Hobbies And Interests
Shrinking Components
Computer chip makers already work at the scale of between 20 to 200 nanometers when they make the microprocessors and memory that go into computers, game consoles and smartphones. Wiring that connects components can be as narrow as 20 to 30 nanometers, or about one-fiftieth the width of a human hair. About a billion transistors, resistors and capacitors fit on a chip a few millimeters square. In 1986, the state of the art was one micron, or 1,000 nanometers; in just a few years, components may be under 1 nanometer, which is only a few atoms in size. The traditional way of making computer chips is to make photographic patterns with ultraviolet light. As light wavelengths get smaller to make smaller transistors, the process, called photolithography, becomes more difficult. Chip makers may have to use other methods to create nanometer-scale circuits.
Nanoelectronics
Smaller transistors let a computer maker pack more complex and sophisticated features into its products. Today, a shirt pocket-sized smartphone has more computing power than a room-sized computer from the 1970s. This trend will continue with nanoelectronics, which will use atom-thin wires and molecule-sized transistors. These devices will continue the trend of more memory and calculating ability in less space and with lower power consumption. As these devices shrink, the electrons that power the circuits will have a greater tendency to ̶0;leak̶1; or drift between components. Electricity is well-behaved in a power cord, but when a wire becomes too small, electrons can jump through narrow electrical insulators, creating problems. While not a shock hazard, this will pose a challenge to engineers designing circuits in the future.
Mechanical Computers
In the 1800s, well before electronics, pioneers proposed and constructed calculating machines made of mechanical gears and levers. Nanotechnology may revive this idea, implementing computers as mechanical systems. The mechanical calculators of the early 20th century were reliable, but slow and clunky compared to electronic machines. However, molecule-sized mechanical parts can operate at speeds closer to electronic circuits, and without needing oil or wearing out. Properly-designed molecular machines will operate with extremely low friction.
Nanobots
The first electronic computers of the 1940s each filled a large room with vacuum tubes and wiring. Only a handful of research machines existed, and few of their users would have guessed that just 70 years later, small toys would have computers, and cars would have a dozen. As the computers have become smaller, costs have also shrunk; simple microprocessors cost just a few dollars. As this trend continues, smaller goods will get computers and software. By making computers out of molecular parts, they can be squeezed to the size of a virus. Microscopic robots called nanobots, having their own simple computers, may someday search your bloodstream for infections or eliminate toxic materials in landfills.
Massive Parallelism
Computer makers today build the fastest machines by linking up thousands of processor chips. They break calculations up into small sub-tasks, assigning each part to one processor, then combining the results. This idea, called massive parallelism, lets a computer user add power to his system simply by plugging in more processors. A processor made with nanotechnology will have linear dimensions up to 1,000 times smaller than current ones, which means that a computer system can fit 1 billion processors in the same three-dimensional volume that now holds just one. Massively-parallel computer systems with millions or billions of processors will help solve tough scientific, technical and commercial problems.