VMware's IPO keeps it real

There is no dispute about the value of virtualisation. With EMC gearing up to sell 10 percent of its VMware subsidiary on the stock market, we'll soon be able to measure the value of the virtual in very real numbers.

At the moment, VMware's revenues are not so much growing strongly as going through the roof. They doubled last quarter to nearly quarter of a billion dollars, substantially outperforming EMC's revenue growth, and it makes sense for the parent company to give VMware more financial autonomy. Not only will it boost EMC's share price, as the parent remains owner of 90 percent of what will doubtless be a vigorously oversubscribed offering, but it gives VMware the clout to compete at the top level for the most important component in IT innovation. People.

It's all about expertise. Virtualisation's importance to the industry is so strong that it may become perversely harder to make money out of the idea. The functions that VMware so successfully sells today will be absorbed into the management and operating systems of tomorrow; the deeper virtualisation is integrated into hardware and software, the more sense it makes. That also makes it harder to make money: the great protoplasmic amoeba of the Windows OS contains the fossilised remains of many good ideas and profitable products. VMware needs to invent new, exciting and useful ways to make virtualisation part of the enterprise, and that means attracting the brightest people to out-think the competition. An IPO, with its attendant options, ready cash and promises of success rewarded, is one way of doing that.

VMware will do well in the short to medium term. With a very strong presence in server management and a strategic move towards more desktop virtualisation, the company is driving the market. Not everything that it says is important now will stay important forever: the much-vaunted power savings from virtualisation are worth having, but you only get them once. Yet the market is new enough, VMware feisty enough and the competition sclerotic enough to leave plenty of room for success.

In the longer term, the outlook for VMware is less clear. If the market swings towards thin-client and service-oriented architectures, then the need for desktop virtualisation may diminish, or be served entirely by the hardware. On the other hand, mobile systems are about to go virtual — Intel's 2008 Menlow platform will introduce the technology to PDA-sized devices, even into mobile phones — a step with tremendous potential for safely extending enterprise IT beyond the office walls. We've been promised that so often that it's slightly difficult to imagine that it might actually happen: VMware cannot afford to let others grab the initiative.

Yet nothing is more virtual than plans for years away. For today, VMware's IPO is very tangible proof of the importance of innovation to IT — and the continuing truth that good ideas well targeted attract ample rewards.