UK: Death of the mainframe. (2 of 4)

UK: Death of the mainframe. (2 of 4) - The success story of Sun highlights the problems experienced by other manufacturers, which are still organised around the 1960s and '70s model, when each company designed its own systems from scratch and there was l

Last Updated: 31 Aug 2010

The success story of Sun highlights the problems experienced by other manufacturers, which are still organised around the 1960s and '70s model, when each company designed its own systems from scratch and there was little interchange of information between the manufacturers.

Today the average product lifecycle is only nine months, and it is increasingly difficult to find the funding for the development of a new range. IBM spent $1 billion on developing its new personal computers and Digital Equipment spends $1.6 billion a year on research.

Frank Owen, technical director of component manufacturer Texas Instruments, believes that semiconductors are a commodity product which can only justify their development costs by mass marketing, and he advises against computer manufacturers trying to do it all themselves. "We believe you have to be a major investor in the semiconductor market to have the right costs," he says. For example, a new manufacturing plant in Italy for advanced semiconductors is costing Texas Instruments $1.2 billion to build. It will take high volume production to recoup that kind of investment.

NCR, one of the top 10 American manufacturers, recently took the bold decision that it would no longer make its own components but would buy in standard chips from Intel for its latest range of computers. Michael Keville, director of product marketing at NCR UK, says that "off the shelf is cheaper", but points out that there are further cost benefits which follow in the software development of a new system. NCR will continue to do some component development in house, but only those which are used for specific functions and give the computers "added value" to the customer.

NCR's move to sharpen its cost base was timely. It has just become the subject of the first ever hostile takeover bid by the giant American Telephone and Telegraph Company (AT and T). AT and T's $90 a share offer, valuing NCR at $6.12 billion, is seen by the industry as the first serious attempt to bring about the long-mooted convergence of telecommunications and computing technology.

It is a sign of AT and T's desperate need to enter this market and meet its arch-rival IBM head on that it has taken the unheard-of step of mounting a hostile bid. But its move is logical. IBM after all is in deep discussion with British Telecom which would create a similar alliance to the one that would emerge from an AT and T victory over NCR. The BT-IBM link, first mooted and abandoned under government pressure six years ago, would cover the supply of a range of advanced computer and communications systems.

Apricot, the British manufacturer recently bought by Mitsubishi, was one company caught out by rising costs. Dr Peter Horne, managing director of the new Mitsubishi subsidiary, blames the shortage of investment funds in part on the narrow views of City analysts who looked only for short-term profits. "Most investors appear to have short horizons, especially when there is no guarantee of success. Apricot needed to invest a lot more and couldn't, because it would have depressed the share price even further."

With a turnover of only £140 million, it could hardly have put in the funding of IBM or DEC; even so, Horne says that it was able to develop the first product of a new range easily, but the money for the rest was difficult to find. The original company was divided with the sale to Mitsubishi, the software business remaining in the hands of its founder, Roger Foster. Under new Mitsubishi ownership, the Apricot manufacturing business is reorganising and investing in new plant in order to become more efficient.

However, the advent of open systems has done more than change the manufacturing cost equation: it has shifted the focus of the entire business away from hardware and on to the design of computer systems and the software that they use.

In the 1960s and '70s it was usually only large businesses which could afford computers. These usually had set tasks to do, which did not interrelate with anything on other computers.

The personal computer in the early 1980s was the catalyst for change. The relatively low cost of a PC meant that computers came within reach of a wider range of businesses. It also meant that large companies could afford to give a PC to employees in their branch offices and let those people have access to corporate databases, or feed information into a head office computer down the line.

This provided the spur for the networking industry which has grown at a rate of 20 to 30% a year. Specialist network companies, led by 3-Com and Novell, have gone from strength to strength in the 1980s.

Find this article useful?

Get more great articles like this in your inbox every lunchtime