Choose timezone
Your profile timezone:
Powered by Indico
v3.3.3-pre
Is Moore's law still alive?
Today’s microelectronic industry manufactures millions of chips with billions of MOSFET transistors that count some 30 Si atom lattice constants across their critical channel dimensions. This corresponds to roughly 100 silicon atoms. Insulators to control the gate are about 10 times thinner, i.e. a handful of atoms thick. Researchers have already shown lab devices that are about 5 times smaller, reducing atom counts to about 20. Such devices are still difficult to build as “printing” shapes at these dimensions remains extremely challenging and progressively more expensive, but even assuming that these difficulties are solved through some clever tricks, from first principles there is not much room left for improving the miniaturization of MOSFET devices.
IEDM and ISSCC are the major annual events where the microelectronics industry reports advancements in the device and circuit areas using technologies such as those mentioned above. People at these conferences are quite concerned that we may be at the final stage of asymptotic development and that the long 40+ years run of unbelievable advancements in microelectronics might be close to the end, but still every year new ideas come out that seem to postpone this long announced plateau.
In this talk, I will try to give a short summary of some of these most interesting and perhaps daring ideas that hopefully will give us another 10 years of progress.