60 years of integrated circuits
60 years of integrated circuits
An integrated circuit is a semiconductor wafer on which thousands or millions of tiny resistors, capacitors, and transistors are fabricated. Sometimes called a chip or microchip. The invention of the integrated circuit made technologies of the Information Age feasible. ICs are now used extensively in all walks of life, from cars to toasters to amusement park rides.
Integrated circuits (ICs) are self-contained circuits with many separate components such as transistors, diodes, resistors and capacitors etched into a tiny silicon chip.
Related Journals of Integrated circuit
Journal of Physical Chemistry & Biophysics, Journal of Electrical & Electronic Systems, Analog Integrated Circuits and Signal Processing, IEEE Radio Frequency Integrated Circuits Symposium, IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, Journal of Integrated Circuits and Systems, Proceedings of the Custom Integrated Circuits Conference.
On 12 September 1958, Jack Kilby of Texas Instruments demonstrated a working integrated circuit1. The circuit was a phase-shift oscillator that used transistor, resistor and capacitor elements built from a single piece of germanium; the elements were connected into a circuit with the help of thin gold wires. A few months later, Robert Noyce, working at Fairchild Semiconductor, proposed a monolithic integrated circuit2. This planar design was based on silicon and used lines of aluminium, deposited on the insulating silicon dioxide layer that can form on the surface of silicon wafers, to connect the different circuit elements in the single chip. By 1960, a team of engineers at Fairchild Semiconductor had turned this design into a reality. Electronic components, which had previously been discrete units connected with individual wires, could now be integrated into the same piece of semiconducting material.
In the 60 years since Kilby’s initial demonstration, progress in
STM8S207C8T6 has been astounding. Noyce would go on to co-found Intel, and just how far the company — and the design of integrated circuits — has come in that time is highlighted in this issue of Nature Electronics. In a News & Views article, Suman Datta of the University of Notre Dame reports on Intel’s 10-nanometre logic technology. With this latest design iteration3, the company has introduced a number of unconventional approaches to improve transistor density and performance, including a technique to reduce the spacing between cells and a method to add gate contacts directly over the active area. As a result, they can deliver around 100 million transistors per square millimetre — a transistor density that is 2.7 times higher than that of their previous 14-nm technology, which was introduced in 2014.
At this level of complexity, developments are far from straightforward. Earlier this year, it emerged that Intel have encountered problems in the manufacturing of the 10-nm chips, leading to delays in mass production4; the chips are now expected to ship in volume in 2019. And in the past few weeks, GlobalFoundries announced5 that they would stop development of their 7-nm chips (thought to be comparable to Intel’s 10-nm technology). The continued scaling of silicon complementary metal–oxide–semiconductor (CMOS) technology beyond these levels is also likely to prove increasingly difficult. But, at the same time, the applications of computers are evolving, and demand the processing of ever larger amounts of data. As a result, the search for strategies and materials beyond silicon, which could help create the next generation of devices and integrated circuits, remains vital.
Carbon nanotubes are among the contenders fighting for a place in the future of electronics, and in our Reverse Engineering column in this issue, Cees Dekker recounts how the first carbon nanotube transistor was built back in 1998. A related contender in this fight is two-dimensional materials, as well as the vertical stacks of different two-dimensional materials known as van der Waals heterostructures. These materials have been used to build a range of promising devices and some basic circuits — even a microprocessor6. The unique challenges involved in trying to build practical integrated circuits from two-dimensional materials are just starting to be addressed, but innovative ideas are emerging. For example, in an Article in this issue, Moon-Ho Jo and colleagues illustrate how a scanning light probe can be used to write monolithic integrated circuits
For ST on two-dimensional molybdenum ditelluride (MoTe2).
The researchers — who are based at the Institute for Basic Science in Pohang, Pohang University of Science and Technology, the Korea Institute of Materials Science, and Yonsei University — first pattern gold electrodes onto the MoTe2. Then, by shining the light probe (a visible laser) onto the electrodes, the semiconducting MoTe2 beneath can be converted from an n-type semiconductor to a p-type semiconductor. (With silicon CMOS technology, such doping is typically achieved using ion implantation.) The approach allows the two-dimensional material to be doped precisely and quickly, and Jo and colleagues use it to create arrays of bipolar junction transistors and circular diodes.
Integrated circuits are the basis of so much of modern technology and here at Nature Electronics we aim to also consider the wider social, ethical and legal issues that surround the implementation of such technology. To this end, this issue sees the start of our Books & Arts section. Here, Arlindo Oliveira of the Instituto Superior Técnico in Portugal reviews Hello World, a book by Hannah Fry on the roles — both good and bad — that algorithms play in everyday life. Then Christiana Varnava from our editorial team reviews The Future Starts Here, an exhibition at the Victoria and Albert Museum in London that brings together a collection of emerging technologies in order to explore the ways they could shape society in the years to come.
An integrated circuit consists of capacitors, resistors, transistors, and other metallic connections required for a complete electrical circuit. The most popular electrical circuits are the MOSFET circuits because the switching time can be easily reduced. In addition, the transistors switch faster if the devices are made smaller. However, it is important to keep in check the power dissipated by a transistor so that the amount of heat produced in an integrated circuit can be kept under control.
Due to the improvement of the technology in building integrated circuits, primarily due to the decrease in the individual devices as well as in the increase in the area of the circuit, there has been a rapid growth in the number of transistors on an integrated circuit since the first such circuit was fabricated in 1961 with only four transistors. At present, a typical integrated circuit
For TI has about 80 million transistors. The single most important criterion is to keep in check the enormous heat produced by such circuits.
The ICs or chips used in a PCB do various tasks, such as signal acquisition, transformation, processing, and transfer. Some of these chips (for example, an encryption or image compression chip) work on digital signals and are called digital ICs, whereas others work on analog or both types of signals, and called analog/mixed-signal (AMS) chips. Examples of the latter type include voltage regulators, power amplifiers, and signal converters. The ICs can also be classified based on their usage model and availability in the market. Application-specific integrated circuits (ASIC) represent a class of ICs, which contain customized functionalities, such as signal processing or security functions, and meet specific performance targets that are not readily available in the market. On the other hand, commercial off-the-shelf (COTS) ICs are the ones, which are already available in the market, often providing flexibility and programmability to support diverse system design needs. These products can be used out-of-the-box, but often needs to be configured for a target application. Examples of COTS components include field programmable gate arrays (
FPGA), microcontrollers/processors, and data converters. The distinction between ASIC and COTS is often subtle, and when a chip manufacturer decides to sell its ASICs into the market, they can become “off-the-shelf” to the original equipment manufacturers (OEMs), who build various computing systems using them.
To develop integrated circuits, we need to accelerate the development of bases in both north and south China, and begin establishing research and production consortiums spread over Shanghai, Wuxi, and Shaoxing to create the conditions for large-scale development during the Seventh Five-Year Program period. Before 1985, we need to focus on mastering the technology for industrialized mass production of small- and medium-scale integrated circuits and achieving a breakthrough in the technology for the industrialized mass production of medium- and low-grade, large-scale integrated circuits. We need to strive to solve the difficult problem of ensuring the quality of small- and medium-scale integrated circuits having anywhere from several tens up to a thousand components on a wafer produced on a large production line and do everything possible to increase yield and decrease the failure rate. Products made in accordance with the seven special standards1 should have tolerances of 10?7 m and consumer electronics should have tolerances of 10?6 m. In addition, one or two production lines should have the capacity to produce more than 10 million items per year and the whole country should be able to produce 50 million per year at a cost that enables them to sell at the same price that imported integrated circuits cost in the late 1970s. We need to attain stable batch production of circuits for memory, 4- and 8-bit microcomputers, computers, instruments and meters, communications equipment, electronic clocks and watches, and TV speakers, all of which have a degree of integration of around 10,000 components, and we need to produce a total of 3–5 million of them per year. At the same time, we need to actively carry out R&D on high-grade, large- and very-large-scale integrated circuits. We need to finalize the design of 16K single-supply NMOS dynamic memory cells and 4K CMOS static memory cells and put them into production, and build satisfactory prototypes of 16K NMOS static memory cells, 64K NMOS dynamic memory cells, 16-bit microprocessor circuits, and ultra-high speed GaAs circuits.The worldwide development of integrated circuits has become more rapid, and a new generation seems to emerge almost every 3 years. Science and technology have achieved astonishing progress over the 20th century, and the next century will bring even greater breakthroughs in the areas of physical science, information science, bioengineering, materials science, cosmology, and environmental science. While developing new and high technologies and the industries using them, it is necessary not only to do good R&D work but also to take care to build a complete set of effective mechanisms. If China wants to develop a technology or product, it must consider the market for it. Without the impetus of market demand, it is very difficult to develop. The country should consider the development of integrated circuits. One major problem is that a few foreign companies have a huge share of Chinese electronic products market. Integrated circuits in these electronic products find their way to China installed in import equipment. The country does not have a market for the integrated circuits they produce, so there is no way to mass produce them.