1958: New hire Jack Kilby shows his Texas Instruments colleagues a little something he’s built. A very little something: a working integrated circuit on a piece of semiconductor material. The world will soon change.

Electronics had relied on vacuum tubes for half a century before Bell Labs invented the transistor in 1947. Transistors were tinier, more reliable, longer-lasting, cooler and more energy-efficient. But connecting hundreds or thousands of them in a complex circuit required wire and solder. That cost money, took time and created thousands of ways for the circuit to fail.

Texas Instruments, or TI to insiders, was working on the U.S. Army Signal Corps’ Micro-Module program when Kilby joined the firm in 1958. Micro-Modules proposed to make all components the same size, so they could be snapped together to create circuits without wire or solder. Most of the company’s employees went on a two-week vacation in July, but Kilby hadn’t earned any vacation time yet.

He used his solitude to good effect.

“Further thought led me to the conclusion that semiconductors were all that were really required,” Kilby later wrote. “[R]esistors and capacitors [passive devices], in particular, could be made from the same material as the active devices [transistors]. I also realized that, since all of the components could be made of a single material, they could also be made in situ interconnected to form a complete circuit.”

See Also:

Photo Gallery
Faster, Smaller, Cheaper

Photo Gallery
Once-Mighty Bell Labs Leaves Behind Transistor, Laser, 6 Nobels

Kilby constructed a prototype of the integrated circuit by September. It was a sliver (a chip, you might say) of germanium with wires sticking out, glued to a glass slide about the size of a thumbnail.

The stakes were high for the new guy. Among those assembled for the Sept. 12 demonstration were then-chairman Mark Shepherd and other execs.

Kilby connected his device to an oscilloscope and threw the switch. There on the screen, a continuous sine curve pulsed, and a new era began.

As recounted many times in This Day in Tech and elsewhere, there’s often an unheralded precursor or prior claim. In Kilby’s case, that would be British radar scientist Geoffrey W.A. Dummer, who presented the concept of a miniaturized, integrated circuit at a 1952 electronics symposium in Washington, D.C.

He wanted to put an entire circuit on a piece of silicon just half-an-inch square. But his prototype failed, the Ministry of Defense was unimpressed, and the idea died on the organizational vine.

And as is often the case — again, oft recounted here — scientific and technological advances frequently occur with nearly simultaneous independent discovery or invention. In this case that would be Fairchild Semiconductor engineer Robert Noyce, who was working on an integrated circuit using silicon instead of germanium.

Kilby and TI were first to file for a patent for “miniaturized electronic circuits” in February of 1959. Noyce and Fairchild filed their application for a silicon-based integrated circuit six weeks later in April. It was granted in 1961, and TI didn’t get its patent until 1964.

Fairchild and TI engaged in a lengthy legal battle before agreeing to cross-license their technologies. Noyce’s silicon chip eventually triumphed over Kilby’s germanium. Noyce went on to co-found Intel with Gordon Moore.

Kilby went on to share the 2000 Nobel Prize in Physics. The Nobel website acknowledges that “Kilby and Noyce are considered to be co-inventors of the integrated circuit.” However, Noyce died in 1990, and Nobel rules prohibit granting the prize posthumously. Earth to earth, ashes to ashes, dust to dust, germanium to silicon.

Kilby died in 2005 in a world where microchips permeate every aspect of our daily life, from the inner space of our bodies to the outer space of the cosmos, at home, at play and on the job, in our cars, in our ears … indispensable.

Source: Texas Instruments, Today in Science

Photo: Jack Kilby’s pioneering integrated circuit was thumbnail size. (Courtesy Texas Instruments/Corbis)

This article first appeared on Wired.com Sept. 12, 2008.