Gordon Moore, the co-founder of Intel who died earlier this year, is famous for forecasting a continuous rise in the density of transistors that we can pack onto semiconductor chips. James McKenzie looks at how “Moore’s law” is still going strong after almost six decades, but warns that further progress is becoming harder and ever more expensive to sustain
When the Taiwan Semiconductor Manufacturing Company (TSMC) announced last year that it was planning to build a new factory to produce integrated circuits, it wasn’t just the eye-watering $33bn price tag that caught my eye. What also struck me is that the plant, which is set to open in 2025 in the city of Hsinchu, will make the world’s first “2 nanometre” chips. Smaller, faster and up to 30% more efficient than any microchip that has come before, TSMC’s chips will be sold to the likes of Apple – the company’s biggest customer – powering everything from smartphones to laptops.
But our ability to build such tiny, powerful chips shouldn’t surprise us. After all, the engineer Gordon Moore – who died on 24 March this year, aged 94 – famously predicted back in 1965 that the number of transistors we can squeeze onto an integrated circuit ought to double every year. Writing for the magazine Electronics (38 114), Moore reckoned that by 1975 it should be possible to fit a quarter of a million components on to a single silicon chip with an area of one square inch (6.25 cm2).
To read more, click here.