Let's Get Small: The Shrinking World of Microelectronics

[Princess Royal]

Let's Get Small: The Shrinking World of Microelectronics

Small Beginnings: From Tubes to Transistors

Electronics have become so prevalent in our world—in computers, cell phones, airplane control systems, space ships, DVD players, coffeemakers, etc., that it’s difficult to imagine what life would be like without them. You couldn’t read this page without them, couldn’t walk through an automatic door at the supermarket, or have the bar code of your soda scanned, or have the cash register figure out your change, or pay with your debit card, or…well, you get the idea—our culture is powered by electronics.

A portion of ENIAC. A modern pocket-size calculator has more computing power.

It wasn’t always like this of course. At one time electronics were relegated to just a few areas, such as radio and television. A big reason for this was because electronics themselves were big. If you’ve ever seen pictures of early TV sets and radios from the 1940s and 1950s they were large, cabinet-size devices that looked more like furniture than like cutting-edge electronics. And computers? The predecessors of the latest 12 inch, five pound laptops were machines like ENIAC, the world’s first general purpose electronic computer, which was developed in the 1940s. ENIAC was so large it filled entire rooms! You would think with all that bulk it was powerful too. Wrong. Although ENIAC was a marvel for its time, its computing power is dwarfed by a simple modern pocket calculator. So, how did electronics infiltrate just about every appliance we use? They got smaller, and smaller, and smaller. Engineers have spent a good part of the last 50 years shrinking electronic components. This is the field of “microelectronics,” the guts of modern electronics.

In the early days of electronics, that is before the 1950s, the basic electronic device was the electron tube (which is also commonly known as a vacuum tube), which had begun life years earlier as a modified light bulb, and stayed about that size. Electron tubes made early electronics such as radio possible, but they had some serious limitations. Their filaments burned out just like a light bulb, and to make something work you needed lots of them. ENIAC, for example, needed 18,000 tubes to function. But electron tubes were also incredibly useful. In a radio or phonograph, they could take an extremely weak signal and amplify it loudly enough so that it could fill a room. The electron tube could also be used like a switch, but unlike a regular switch it had no moving parts and so it could switch on and off incredibly fast. Computer engineers, who used electrical switches to construct elaborate “logic” circuits, chose to use the electron tube despite its size and tendency to fail.

During World War II, things began to change. Engineers undertook a bold experiment to try to pack an entire radar set into an artillery shell. They called their new device a “proximity fuse,” because it could destroy by being near a target rather than requiring a direct hit. Even though they were a success, proximity fuses still relied on electron tubes, albeit, quite tiny ones. After the war, as missiles and rockets emerged, there was an increasing need for compact, rugged electronic systems for communication and navigation. The search was on for smaller and smaller electron tubes.

While some engineers worked on building better and smaller electron tubes, others were looking for ways to do away with tubes altogether and turned to semiconductors, a class of materials valued because they could be used as diodes (a diode is a one-way valve for electricity). One was Russell Ohl of Bell Telephone Laboratories. Ohl and his fellow researchers discovered that putting two slightly different types of a semiconductor called germanium together produced a device that acted like a electron tube diode.

A replica of the point-contact transistor created by John Bardeen and Walter Brattain, under the supervision of William Shockley in 1947.

Ohl’s work was important, but an even bigger discovery was made in 1947 when John Bardeen and Walter Brattain stumbled on the “transistor,” a slice of germanium with a few carefully placed wires touching it, that was not only a valve but also an amplifier. This was the point-contact transistor. As an added bonus, the transistor produced a fraction of the waste heat and was tiny compared to an amplifier tube—the whole device could fit on the end of a finger. Not long afterwards William Shockley, also of Bell Labs, made the fragile transistor into a rugged and practical device when he invented the “junction” transistor, a sandwich made up of layers of germanium. Bell Labs announced the point-contact transistor in 1948 and the junction transistor in 1951. The germanium transistor was a milestone, but it was unreliable and engineers sought out new materials with which to construct transistors.

The first commercially produced silicon transistor, developed by Texas Instruments in the early 1950s.

They found an answer in silicon, another semiconductor that had been used in diodes. Silicon proved to be a better material for making transistors. It was this type of transistor, introduced by Texas Instruments in 1954, that revolutionized the technological world. Missiles became more accurate with onboard transistor guidance systems and computers became small enough to fit on board an aircraft. Perhaps the most famous transistorized product from this era was the pocketsize radio. By the end of the 1950s, the little transistor had replaced the hot, unreliable electron tube in nearly every existing type of electronic system. It also made electronic devices smaller, cooler (in temperature, that is), and less expensive. But engineers were not satisfied—they wanted to make things even smaller.

Transistors Launch the Computer Revolution

If you ask someone who lived during the late 1950s or 1960s what they associated with the transistor, there is a good chance they’ll say “transistor radio.” And with good reason. The transistor radio revolutionized the way people listened to music, because it made radios smaller and portable. But, nice as a hand-held radio is, the real transistor revolution was taking place in the field of computers.

In a computer the transistor is usually used as a switch rather than an amplifier. Thousands and later tens of thousands of these switches were needed to make up the complicated logic circuits that allowed computers to compute. Unlike the earlier electron tubes (often called vacuum tubes), transistors allowed the design of much smaller, more reliable computers—they also addressed the seemingly insatiable need for speed.

The speed at which a computer can perform calculations depends heavily on the speed at which transistors can switch from “on” to “off.” In other words, the faster the transistors, the faster the computer. Researchers found that making transistors switch faster required that the transistors themselves be smaller and smaller, because of the way electrons move around in semiconductors—if there is less material to move through, the electrons can move faster. By the 1970s, mass-production techniques allowed nearly microscopic transistors to be produced by the thousands on round silicon wafers. These were cut up into individual pieces and mounted inside a package for easier handling and assembly. The packaged, individual transistors were then wired into circuits along with other components such as resistors and capacitors.

As computers were produced in larger numbers, some kinds of logic circuits became fairly standardized. Engineers reasoned that standard circuits should be designed as units, in order to make them more compact. There were many proposals for doing this, but British engineer G. W. A. Dummer proposed the idea of making the entire circuit directly on a silicon wafer, instead of assembling the circuits from individual transistors and other components. Two engineers, Jack Kilby at Texas Instruments and Robert Noyce at Fairchild Semiconductor, invented such circuits—called integrated circuits—in 1958.

This device, developed by Robert Noyce in the late 1950s, was the first commercially available integrated circuit.

The first integrated circuits (ICs) were very simple and merely demonstrated the concept. But the idea of fabricating an entire circuit on a silicon wafer or “chip” with one process was a real breakthrough. Integrated circuits were so expensive that the first ones were purchased only by the military, which could justify the cost for top-notch performance. A little later, however, the integrated circuit would be mass-produced (largely to meet the needs of NASA’s Apollo program and the United States’ missile programs). When this happened it would revolutionize the design of computers.

Chips, Anyone?

Integrated circuits (ICs) seem to be nearly everywhere—they’re in places such as your car’s engine and your car’s radio, telephones, iPods, and home thermostats; they’re in virtually all the technologies you interact with every day from ATMs to X-ray machines. And, of course, they’re in computers. Computers were one of the first places where ICs took hold, and they remain among the most recognizable technologies equipped with ICs.

Despite their increasingly small size, computers are extremely complicated technological systems. Inside a computer are a whole range of different chips that do everything from regulating power supplies and internal temperatures, to running sound and video systems, to controlling the spinning of hard drives and DVD burners. The most familiar chips are memory chips and microprocessors.
Memory chips store information, such as programs and data. The “main” memory chips that you see advertised are usually for storage of program data. These chips lose their data when power to the computer is turned off. Other memory chips store data permanently or until you change it, and there is some memory built into microprocessors and other types of chips.

The first Mosfet transistor, designed by M. M. Atalla, D. Kahng, and E. Labate in late 1959.

Inside a typical main memory chip are tens of thousands or even millions of transistors—often in the form of a transistor called the metal oxide semiconductor or MOS, a device that was invented by Dawon Kahng and M. M. Attala. MOS transistors store information by switching on or off. In every computer, every piece of data is translated into a binary “code” of 0s and 1s. The letter “A” for example is translated into a binary number, 01000001. Then 01000001 is represented inside the chip as a set of transistors switched on (1) or off (0). A program like a web browser that deals with large amounts of text, displays pictures, accepts input from the user, and communicates with other computers needs millions of transistors to store all the coded information that passes through.

The Intel 4004 microprocessor, which was introduced in 1971. The 4004 contained only 2300 transistors and performed 60,000 calculations per second.

The microprocessor is another famous chip that resides in every computer. Unlike a memory chip, the microprocessor has many different functions, all carried out on one chip. Early computers had separate units (sometimes housed in different cabinets) for their mathematical and logic units, synchronization circuits or “clocks,” register units where various logic operations take place, buffers where data is held, circuits to accept data from the outside world, and so on. To make computers smaller, more energy efficient, and to move data around inside them more quickly, engineers began “integrating” those separate units onto one or more chips, then integrating those chips into a single “microprocessor,” or, in cases where engineers wanted to put a tiny computer into an industrial machine, a “microcontroller.” Gary Boone and others at Texas Instruments, and Federico Faggin, Stanley Mazor, Tedd Hoff and others at Intel Corporation developed the first microprocessors and microcontrollers.

Intel's Pentium 4 contains tens of millions of transistors.

A chip is more than just a home for transistors. It also contains other elements needed to make a circuit, such as resistors, capacitors, and interconnecting conductors. But the usual way of comparing chips is to discuss the number of transistors on them. The first integrated circuits invented in 1958 had just a few transistors. The latest microprocessors have over 40 million. Intel executive Gordon Moore was the first to observe this growth and the increase in numbers is often known as Moore’s Law.

To pack so many transistors and circuit elements onto one chip engineers have had to shrink the size of the parts. These smaller parts are, in fact, one of the major reasons for innovation in the integrated circuit field. The transistors that were about a centimeter wide in 1959 are now less than 200 billionths of a meter wide. That is so small that engineers are already predicting that the next generation of chips will have to be constructed in entirely new ways, perhaps assembled from individual molecules. This exciting new field is called “nanotechnology,” and it may open up entirely new directions for electronics in the 21st century.

Nanotechnology

With the integrated circuit growing smaller and smaller over the last decades, one might wonder, can they get any tinier? Engineers working in the field of nanotechnology believe they can and will. Nanotechnology refers to any new technology—a transistor, a tiny machine, a chemical—that is put together atom-by-atom or molecule-by-molecule. It usually also refers to the size of these technologies, which is defined as being 100 nanometers or less. A nanometer is one billionth of a meter. By comparison, today the smallest transistors on an IC are about 200 nanometers in size.
Renowned physicist Richard Feynman introduced the basic idea for nanotechnology in a 1959 speech called “There’s Plenty of Room at the Bottom.” Feynman predicted that tiny assembly machines made from a few molecules of matter could be built, and that these assemblers would be used to make other microscopic products. The result would be a system of production that would revolutionize the way things are made.

Micromachining involves the creation of microscopic mechanical devices, like that shown here. The legs belong to a spider mite placed to demonstrate scale.

In the 1990s “micromachining” emerged as one of the first practical approaches to creating nanotechnologies. Using etching techniques pioneered in the field of integrated circuits, engineers began building microscopic machines with tiny gears, levers, and rotors. While most of these were simply demonstrations that such things could be built, engineers believed that these machines would soon be used in practical systems, such as microscopic, implantable, or injectable pumps to deliver drugs inside the body. Because of its relatively large scale, not everyone today agrees micromachining should still be part of the nanotechnology field, but it did spawn the important field of micro-electro-mechanical systems (MEMs). MEMs are currently used with integrated circuits, where tiny machines are combined with electronics on a silicon chip.

The connection between nanotechnology and electronics grew stronger when chip designers began to approach the limits of the miniaturization by conventional techniques. In the mid-1960s “Moore’s Law” predicted that the size of features on integrated circuits would shrink dramatically over time and, in fact, transistors and other chip components shrank rapidly over the next four decades. But the photolithographic etching processes used to make transistors on an IC impose physical limits on the size of the transistors.

Nanotubes, which are made in a flask by a chemical process.

Many engineers and scientists are currently working on new, nanotechnological solutions to this problem, using tools such as the atomic force microscope (ATF) to build functional transistors from just a few atoms. They hope to find ways to build entire integrated circuits “from the bottom up,” by assembling them from atoms, rather than using today’s “top down” methods. Recently, nanoscale transistors have been demonstrated using materials called nanotubes, which are custom-made variations on a complex carbon molecule called a buckyball. If nanotechnology is the wave of the future, what is it doing for us today? Chemists have introduced new materials such as improved plastics that are stronger and better than earlier types of plastics. Another exciting area of progress is in quantum dots, which are microscopic crystals of semiconducting material that emit light when they are exposed to strong ultraviolet light. These dots can be used to detect cancer cells, and may soon be used to illuminate living spaces.

Organic LEDs consist of layers of organic thin films sandwiched between two conductors. When an electric current is applied, bright, visible light is emitted. The devices are lightweight, durable, flexible, power efficient, and hence ideal for portable applications. Here, a prototype flexible organic LED from Universal Display flashes the corporate logo.

In electronics, nanotechnology is making an impact in cell phone and computer displays, where organic LEDs (OLEDs) are in production utilizing nano-engineered thin-film layers. Most computer hard discs are also made using a combination of a nano-engineered recording medium and a sensitive type of recording head made of giant magnetoresistive (GMR) materials. Filters using nanoparticles are capable of removing bacteria and viruses from drinking water in addition to larger particles. If nothing else, nanotechnology has helped us cut down on our dry cleaning bill: In 2003 the clothing store The Gap began selling trousers impregnated with a new stain resistant chemical developed through nano-engineering.

Other researchers are focusing their efforts on studying the way nanotechnologies will work, because at the nano-scale, the normal rules about the behavior of electrons, photons, and matter have to be thrown out. In fact, computer designers anticipate that future computers based on nanotechnology may eliminate transistors altogether. Another line of research is aimed at using DNA—the same material our bodies use to store genetic information. This would require the construction of custom DNA molecules and a way to get information in and out of the “computer” which might take the form of a flask of millions of molecules, suspended in a liquid. Nanotechnology is so new, and so little understood, that it is difficult to predict how it will develop. Many engineers, however, believe that it holds the key to the next generation of electronic devices, which will demand faster computational speeds and pack more components into smaller spaces than has been possible before.



http://www.ieee-virtual-museum.org/e...70&lid=1&seq=1

[Muhammad Irfanullah]

Thnx for such a useful & informtive post.you did capture all the 5 Computer's Generations in one page in such a easy way that even the begginers,who do not know much about it, can understand the speedy revolutions of the computer technology.