The History of the IC (Integrated Circuit): From One Transistor to Billions

You probably carry dozens of chips every day without noticing: in your phone, your watch, your car and most home appliances. These chips are integrated circuits (ICs). An IC puts transistors, resistors, capacitors and the wiring between them onto one small piece of semiconductor material. That idea changed electronics completely.

Before the IC: the problem with discrete circuits

In the mid-twentieth century, electronic circuits were built from separate parts such as vacuum tubes, resistors and capacitors, soldered together by hand. The early computer ENIAC used tens of thousands of vacuum tubes and filled a large room. Tubes ran hot, broke often and took up space, so making a circuit more complex was difficult.

The transistor, invented at Bell Labs in 1947, reduced size and power use. But the old problem remained. When a circuit has thousands of parts, every hand-made connection is a chance for failure. Engineers called this the tyranny of numbers: the more parts you need, the more connections you must make, and the number of connections grows very quickly.

1958: Jack Kilby and the first germanium chip

Jack Kilby, an engineer at Texas Instruments, asked a simple question: why not make the transistor, resistor and capacitor from the same piece of material? In September 1958, while most of the company was on holiday, he demonstrated a working circuit built on germanium. Thin gold wires connected the parts on a small bar. It looked crude compared with modern chips, but it proved that a whole circuit could live on one piece of material. The circuit produced an oscillation, which showed that it really worked.

Kilby shared the 2000 Nobel Prize in Physics for this work, together with two other researchers in semiconductor devices.

1959: Robert Noyce and the planar silicon IC

A few months later, Robert Noyce at Fairchild Semiconductor developed a more practical version. He used silicon instead of germanium and the planar process, which grows a protective oxide layer on the surface of the silicon. Patterns were printed with photography and etched with acid. The connections were thin metal films on the surface, so there were no hand-attached gold wires. This method was more reliable, easier to repeat in production, and became the foundation of chip manufacturing.

The 1960s: ICs enter the market and the military

The first big buyers were space and defence projects. The US Minuteman missile needed a guidance computer that was small, light and reliable, and ICs fitted that need well. Manufacturers sold many chips to government agencies in the early 1960s. The price per chip was still high, but rising production volumes brought it down quickly.

In 1965, Gordon Moore of Fairchild wrote an article predicting that the number of transistors on a chip would double roughly every one to two years. This prediction became known as Moore's Law. It gave the industry a target and drove engineers to keep finding new improvements for several decades.

The 1970s: microprocessors and MOS technology

In 1971, Intel released the Intel 4004, the first microprocessor that put a complete processing unit on one chip. It had about 2,300 transistors and ran slowly by today's standards, but it opened the door to small computers in calculators, controllers and later personal computers.

At the same time, MOS (metal-oxide-semiconductor) technology became practical and used less power than earlier designs. The number of transistors per chip grew faster, and medium-scale and large-scale integration (MSI and LSI) became normal.

The 1980s and 1990s: VLSI and the personal computer

Chips entered the VLSI (very large scale integration) era, with hundreds of thousands and then millions of transistors on one chip. The Intel 386, introduced in 1985, had about 275,000 transistors. The demand for personal computers made the chip market much larger, and competition pushed the feature sizes down from several micrometres to below one micrometre.

CMOS became the main technology in this period because it used less power. This mattered for computers, and later for mobile phones and other portable devices.

The 21st century: billions of transistors

Since about 2000, the main problems for the industry have been heat, leakage current and manufacturing cost, not just smaller size. Engineers moved to three-dimensional transistor structures such as FinFETs, stacked chips and chiplets, which are smaller blocks combined into one package.

Modern chips contain tens of billions of transistors. Many companies design their own chips for phones, computers and servers. These chips make it possible to process images, sound and artificial intelligence inside small devices.

Lessons from the history of the IC

Conclusion

From a germanium circuit with a few gold wires to silicon chips with tens of billions of transistors, the history of the IC is the story of making complex things smaller, cheaper and more reliable. Every time you use a phone or a calculator, you use the result of that idea.