Sun Microsystems Inc’s chief executive Scott McNealy doesn’t know how the company will make a Network Computer to sell for the fabled $500, but says purchase price is irrelevant because reduced cost of ownership is the key. In light-hearted mood McNealy, referring to Oracle Corp chief Larry Ellison’s predictions of the $500 machine said, When you’re in the software business it is easy to price hardware. I might go round telling people that Java Office software will be free he quipped. Whatever the purchase price of the Network Computer, McNealy is convinced that the return on investment for any company abandoning the desktop personal computer route will be enormous, due to huge savings in ongoing support and maintenance. For Sun, he says, the Network Computer is the company’s entry into the passenger car business, where previously it had been sucessful in commercial vehicles. He is banking on the product to take Sun into the high volume business, which, in turn will encourage increased sales in almost all other parts of its business, particularly high end servers to support the network clients. McNealy, who makes no secret of his anti-Microsoft Corp and Intel Corp ambitions, calls the desktop computer an activity station, and office suites activity generators rather than productivity tools. He looks forward to the day when his children will turn to him and ask incredulously, Daddy did you really have a computer at home? McNealy likens each household with its own computer to each of them having a well in the back garden for water. Who wants the worries of a well when the city pipes in the water through its network, he asks. Similarly, he uses the telephone on the desk analogy. It would be like having a switchboard on every desk, instead of just a telephone handset.
He knows that the picture he paints is somewhat Utopian, and is not for one moment hailing the death of the personal computer, which he likens to the common cold, it won’t go away. However, as Sun did once before when it announced it was coming off its mainframes to practice what it preached about client-server, the company has committed to use Network Computers internally as quickly as possible. McNealy says he is aiming to have 2,500 users converted by June next year, and there are bonuses riding on it, although he admits it will be an ambitious target. The company is overwhelmed at the moment with interest in its Network Computer technology. McNealy admits that Sun is not sure what the first killer application will be. As a result, it does not seem to have really sorted out its marketing strategy, uncertain which market sectors to hit first. He talks a lot about bank tellers, ideally suited to the technology because the end user typically uses the same dozen or so parts of an application every day. These could be downloaded to a Java client every night, he says, or even once a month if there are bandwidth issues. These types of applications are not likely to change that often. The company is poised to take off with Java and Network Computers, although as McNealy points out, it is doing very nicely from its existing power desktop and high-end server business, with analysts predicting around $9bn revenue this year. He would like to say the company was visionary when it developed Java, he said, but admits it was merely an accident of good fortune and a lot of hard work. The company will not make money out of selling Java, McNealy insists, it will make money only out of using it – he does not believe anyone has the right to own the spoken and written languages of computing. The man is making no predictions on what proportion of Sun’s business the Network Computer will represent in the year 2000. It would be like guessing which one of your six children will do best at college, he said. But he has no doubt that the Network Computer model will take off. He only hopes it will not happen so quickly that Sun cannot cope with the demand, and the likes of IBM Corp, Intel and Compaq Computer Corp clean up, he says with his tongue only slightly in his cheek.
By Joanne Wallen
