How things have changed. Now the latest chip technology goes by names such as 'sub-micron' and 'deep sub-micron' - one micron being 0.0001 of a centimetre, or one-millionth of a metre.

Hong Kong's chip-makers are more likely to be making the circuitry behind singing greeting cards than the chips powering the world's most advanced computers and electronic devices.

'Hong Kong was early in the wave,' Mr Cheng said.

'But when the technology came around, we missed out.' Hong Kong's non-interventionist government was not keen to subsidise the capital-intensive chip industry, Mr Cheng said, and development of the property and financial services sectors was a much higher priority.

'Taiwan and Singapore were later, but they went in that direction,' Mr Cheng said.

These facts alone have prompted many sceptics to dismiss the idea of building the type of facilities now proposed by the Government and the companies behind the US$1.2 billion Silicon Harbour project.

However, Mr Cheng and the engineering faculties at Hong Kong's universities say the project may be the best chance that Hong Kong has had in a long time to regain its standing in the field of chip design and manufacturing.

The next wave of technology involves building customised application-specific chips that will go into the mobile phones and electronic gadgets that Asians are expected to buy in ever-greater numbers.

'The mainland and Hong Kong consume $6 billion in chips every year,' said Ping Ko, dean of engineering at Hong Kong University of Science and Technology (HKUST).

'They don't have that kind of fabrication, therefore it has to be imported. In the medium to long term, China collectively needs to produce its own.' Mr Cheng said though Hong Kong might not be competitive with other countries when it came to standardised commercial chip-making or in developing new technologies, it had always been good at customising.

The Silicon Harbour proponents were 'not talking nonsense' when they said Silicon Harbour would be 'producing for this part of the world, and close to the market.' Though few details have been made available by the Government or by H&Q Asia Pacific - the venture capital firm behind the project - Hong Kong's micro-electronics experts say the size of the planned investment indicates that manufacturing will be a part of it.

That is also good news because the massive investment in equipment will mean that operations cannot be moved out of Hong Kong as soon as the local economy turns sour, unlike service industries that need only high-speed communications lines and can relocate easily.

'Of the hardware side, semiconductors are the most appropriate for Hong Kong. They provide the highest revenue per square foot,' Mr Ko said.

Land requirements would be small and the high cost of labour would not be an issue because relatively few scientists would be needed and their salaries would represent a small portion of the total cost, he said.

According to Johnny Sin, a micro-electronics specialist at HKUST, the obstacles to the success of such projects included government agencies that did not have a sufficient understanding of the technology.

Mr Sin's program's newly opened micro-electronics lab was delayed for two years because safety inspectors and other officials did not understand the project, he said.

'Overall, I don't think they had a direction. I think it is getting better now that they are willing to bring in foreign experts like [Chang-lin] Tien,' Mr Sin said.

Mr Ko, who specialises in chip design, agreed.

'I have to give credit to them, they are changing. But as far as their capacity for understanding the industry, it's still very low,' Mr Ko said.

Hong Kong's schools were equipped to provide the kind of graduates the Silicon Harbour projects needed, Mr Ko and his colleagues said. Experienced designers would have to be recruited from overseas for many of the top jobs. However, for less specialised jobs, the thousand or so electronics engineering and applied science graduates from Hong Kong's universities each year should be sufficient, they said.

Currently, many of these graduates could not find jobs in the field, Philip Chan, head of HKUST's electronic engineering department, said.

'If they insist on having this type of job, it means going overseas,' he said, to places such as the United States or Singapore.

Those who stayed in Hong Kong often ended up on the services side of the business, working for banks or as marketing engineers, Mr Chan said.

He hoped that would change with projects such as Silicon Harbour. A wafer manufacturing plant would probably attract a nearby cluster of related companies, he said.

If such plans come to fruition, they could help vault Hong Kong back to the top of the technology heap, though Mr Ko and his colleagues are reserving their enthusiasm until more details are known.

'Right now, I would treat it as a little bit more concrete than a rumour,' he said.