Moore’s law has governed the computer industry for almost half a century, but the improvements in performance it predicts may longer be achieved simply by cramming more transistors onto a wafer of silicon.
Every 18 months to two years the number of transistors on a chip doubles because each transistor is half the size of the previous generation. And thanks to clever chip design, power consumption has remained constant, rather than increasing in line with the number of transistors.
This is why it is possible today to have a chip with millions of transistors.
"The problem with Moore’s Law is it is reaching a limit," says IBM fellow Bernie Meyerson. "You do realise that at some point you could have transistors that are one atom thick. When you cut them in half that’s nuclear fission. Let me know when you’re going to do that so I can be far away."
This limit will be hit in just a few generations of chip design, according to Meyerson, given that 14nm fabrication is now reaching production.
Meyerson predicts that silicon chips will not suddenly be replaced by a new technology – because nothing else currently exists that can be used on an industrial scale to manufacture chips. Instead, he expects complex software systems and specialised processors such as field programmable gate arrays (FPGAs) and GPUs (graphics processing unit) will see more mainstream usage to allow the IT industry to continue to yield step improvements in performance.