Sun banks on new technology

Sun Microsystems is apparently riding high, turning in financial results that are the envy of the industry and successfully moving into new ecommerce markets. For the moment, it is doing well.

Yet the industry is littered with companies that failed to make the transition from one generation of technology to another. Casualties such as Digital Equipment, Honeywell and Burroughs demonstrate the difficulties of staying on the high wire of innovation and profit.

So are there any signs that the supplier that made Unix a household name in business circles, and developed the Java language to boot, is beginning to develop a middle-aged paunch?

Last quarter, Sun's revenue increased to $5bn, an impressive rise of 60 per cent over the same quarter last year.

But much, as always, depends on chairman Scott McNealy, who is still intimately concerned with Sun's technology direction. After all, he keeps the telephone number of Sun Labs director Jim Mitchell on his speed dial so that he can stay in close touch with him at all times.

The man who at one time could be relied on to insult Bill Gates at the sight of a reporter's pen, however, is now playing more of a family role. These days, he rations his public appearances, preferring instead to busy himself with geeing up the troops.

The word is that McNealy has toned down his pugnacious style and is concentrating on running his company on the advice of his friend Jack Welsh, chief executive of the long established General Electric Corporation.

McNealy still commands the loyalty of his employees, however. "Sun people love him. He's the best reason staff have to stay with the company," said one senior Sun executive in Europe.

But others complain of increasing bureaucracy within the vendor as its ranks swell. Last financial year, 12,000 people around the world joined Sun.

New ideas
Whatever its management style, however, the firm has to keep up a fast pace of innovation. The opportunity to generate profit from some products lasts as little as six months, so a steady stream of new ideas is needed.

As a result, Sun's annual expenditure on research and development is now approaching $2bn, or 10 per cent of its turnover. But the company prides itself on being able to turn its laboratory research efforts into commercial products more consistently than most.

"What distinguishes us from Xerox Parc is we believe if you don't put things into use, the company doesn't see a direct benefit from new products, increased profits or a better share price," said Mitchell.

Winners that Sun Labs have produced in the past include Java, the Jini connectivity technology, the Sun Ray desktop system, and UltraSparc microprocessors.

This is because five years ago, Sun backed three trends that it believed would be key to its business strategy: increased bandwidth, internet services and Java. Hardly a gamble, some might say, but nonetheless, these choices turned out to have been the right ones.

This year, however, Sun has placed another three bets on the future: on massive scalability, continuous real-time computing and integrated stacks.

Massive scalability relates to the internet, which requires systems and applications that can be scaled up to cater for the predicted one billion handheld devices that will be connected to it within a few years. This compares with the 100 million computers hooked up to the web at present.

Continuous real-time computing is enabled by wireless and optical networks that connect users continuously to the internet. Finally, integrated stacks of microprocessors, storage, system software and middleware will be required to support these large-scale and continuous real-time deployments.

Research laboratory
How well these bets pay off, however, is very much in the hands of Sun's top technologists. Earlier this month, the company's laboratory chiefs and star researchers gathered in a castle outside Grenoble in France to toast the supplier's decision to set up a research laboratory in Europe.

Although Sun has yet to construct the facility, and plans to hire just 30 researchers to work in it over the next two years, its decision is significant. The labs will play a pivotal role in Sun's development of continuous computing technologies, especially wireless, an area in which Europe currently leads the US.

Last month, Sun announced it would spend $100m on providing facilities and resources for wireless customers. As part of its so-called iForce wireless initiative, Sun said it is teaming up with more than 50 wireless system, package and content providers to deliver hundreds of new applications and services.

The vendor's new wireless organisation will set up a global Wireless 'excellence centre' in Stockholm, which will open early next year. The centre will work with global carriers, network equipment providers and independent software vendors to develop wireless systems.

As a result, a team of executives, led by Sun's chief technology officer Greg Papadopoulos, is scouring European universities, institutes and commercial research centres for talented technologists who are capable of coming up with Java-style breakthroughs. These are not run-of the-mill engineers, but serious computer scientists.

Although Sun is not yet certain which research projects will be carried out at Grenoble, the internet protocol network, telecoms convergence, internet security and ubiquitous computing are at the top of the company's list.

"Since Europe is so strong in telecommunications, especially wireless, setting up a research centre here is a sensible thing for Sun to be doing," confirmed Mitchell.

But the relationships that the firm establishes with European researchers will be just as important as the staff it hires. "We accept that most of the good ideas will happen outside Sun," acknowledges Jeff Rulifson, director of Sun's labs in Grenoble. Rulifson points to links that were recently forged with scientists in Delhi as evidence of the global nature of the quest for new ideas.

So it will not be through its own ground-breaking innovations that Sun will maintain its grip on the technology tight-rope, but rather the way the company responds to technology changes. And it will also come down to McNealy's ability to select the best path for Sun and to co-ordinate the efforts of a growing army of researchers and engineers.

Key points

• Sun Microsystems is spending up to $2bn, or 10 per cent of its turnover, on research and development
• Sun is targeting Europe to find researchers for wireless technology and will open labs in Grenoble and Stockholm
• The iForce initiative will see Sun spend $100m on developing wireless services.