sciencebriefs
13:00in productionCh. 1 · The tyranny of numbers/ 13:00 · ceiling 15 min
Computing & AI · Engineering

Integrated circuit

1958

Two engineers solved the same wiring problem months apart in 1958 and 1959, and it was the less-famous first version, not the one modern chips descend from, that reached the Nobel committee.

Jack Kilby demonstrated the first working integrated circuit at Texas Instruments on 12 September 1958, a germanium device that put several components on one piece of semiconductor but still needed fine gold wires connecting them by hand, a method too delicate for mass production. Robert Noyce at Fairchild Semiconductor solved the same problem differently a few months later, using silicon and a planar process that let the connections be built into the chip itself, and it is Noyce's approach that modern chips actually descend from. Both men are credited as co-inventors, but only Kilby received a Nobel Prize, awarded in 2000, since Noyce had died a decade earlier and the prize is not given posthumously. The underlying idea, packing a circuit onto one piece of material instead of wiring separate parts together, is what let electronics keep shrinking for decades afterward.

Chapters & takeaways6
  1. 0:08
    The tyranny of numbers

    Engineers of the 1950s faced a wiring problem: more components meant more hand-soldered connections and more ways for a circuit to fail.

  2. 2:10
    Kilby's germanium demonstration

    On 12 September 1958, Kilby showed Texas Instruments management a working circuit built on one piece of germanium, still wired together with fine gold leads.

  3. 4:20
    Noyce's silicon answer

    Months later, Robert Noyce at Fairchild used a planar silicon process that built the interconnections into the chip itself, without external wiring.

  4. 6:30
    Two inventors, one credit

    Kilby and Noyce are both recognised as co-inventors of the integrated circuit, arrived at independently rather than in collaboration.

  5. 8:40
    A prize that came late

    Kilby received the Nobel Prize in Physics in 2000; Noyce, who died in 1990, could not share it because the prize excludes the deceased.

  6. 10:50
    From one chip to billions of transistors

    Noyce's manufacturable approach became the basis for the steady miniaturisation that later became known as Moore's law.

Worth your time?

Yes. Study the whole thing.

4/ 5
What works
  • it draws a clear, specific line between Kilby's proof of concept and Noyce's manufacturable design
  • the Nobel Prize timing is explained honestly rather than glossed over
  • it treats both inventors as legitimate rather than picking a winner
What does not
  • the planar process that made Noyce's version practical is described only in outline
  • the patent dispute between the two companies gets little space given how consequential it was
Study it if
  • anyone who wants the real story behind the chip in every device they own
  • readers interested in how simultaneous invention gets sorted into credit and prizes
  • people curious why one version of an idea wins out commercially over another
Skip it if
  • readers wanting a manufacturing-level explanation of photolithography
  • anyone mainly interested in Moore's law itself rather than its origin
The written brief4 min read

The tyranny of numbers

By the late 1950s, electronic circuits had grown complicated enough that simply wiring together individual transistors, resistors, and capacitors by hand had become a real bottleneck, a problem engineers of the time called the tyranny of numbers: every added component meant another soldered connection and another chance for the whole circuit to fail. Jack Kilby, newly hired at Texas Instruments, worked through the summer of 1958 toward a different approach, reasoning that if all the components were made from the same piece of semiconductor material rather than assembled from separate parts, the wiring problem would shrink along with the circuit. On 12 September 1958 he demonstrated a working version of this idea to company management, built on germanium, proving with an oscilloscope that a genuine circuit could be formed this way.

Kilby’s germanium demonstration

Kilby’s demonstration circuit worked, but it was not yet something that could be manufactured in volume, since the individual elements on the germanium piece still had to be connected to each other with fine gold wires attached by hand, a delicate and slow process. A few months later, working independently at Fairchild Semiconductor, Robert Noyce arrived at a different solution to the same underlying problem. Using silicon rather than germanium and a planar process that had recently been developed for building transistors, Noyce found a way to lay down the interconnections between components as part of the same manufacturing steps that built the components themselves, eliminating the hand-wired connections entirely and making the design suitable for mass production in a way Kilby’s original was not.

Noyce’s silicon answer

Both approaches are recognised, correctly, as genuine and independent solutions to the same problem, arrived at within roughly a year of each other by engineers who were not working together or aware of each other’s specific progress. Kilby’s germanium demonstration stands up as the first working proof that the underlying idea, one piece of material instead of many wired components, could actually function as a circuit. Noyce’s silicon planar version stands up as the design that could actually be built at scale, and it is this version, not Kilby’s original germanium prototype, that modern integrated circuits trace their lineage back to. The distinction between the two men’s contributions, first working demonstration versus first manufacturable design, holds up as the accurate way to describe what each achieved.

Two inventors, one credit

What does not hold up as cleanly is any simple story of a single moment of invention. The two companies pursued overlapping patents, and for years the commercial and legal picture around who owned rights to the integrated circuit was contested rather than settled, reflecting the fact that two teams had solved closely related problems in parallel rather than one clearly preceding the other in every respect. The Nobel Prize outcome further complicates any tidy narrative: Kilby received the award in 2000, sharing it with two other physicists for separate work on semiconductor heterostructures, while Noyce, who had died in 1990, was never eligible, since the prize is not awarded posthumously. That is an accident of timing and mortality, not a verdict on whose contribution mattered more.

A prize that came late

The significance of the integrated circuit lies less in either man’s specific device than in what the general idea made possible once Noyce’s manufacturable version took hold: circuits could now be shrunk and multiplied at a pace that hand-wired electronics never could have matched. That manufacturability is what allowed the steady doubling of components per chip that later became known as Moore’s law, and it is the reason a modern chip can carry billions of transistors on an area the size of a fingernail. Nearly every piece of electronics built since, computers, phones, cars, medical devices, depends on descendants of Noyce’s planar silicon approach rather than on Kilby’s original wired germanium prototype, even though it was Kilby’s demonstration that first proved the concept could work at all.

From one chip to billions of transistors

This is a good use of an hour for anyone who assumes chip history is a single clean invention story, because the reality, two engineers solving the same problem independently within months, with only one of them living long enough to collect the Nobel recognition, is more interesting and more honest. It rewards attention mainly in the middle stretch, where the real technical difference between Kilby’s and Noyce’s approaches gets explained, since that is where the actual reason modern chips look the way they do becomes clear. Readers who want the deeper manufacturing chemistry of photolithography, or the legal detail of the patent fight between Texas Instruments and Fairchild, will need to look elsewhere, but as an account of how a shared problem produced two different, unequally remembered answers, it holds together well.

Same field · Computing & AI4 of 46
Up next in Science

Keeling Curve

Charles Keeling · 11:43

The Keeling Curve did not discover climate change — it discovered that CO₂ is rising, seasonally and relentlessly, and that we can measure it without ambiguity.

11:43