Intel redefined its business when the chip was down

ON NOVEMBER 22nd, 1994, the siege began. Andrew S

ON NOVEMBER 22nd, 1994, the siege began. Andrew S. Grove, the Hungarian emigrant who had risen to become head of the world's most successful computer chip company, received an urgent call from Intel's PR department. A crew from CNN was on its way to the company, armed with a story about a flaw in the much vaunted, latest generation, Pentium chip.

At that moment Mr Grove made a mistake. He already knew about the "floating point" flaw in Intel's new chip, and that, it occurred just once in every nine billion times the computer carried out a division calculation. The average spreadsheet user would run into the problem once in every 27,000 years of use. Intel would tough it out, he decided, and not recall the problem chips.

All of the major US and international newspapers picked up the CNN story, and dozens of television crews camped outside Intel's headquarters in Santa Clara, California. Intel announced a replacement policy, but only for people whose use pattern suggested they needed it.

Then IBM said it would stop shipments of all Pentium based computers.

READ MORE

Eventually, Mr Grove relented. Anyone of the millions of computer owners who wanted a new chip could have it. The company took a write off of $475 million (£308 million).

"And," he says now, "we embarked on a whole new way of doing business."

It took the Pentium episode for Intel to realise that its market had truly changed, he believes and that the end users, not the company's immediate computer manufacturing customers, were now in the driving seat.

For the 59 year old Mr Grove, the Pentium error of judgment was most uncharacteristic. But his reaction to it, demanding a new openness to change throughout the organisation, was typical.

He has been with Intel since it was set up in 1968 and as he climbed the managerial ladder, the young Andy Grove made quite an impression. By 1971, he had instituted a sign in list for all employees who arrived at work after eight o'clock in the morning. As he saw it, such people were slackers and should be forced to confront their tardiness.

By 1979 he was president of the company and in 1987 he became chief executive.

At the company's headquarters, he sits in a cubicle like everyone else. He encourages staff to stop by his desk to debate his decisions. It's not that be wants Intel to be a democracy, he just thinks that creative conflict and discussion make for a healthy organisation.

He appears to be right. Under Mr Grove, Intel has captured close to 80 per cent of the microchip market, and its gross profit margin stretches to 60 per cent.

To reach this peak, Intel has had to be adaptable and Mr Grove sees this as the key to all business success. He has recently published a book, Only the Paranoid Survive, outlining how a small technological advance can sometimes change a business forever.

The trick is to recognise the moment - he calls it a "Strategic Inflection Point" - at which a company must either remain on its current path, and decline, or initiate radical change, and soar. The problem is that these key technological changes are happening more frequently.

"Remember that 90 per cent of all engineers who have ever lived are alive today So the number of technical people is accelerating and consequently the amount of technology is accelerating at a similar rate," he says.

The Internet is just the latest example. It has infused Intel from top to bottom, he adds, with particular emphasis on whether the idea of a "network computer" - a dumb terminal linked by phone or fibre to a central server - could spell trouble for Intel's sales.

"It is going to come about, but in a way that is different from its original conception indicated. It will be more balanced than just a dumb terminal. Partly because network connections, even telephone connections, are not sufficiently reliable and sufficiently robust to provide 100 per cent safety for computation," Mr Grove believes.

In fact, due to increasing demands for data transfer of sound, images and video, the conventional networks will be strained even more.

"Consequently the way people will implement these things will be with continued reliance on the client as well as on the server. When you take that - and that because of the reach of the Internet - computer use is going to become even more pervasive, the net result is going to be a very strong positive for Intel".

Mr Grove is impressed by the report of the Irish Government's Information Society commission, which suggests tax incentives and better education to help make Ireland a technology hub in the next century. But he insists that the market, not government, should choose the technology.

"I think it's a wonderful proposition. I don't know how big the tax breaks are, but the general direction that they are going is excellent . . . But as wireless technologies come in, both land based and satellite based, they will provide very economically interesting alternatives to fibre optics, and threaten the business model of the land line based companies.

"I don't favour government getting into those things at all, if for no other reason than government is unlikely to be able to pick the right technological direction. Nobody can! I don't know it either! It's competition that determines that," he adds.

However, the education policies of several previous governments have helped Intel's success in Ireland, he says.

"Intel was located in Ireland because we felt that there was an education system in Ireland that was producing a larger number of technical people than there was job opportunities for. So they were plentiful and available."

"Consequently we had a superb start up. So we are just building on exactly the same strengths that attracted us there in the first place."

Strangely, for his native land he seems to have less interest. He still speaks in a Hungarian accent, but in 40 years has never returned to Budapest.

"I have no contact with Hungary ... I left that part of my life behind. I'm not really curious. I miss it, as I remember it, but not enough to open that phase up again."

He has made his life in America, and he is happy married for 39 years, with two grown daughters, and only vague plans to retire.

"The first chief executive of Intel gave way to the second, then stuck around in a particular capacity that appeals to him. The second one gave way to me, and he's still around working part time. I'd like to figure out my way of doing something similar as the years go on. To diminish my activity but, not cut it off. And not yet!"