because the technology has advanced. We complain about the demands of ubiquitous connectivity as we attach cell phones to our belts and of information overload as we put more and more material on our Internet servers. Over this time span, transistors have gone from macroscopic, ~1-millimeter junction-length devices, to ~90-nanometer gates in the latest commercial chips and to ~10-nanometer gates in laboratory devices. This linear scaling clearly must end as devices approach the size of atoms (~0.2 nanometers). This does not, however, mean that progress in electronics and information technology will come to a halt. The integrated circuit paradigm that has enabled this dramatic scaling improvement is a planar, two-dimensional concept based on an interconnection of three-terminal switching elements (transistors).2 Moving to a volumetric approach, new materials, and different computing strategies will probably allow continuation and even acceleration of the capabilities and function per weight/volume/power of electronics. The practical success of miniaturization has been the result of the accompanying dramatic reduction in cost per function achieved by the integration of so many electronic devices onto single chips and using parallel, or batch, fabrication technologies to allow this cost scaling.

Less well understood is the acceleration in other micro- and nanotechnologies, which is being driven by miniaturization and is contributing to the increasing density of information transmitted, stored, and processed. The growth in magnetic information storage in recent years has been even more rapid than growth in electronic information processing.3 Advances in magnetic memory storage range from new giant magnetoresistive nanoscale layered materials to read heads flying 10 nanometers over the surface of magnetic discs moving at speeds of 20 meters/second. To appreciate the challenge in control of tolerances for this technology, scaling to the macro world by the relative lengths of a magnetic read head and an F-18 jet fighter would correspond to flying the F-18 only 100 micrometers above the ground, which has been polished to a smoothness of 10 micrometers and staying on course within an accuracy of 100 micrometers. Optical information transmission has also been increasing at growth rates comparable to that for magnetic memories, aided by control of materials—for example, in optical fibers with ultrahigh-purity microscale cores and semiconducting lasers with nanoscale quantum wells.

Mechanical devices at the microscale and below promise to further extend the reach of miniaturized technologies. Microelectromechanical systems (MEMS) build on the manufacturing paradigm of microelectronics and offer the promise of large-scale batch fabrication at low cost. Currently this emerging technology is primarily focused on simple devices such as inertial sensors for air bag release in automobiles and microscale mirrors for optical projection and switching. However, future applications of MEMS for airfoil control, inertial sensing, or satellite maneuverability could significantly broaden the scope of this technology. The integration of MEMS technologies with electronics and optics is also being explored for chemical sensing, so-called lab-on-a-chip systems. Indeed, the current



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement