Intel is reaching the limits of the miniaturisation of current technology, according to one of its researchers. Moore's Law, which says computing power will double every 18 months, was thought to be good for another 10 to 15 years. But it may not last that long, Paul Packan said. A major problem is with the dopants used to carry charges, he wrote in Science magazine. Transistor gates use dopant atoms with either three or five outer orbital electrons spread though a lattice of silicon atoms, each with four outer electrons. The dopants provide positive or negative charge paths which can be switched on or off by the gate voltage, an effect which forms the basis of all digital computers. The charge the gates have to carry is roughly the same, however small they get. The result is that dopant densities have increased a hundredfold in 20 years to about one per cent of the silicon lattice. Above this level they have a tendency to clump, Packan said. 'Unfortunately the charge concentrations needed for current process technologies are at the solid solubility limit for the dopant atoms currently in use. 'New dopant atoms have been evaluated but none have yet been found to create higher concentrations of mobile charge. Thus, unless new methods are developed, future scaling of the transistor will result in a loss of total charge, an increase in resistance, and a potential decrease in performance,' said Packan. Another problem is quantum tunnelling of electrons across insulating layers as they reach the depth of around five atoms. Gordon Graylish, Intel's European marketing director, said Packan's paper had been published to make people aware of the limits. He added: 'We are still evaluating new materials and we can miniaturise in other ways - by increasing the number of layers in a processor, for instance.'
Latest Tesla news: Tesla share price continues to fall after Saudi Arabia's sovereign wealth fund is linked to investment in rival
SEC 'probe' takes its toll on Tesla as new research suggests that Tesla loses $6,000 on every $35,000 Model 3
RTX 280 Ti will come with 11GB of fast GDDR6 video RAM with a 352-bit memory bus offering 616Gbps
The scale of jobs lost to automation will be at least as large as those in the first three industrial revolutions
10nm Cannon Lake Core i3-8121U CPUs make a rare outing with Intel's NUC mini PC