Technology on tap

Forecast to be the biggest trend to influence enterprise technology this decade, utility computing promises to deliver scalable, efficient IT services on tap.

  • E-Mail
By  Patrick Phelvin Published  January 25, 2004

|~|tap_M.jpg|~|Utility computing promises technology on tap.|~|After two years of excitement and marketing fluff surrounding utility computing, the concept is at last being adopted by some major organisations around the globe. Here in the Middle East, vendors are reporting that a number of companies are making the first tentative enquiries about the model and Saudi Arabia’s Arab National Bank (ANB) is testing a beta version of HP’s Adaptive Enterprise offering. IDC predicts utility computing, which has gained momentum over recent months, is set to boom in 2004, with global spending of US$1 billion on the concept last year set to double in the next 12 months. Gartner, too, predicts utility will be the most significant long-term trend for enterprise computing of recent years. “Utility computing is a major industry trend for the IT services industry. It is not just a vision. We believe utility is really important, we started forecasting in 1999 that one of the major drivers pushing ahead the IT services infrastructure, or any particular outsourcing concept, was going to be standardisation. The best term to capture this revolution is the industrialisation of the IT services industry,” says Claudio Da Rold, vice president & research director, Gartner Europe, Middle East & Africa. The promise of lower total cost of ownership (TCO) for IT systems is the driving force behind utility. Under the broad concept of utility computing, which is touted by HP as the adaptive enterprise, Sun under its N1 Initiative and IBM with computing on-demand, companies will no longer have to invest in systems that can cope with peak demand in terms of processing power and storage. Now, driven by an ever-closer alignment of business objectives and IT investment, the model promises to deliver the infrastructure organisations need, when they need it. The most attractive part of such propositions for end users is that they will be billed only for what they use. Utility computing also means enterprises will be able to focus on running their businesses, rather than worrying whether their IT infrastructure can cope with a growing workload, or that they will achieve significant return on the dollars they have invested in their existing IT infrastructure which, for the most part, lies unused for two thirds of the month. In addition to ANB’s pilot project some companies here have started making technical and pricing enquiries about the model following successful utility projects in the US and Europe. So far, these have generally been international companies with a local operation in the Middle East. HP, for instance, says some of its data centre implementations are now being angled towards its adaptive enterprise vision. “We do have data centres over here which are already implemented and are gearing towards adaptive enterprise,” says Nasser Ballout, HP’s principal consultant in charge of next generation infrastructure practice in the Middle East. “Utility computing is momentous right now, it is growing slowly here because utility computing has to be made as a strategic decision, but it is a trend that has great momentum,” he adds. Sun Microsystems has also had a number of enquiries from Middle East companies looking to explore the potential cost savings of utility computing relating to their data centres. “In the region I have talked to a few customers about billing and leasing models in terms of general enquiries to see if we are able to offer something along these lines,” says David Drinkwater, Sun’s data centre solutions manager for Southern & Eastern EMEA. “70% of IT budgets are spent on current operations, there is a big move on to make those operations more efficient and less costly. What’s leading us towards utility computing is the desire to reduce costs and increase revenues and everybody wants to do that,” he adds. ||**||Grid computing|~|ballout.jpg|~|There is huge momentum behind utility computing says HP’s Nasser Ballout.|~|Oracle also expects the region’s businesses to embrace its flavour of utility, grid computing. Oracle’s 10g automatically provisions processing power and storage throughout an organisation in response to workload demand. The company has already made its 10g Application Server available in the region and the Middle East release of its 10g Database is imminent. “We are anticipating that some of the large organisations that have a fairly large population of servers will be the first to start exploring this technology,” says Oracle Middle East’s marketing manager, Ayman Abouseif. “With grid computing you can transparently move workload between servers. A lot of our customers others in the region will be able to free up a lot of processing power [with 10g]. We believe that people will start playing with it soon. Under the current economic pressures I think people will be forced to look at ways to not have to buy servers at peak for every single application,” he adds. Despite no shortage of people prepared to testify to the significance of the drive towards utility concepts, the actual speed at which true utility computing will be embraced by businesses, and the ability of vendors to actually deliver on their utility visions, remains questionable. One of the main problems facing the adoption of the model worldwide is that many potential utility customers have been left bewildered by the range of solutions now being touted by vendors under the utility computing umbrella. In an effort to clarify the situation, IDC recently listed a number of defining elements of utility computing. It said utility consists of virtualised systems, applications shared among businesses realtime usage-based pricing, and ultimately realtime adaptation of IT systems to changes in business events. While the IT community agrees that utility computing is a good long-term bet and makes solid financial sense, they are divided over how quickly the model will be adopted because of confusion in the market and a lack of clarity over how businesses can leverage the model to reduce costs. Even analysts seem confused about the model. For instance, just days before IDC forecast blockbuster growth for utility computing, senior IDC analyst Frank Gens branded the model “futility computing” and argued that the lack of understanding of the different flavours of the model threatens to undermine the short-term future of the model. “Confusion about the meaning and value of utility computing will persist. The fact that the basic message is still in need of clarification means that leadership continues to be up for grabs in this important and nascent market,” wrote Gens. “Leadership will not be defined by size of advertising budget. In 2004, leadership will begin by moving beyond simple, broad visions of utility computing to segmented, solution-focused views of utility computing and its value, as well as step-by-step adoption road maps for different classes of customer and solution,” he added. This confusion is especially likely to impact the adoption of the model in the Middle East. Because utility is essentially a new business proposition rather than a technology trend, it will be up to the region’s chief financial officers to decide when to make the move, and they are likely to sanction a move before cost benefits of the model have been proved. Also, vendors keen to be seen as driving the utility model forward are applying the utility tag to many projects and products. “There is an old rule that says if your industry is going to become commoditised you will be better being the first doing so,” explains Da Rold. “From the service provider point of view, this is a trend that cannot be stopped and you are best being the first in. When HP started with adaptive enterprise and when IBM started with on demand they were spending a lot on marketing and a lot on R&D, this was a bell for the industry,” he adds. ||**||Into the future|~|Ayman.jpg|~|Oracle’s Ayman Abouseif says economic trends will drive demand for grid computing.|~|Already, this consolidation has led to Veritas purchasing Ejasent for US$59 million in an effort to boost utility offering. Ejasent’s UpScale software transfers applications between severs, while its MicroMeasure product allows companies to track computer resource use for billing purposes. An additional hindrance to the widespread adoption of the utility model is the evidence that end users are more cynical of ambitious cost cutting promises than vendors would like. After all, the likes of HP and IBM must still recoup R&D spend on their hardware regardless of how many CPUs an end user decides to switch on within that system. In addition, vendors are some way from actually being able to deliver a flexible, pay as you go, and scalable solution. “Of course the on-demand, the flexible or the adaptive enterprise are essentially the way providers try to get deals,” adds Da Rold. “It’s part of the style of this industry to both to announce new offerings and new developments and at the same time as selling these to build the references and to experiment and evolve,” he adds. Sun counters this argument by saying that although their N1 customers will be able to reduce their IT budgets, the proposition will be able to appeal to a wider number of customers and therefore, revenues into the company will increase. “We have to respond to our customers’ demands [for utility computing] and our objective is that our customers run their businesses efficiently and productively,” says Drinkard. “We are not going to put ourselves out of business just because of utility computing, the reverse is actually true. We can open up doors to new customers because it [utility] provides financial justifications [for them to invest in IT] that previously may or may not have been there,” he adds. Another constraint on the development of the model is fact that companies have already invested significant amounts of money in their current IT infrastructure are looking to achieve ROI on their initial outlay. In an effort to negate this factor, the main vendors are at pains to point out that their utility offerings are compatible with existing systems. “This [10g] technology is an evolution rather than a revolution from an Oracle point of view,” says Tarek Shahawi, Oracle Middle East’s technology manager. “It is a natural progression of our Real Application Clustering (RAC) technology. Users can buy as they go and mix and match different boxes in the grid, even though some of them might be faster than others,” he adds. Despite these concerns about the utility concept, a tightening of IT budgets looks set to drive the model throughout the world. In the Middle East, it is large enterprises such as banks and telcos that are leading the charge but the trend looks set to move downmarket in the next few years. “If you are talking to an IT manager, their may be some interest in the billing and leasing model from a technical aspect of what can be achieved, but really you have to talk to the CFO level. They are the ones that are actually going to make business decisions about this, definitely that is where future interest is going to lie,” says Drinkwater. Ballout agrees: “Utility is not really to do with the technology, it is to do with the company strategy to move forward. It is and it is really business driven over here and each [company] has a different view about what it wants out of utility computing,” he says. “It is growing slowly, it will start with large organisations and companies that provide services to the public domain,” he predicts. But the utility market is far from mature, and billing and leasing models have yet to be tried and tested. This means the model is best left to those firms that can invest time and money in ensuring current offerings are right for them. “Our current advice to clients is to first enlarge their mind and evaluate new kinds of offerings and then to leverage on the opportunities if those opportunities fit with strategy or the need,” says Da Rold. “The pressure on costs and economic uncertainty are a strong elements to go in this direction. But be aware that this is a long-term trend and not everything is readily available today so it is a move that has some risk embedded. Wait-and-see is probably the savviest approach. It’s important to separate marketing from reality,” he warns.||**||

Add a Comment

Your display name This field is mandatory

Your e-mail address This field is mandatory (Your e-mail address won't be published)

Security code