses.
In late June, Google unveiled both the Nexus, and its own IaaS offering, dubbed Compute Engine. Clearly, the deep-pocketed Google represents a formidable challenger for Amazon, but as yet, there is no information on when Google’s Computer Engine service will be generally available. And as other competitors have found out to their cost in the past, it doesn’t pay to be playing catch-up with Amazon for too long.
Sometimes, the power of the mainframe is the most cost effective answer. Computing's Peter Gothard puts Computing's readers' questions on the future of the mainframe to IBM's Z13 expert Steven Dickens.
This Dummies white paper will help you better understand business process management (BPM)