IT firms need imagination as well as technical brilliance, argues Paul Hunt, managing director of Phoebus Software
In 1965, Gordon Moore, co-founder of Intel, set out a law to which the technology industry has adhered for the last 46 years. He predicted that the number of transistors which can be placed on a silicon chip for the same cost would double every two years. Since then, manufacturers have constantly moved silicon chip technology forward, to the extent that today the greatest concern for those hoping to preserve Moore’s law is that transistors are now so small, they are limited by the size of atoms. This year, the latest silicon chips will contain transistors more than 2000 times smaller than the width of a human hair and it is uncertain that it will ever be possible to produce transistors smaller than this. Moore’s law may, finally, be broken.
Nonetheless, heroic efforts continue to maintain the pace of progress. Work currently underway in the University of Cambridge plans to halve the width of transistors to just 11 nanometres, down from the current limit of 22. But beyond that point, there’s no known way to make components smaller.
Of course, all exponential patterns of growth will eventually reach a limit. If not, technology would eventually advance almost instantaneously. So the solution in the long run cannot lie in finding ever more elaborate ways of driving the size of transistors on a chip down. Another approach is needed. Rather than making components themselves smaller, advances in the long run will need to look at altering the way chips are made, rather than just refining the process. Earlier in the last century a similar situation occurred in the shipping industry as propellers became so large their shafts were no longer able to sustain their weight. The solution was to find new alloys and blade designs with which to drive up efficiency.
Moore’s law isn’t just a pronouncement on chip technology. It’s also a point about the economics of technological progress. No matter how hard you try, you can’t flog the same horse forever. Unless you’re prepared to find new ways of improving performance, you will eventually come up against a brick wall. To match consumers’ expectations for inexorable improvement in performance and cost requires imagination as well as technical brilliance.
In the mortgage world, the same is true. The lenders who use our software systems are constantly looking for ways to drive down costs and improve performance. Sometimes, this is really just a case of refining our products, increasing speed, flexibility and power. But just as there is a limit to how much you can shrink a transistor, there’s a limit also to what you can do in software without thinking outside the box.
Development shouldn’t just be about finding ways to make transistors smaller or propellers larger. It’s about finding new ways in which to make products more efficient. For Phoebus, this means not only striving always to improve our core product, but also looking closely at the business into which our technology is to fit. That’s why we offer our clients a flexible pricing structure that allows them to pay more or less depending on the size of their mortgage books. It’s also why we believe it’s important to offer a full consultancy service which helps our clients set up their businesses in the most efficient way possible.
Technological advances have to live in the real world. While there’s huge value in trying to refine products that already exist and to make the same level of performance possible for less, it’s also vital to see the wood for the trees. Smaller, faster and quicker are all important, but the requirements of businesses are more complex than any single product. Those of us who build products for business must remember that.
