But the second generation never used vacuum tubes. In 1947, the invention of the transistor-a thumbnail-sized sandwich of solid material which performed all the functions of a vacuum tube-ushered in an era of “solid state” electronic devices which drew little power, generated little heat, and were smaller and more reliable than the tubes they replaced. Silicon technology provided the basis for three generations of increasingly compact, reliable, and cheap computers over the next twenty years.
But by the 1970s, computer designers began to confront the inherent limitations of silicon technology. Although circuits had been shrunk to microscopic dimensions, computation speed was still dependent on circuit length. To miniaturize circuits still more, where distances were already on the order of millionths of an inch, brought back an old problem: heat. Smaller circuits would literally melt from the heat produced. What was needed was some method to eliminate heat and reduce resistance at the same time.
It had been known since the 1950s that many metals when cooled to extremely low temperatures became “superconducting,” permitting the unimpeded flow of electrons through them. In 1977, IBM announced it was designing an ultra-high-speed computer the size of a grapefruit, chilled with liquid nitrogen. The superconducting computer required a radically new technology, and a new range of low temperature construction materials.
Doped diamonds would be used extensively throughout.
Several days later, the ERTS canteen came up with an alternative explanation. According to the new theory, the 1970s had been a decade of unprecedented growth in computers. Although the first computer manufacturers in the 1940s had predicted that four computers would do the computing work of the entire world for the foreseeable future, experts anticipated that by 1990 there would actually be one billion computers-most of them linked by communications networks to other computers. Such networks didn’t exist, and might even be theoretically impossible. (A 1975 study by the Hanover Institute concluded there was insufficient metal in the earth’s crust to construct the necessary computer transmission lines.)
According to Harvey Rumbaugh, the 1980s would be characterized by a critical shortage of computer data transmission systems: “Just as the fossil fuel shortage took the industrialized world by surprise in the 1970s, so will the data transmission shortage take the world by surprise in the next ten years. People were denied movement in the 1970s; but they will be denied information in the 1980s, and it remains to be seen which shortage will prove more frustrating.”
Laser light represented the only hope for handling these massive data requirements, since laser channels carried twenty thousand times the information of an ordinary metal coaxial trunk line. Laser transmission demanded whole new technologies-including thin-spun fiber optics, and doped semiconducting diamonds, which Rumbaugh predicted would be “more valuable than oil” in the coming years.
Even further, Rumbaugh anticipated that within ten years electricity itself would become obsolete. Future computers would utilize only light circuits, and interface with light transmission data systems. The reason was speed. “Light,” Rumbaugh said, “moves at the speed of light. Electricity doesn’t. We are living in the final years of microelectronic technology.”
Certainly microelectronics did not look like a moribund technology. In 1979, microelectronics was a major industry throughout the industrialized world, accounting for eighty billion dollars annually in the United States alone; six of the top twenty corporations in the Fortune 500 were deeply involved in microelectronics. These companies had a history of extraordinary competition and advance, over a period of less than thirty years.
In 1958, a manufacturer could fit 10 electronic components onto a single silicon chip. By 1970, it was possible to fit 100 units onto a chip of the same size-a tenfold increase in slightly more than a decade.
But by 1972, it was possible to fit 1,000 units on a chip, and by 1974, 10,000 units. It was expected that by 1980, there would be one million units on a single chip the size of a thumbnail, but, using electronic photo projection, this goal was actually realized in 1978. By the spring of 1979, the new goal was ten million units-or, even better, one billion units- on a single silicon chip by 1980. But nobody expected to wait past June or July of 1979 for this development.
Such advances within an industry are unprecedented. Comparison to older manufacturing technologies makes this clear. Detroit was content to make trivial product design changes at three-year intervals, but the electronics industry routinely expected order of magnitude advances in the same time. (To keep pace, Detroit would have had to increase automobile gas mileage from 8 miles per gallon in 1970 to 80,000,000 miles per gallon in 1979. Instead, Detroit went from 8 to 16 miles per gallon during that time, further evidence of the coming demise of the automotive industry as the center of the American economy.)
In such a competitive market, everyone worried about foreign powers, particularly Japan, which since 1973 had maintained a Japanese Cultural Exchange in San Jose-which some considered a cover organization for well-financed industrial espionage.
The Blue Contract could only be understood in the light of an industry making major advances every few months. Travis had said that the Blue Contract was “the biggest thing we’ll see in the next ten years. Whoever finds those diamonds has a jump on the technology for at least five years. Five years. Do you know what that means?”
Ross knew what it meant. In an industry where competitive edges were measured in months, companies had made fortunes by beating competitors by a matter of weeks with some new techniques or device; Syntel in California had been the first to make a 256K memory chip while everyone else was still making 16K chips and dreaming of 64K chips. Syntel kept their advantage for only sixteen weeks, but realized a profit of more than a hundred and thirty million dollars.
“And we’re talking about five years,” Travis said. “That’s an advantage measured in billions of dollars, maybe tens of billions of dollars. If we can get to those diamonds.”
These were the reasons for the extraordinary pressure Ross felt as she continued to work with the computer. At the age of twenty-four, she was team leader in a high-technology race involving a half-dozen nations around the globe, all secretly pitting their business and industrial resources against one another.
The stakes made any conventional race seem ludicrous. Travis told her before she left, “Don’t be afraid when the pressure makes you crazy. You have billions of dollars riding on your shoulders. Just do the best you can.”
Doing the best she could, she managed to reduce the expedition timeline by another three hours and thirty-seven minutes-but they were still slightly behind the consortium projection. Not too far to make up the time, especially with Munro’s cold-blooded shortcuts, but nevertheless behind- which could mean total disaster in a winner-take-all race.
And then she received bad news.
The screen printed PIGGYBACK SLURP / ALL BETS OFF.
“Hell,” Ross said. She felt suddenly tired. Because if there really had been a piggyback slurp, their chances of winning the race were vanishing-before any of them had even set foot in the rain forests of central Africa.