Chapter 4 - The Competitive Limit of Soft Technology

Tailgaters seldom gain permanent advantage from technological superiority alone. The incumbent simply has too many other strengths. Dr. Gene’s eponymously named Amdahl Corporation initially hoped to leapfrog IBM with advanced semiconductor components. But IBM countered, first with an unforeseen product upgrade, then with denser logic chips, and finally with a radical shift in chip technology. Like a self-righting punching bag, Amdahl repeatedly bounced back into the lead but was ultimately crushed under the enormous power of IBM’s superior research, business scale, and market positioning.

Dr. Gene’s first plug-compatible mainframe was targeted at IBM’s 370/165 (on which he’d done some early work) but was whacked by the unexpected announcement of the 370/168 and, therefore, dead on arrival.

Many years later, Amdahl recalled how he’d been blindsided: Lacking “the virtual memory [of the 370/168] ... we dropped the first model and put aside any expectations of being able to ship a product for nearly a year and a half.”1 When the product delays proved ongoing, Amdahl’s investors—including Fujitsu, the German minicomputer maker Nixdorf, and Edgar Heiser, a Chicago-based venture capitalist— got testy. And who could blame them with nothing to show for the $27 million they’d invested to date? "Tension in the boardroom reached a peak in early 1974, when Amdahl’s enterprise almost went under," revealed Fortune magazine in September 1977. "Chief among the problems was a collapse in management."

The board forced Amdahl out as CEO and pushed his two founding colleagues out the door altogether. Five vice presidents subsequently departed after the board enlisted the scrappy, white-haired Eugene R. White to run the company. White, a former GE and Fairchild Semiconductor executive, brought the executive presence Gene Amdahl lacked. And both Genes deserved equal credit for the early company’s success.

On arrival, White found the company without management, product, customers, money, or morale. The most immediate tasks were to reorient the design team toward the 370/168 and to resolve the funding shortage. For that, the two Genes approached Fujitsu executive Toshio Ikeda, a.k.a. “Mr. Computer of Japan,” and came away with $35 million on top of the $11 million Fujitsu had previously invested, in return for a 41 percent equity stake and access to Amdahl’s intellectual property. The agreement also called for the Japanese giant to manufacture ten mainframes based on an Amdahl-built prototype in exchange for exclusive rights in some markets.

When the brilliant and somewhat eccentric Ikeda (he reputedly kept a computer team awake during one grueling night by belting out Schubert’s Winterreise song cycle in German) suffered a fatal fall at Narita Airport a short while later, Fujitsu’s relationship with Amdahl took on a much harsher tone. Amdahl CFO Cliff Madden, who remembered Ikeda as “very much the rising star,” said that “his successor was well regarded, but didn’t have nearly the same stature. The circumstances gave other powerful factions within the company, such as international marketing, a bigger voice,” he told us.

In the early years of the partnership, any animosity was likely cloaked by high hopes for Amdahl’s success. Fujitsu manufactured the first boxes to help Amdahl conserve capital, and its engineers worked “shoulder to shoulder in Sunnyvale [California] throughout the development of the Model 470/V6,” Amdahl’s first successful product, related Gene Amdahl in a December 1975 Infosystems interview. The first six processors were installed at the National Aeronautics and Space Administration (NASA), Computer Usage Corporation, and four universities. The seventh spelled commercial credibility when Dave Blackwell, the doughty chief information officer of the Massachusetts Mutual Life Insurance Company, took the career bet of persuading top management to buy from the upstart Amdahl, even as a near-hysterical IBM dispatched fleets of blue helicopters to land on Mass Mutual’s lawn, dispensing a mix of dire threats and halcyon promises. Amdahl had secured a beachhead in commercial data processing.

At our first research visit to Amdahl in March 1976, the tough talking Gene White boasted of twelve letters of intent to purchase, twenty excellent prospects, and 121 that he categorized as “Class A.” Since half the letters of intent came from government agencies, which were required to accept the lowest bid from a group of competitors, we viewed White’s enthusiasm with a jaundiced eye. But we were wrong. Business boomed, and the future looked rosy, with twenty-seven customer installations and a fifty-million-dollar public offering in the wings. “Amdahl in ‘76: from oblivion to a comer,” gushed a Datamation headline in February 1977.