Top
Free Shipping On Orders Over $35!
Get Rewarded for Ordering Your Textbooks! Enroll Now
Customer Reviews Read Reviews
Write a Review
|
|||
|
|||
|
|||
|
|||
|
|||
|
|||
|
|||
|
|||
|
|||
|
|||
|
|||
|
|||
|
|||
|
|||
|
|||
|
The New copy of this book will include any supplemental materials advertised. Please check the title of the book to determine if it should include any access cards, study guides, lab manuals, CDs, etc.
The Used, Rental and eBook copies of this book are not guaranteed to include any supplemental materials. Typically, only the book itself is included. This is true even if the title states it includes any access cards, study guides, lab manuals, CDs, etc.
Chapter One
Beginnings
Like the musical genius of a Mozart or a Gershwin, entrepreneurial ability often manifests itself very early in life. As a teenager in New Delhi, India, Vinod Khosla already knew he wanted to start his own company. With a degree in electrical engineering from the Indian Institute of Technology, the intense Khosla came to the United States after he failed with a start-up in his homeland. He received a master's degree in biomedical engineering from Carnegie-Mellon University in Pittsburgh before trekking west to the Stanford University MBA program in Palo Alto, California. This was in the late 1970s, when the recently named Silicon Valley was becoming a mecca for entrepreneurs. Apple Computer's two founders had developed the first real personal computer in a garage, while Intel had just invented the microprocessor. Stanford and the nearby University of California at Berkeley provided the computer laboratories in which countless students honed the skills to make their dreams come true.
After Khosla graduated from Stanford in 1980, he wrote 400 letters to companies that met his criteria: less than three years old, fewer than a hundred employees. He was already demonstrating those qualities of near fanatical determination and attention to detail that would pave the way for his triumph and later his downfall. None of the 400 letters paid off, but a connection at Stanford proved fruitful and, in early 1980, he helped start Daisy Systems, which specialized in computer-aided design (CAD). While Daisy eventually foundered after a promising start, the year-and-a-half he spent there gave Khosla the impetus to found Sun. At Daisy, he was searching for a better computer to handle CAD. The prevailing model required that engineers take turns, or share time, on a larger minicomputer. Few people thought it was feasible to give every engineer a computer with enough power to do CAD or CAM (computer-aided manufacturing), but Khosla recognized that the solution might lie in the new workstation, an individual computer that was relatively inexpensive, yet able to handle complex tasks previously done by minicomputers. Khosla also wanted the workstation to plug into a network of the type he had seen running at Stanford, called the Ethernet, which would allow engineers to collaborate electronically. This was a revolutionary concept at the time.
In 1981, came one of those fortuitous meetings that are the stuff of Silicon Valley legends: a secretary connected Khosla with a graduate student who was working on a project called the Stanford University Network, or SUN. This was fellow immigrant Andy Bechtolsheim, a tall, blond, introverted German, whose appearance and manner contrasted starkly with Khosla's. They spoke to each other in differently accented English, but Khosla grasped that Bechtolsheim had the technology he needed to make his vision of a workstation reality. Bechtolsheim had more modest ambitions: he was licensing his design to local companies for $10,000 apiece. That seemed like a great business while he finished his Ph.D. But Khosla told him bluntly, "I want the goose that laid the golden egg."
Bechtolsheim never did get his Ph.D. He and Khosla wrote a terse six-page business plan outlining how they would build and sell the SUN workstation. Dated February 12, 1982, the plan listed more than a dozen competitors, ranging from minicomputer manufacturers like Digital Equipment, Data General, and Hewlett-Packard, to personal computer vendors like Apple and Tandy. Workstations, priced then from $25,000 to $100,000, were smaller, cheaper, and more flexible than minicomputers, but more powerful than PCs. Another start-up, Apollo Computer of Chelmsford, Massachusetts, founded just two years earlier, had already demonstrated there was a market for workstations. (Microsoft, at the time an obscure vendor of PC software up in Redmond, Washington, was not even on Sun's radar screen, though it had already concluded a crucial deal to provide IBM with an operating system for its PC. For the rest of the decade, Sun and Microsoft would grow into hugely successful companies on parallel tracks, remotely aware of each other's existence.)
Sun was grounded in the notion of open systems and networking. Apollo, its chief competitor, had developed an elegant hardware-software combination, but both of those elements were proprietary. That is, they were unique to Apollo and totally under its control. By contrast, Sun intended to use standardized components. For an operating system, it would use Unix, which was developed by AT&T and widely licensed to universities, research labs, and companies. The Sun hardware was based on semiconductors from Motorola, available to anyone who wanted to buy them. The value added came in how Bechtolsheim put these chips together in the guts of the Sun workstation. Using off-the-shelf components, rather than building from scratch, enabled Sun to price its machines at less than half the cost of the competition. According to the business plan, Sun's machines would carry an end-user price of $10,000 to $20,000, whereas Apollo's cheapest machine sold for $25,000.
On the networking side, the Ethernet link would allow potential users to exchange electronic messages and share services such as file storage and printing. "To me the vision was distributed computing," Khosla says. Because the Ethernet allowed users to share information, "you didn't need all the computing power in the same place. Technology had changed the economics of the computer." Though open systems and networking are trumpeted today by almost every computer manufacturer, the concepts were new in the early 1980s. Leading computer manufacturers like IBM and Hewlett-Packard each offered separate, proprietary technology that did not interoperate with each other. Customers were forced to make a choice and stick with it, unless they wanted to scrap all those expensive machines and start over. Sun's technology was quicker to put together, much like the Ford assembly line replaced handcrafted automobile manufacturing, and thus it was more responsive to market changes.
SHOW ME THE MONEY
Before Khosla and Bechtolsheim could start their company, they needed funding. In the early 1980s, the elaborate structure of the venture capital community in Silicon Valley had not yet fully developed. Today, aspiring entrepreneurs must have a sophisticated, multifaceted business plan and an oral presentation before they begin parading before various venture capital firms like beauty contestants seeking a crown. In 1982, the venture capital world was smaller and more informal. Armed with their six-page business plan, plus the demonstrable fact that they had a working model of their product running at Stanford, Khosla and Bechtolsheim met with Bob Sackman, a veteran engineer turned investor whom Khosla knew from Daisy Systems. In Silicon Valley, would-be entrepreneurs cultivate a network of knowledgeable contacts they can turn to time and again. Sackman, with 30 years' experience in the electronics industry, had cofounded U.S. Venture Partners of Menlo Park, California. Another potential investor at the meeting was Doug Broyles, representing West Coast Venture Capital, also of Menlo Park. Within five days, the two venture capitalists wrote checks totaling $284,000 to fund the young company. "It was really on trust," Khosla told Professor Amar Bhide of Harvard Business School in a 1989 case study on Sun. "There was very little due diligence on their part--they just believed in the concept."
The business plan that got Sackman to invest was short, simple, and maybe a trifle naive, but it reflected qualities that would persist at Sun. The stated mission was to "develop, manufacture, market, and support graphics workstations for the CAD/CAM marketplace ... Maintain lead with the best cost/performance product on the market." Computer-aided design, which created blueprints for products from airplanes to chips, was a market of sophisticated engineers who demanded technical excellence and had no qualms about taking a chance on a new company that could supply it. Although Sun's market has since expanded upward to corporate information managers, the company has never lost its devotion to being a technological leader with its products, rather than being merely an integrator or solutions provider using technology from elsewhere. Ask any of the other top executives at Sun to define the company, and they will describe it first and foremost as a products company that controls its own intellectual property.
The tentative two-year forecast by Khosla and Bechtolsheim suggested that Sun would achieve total sales of $4 million in its first full year of operation, fiscal 1983, and $10 million in its second year. In addition, they expected to "manage company growth to produce break-even cash flow by the end of the first year." In fact, the company had sales of $8.6 million in its first fiscal year and $39 million in its second. Operating income was $588,000 in that first year and $3.7 million in the second. Despite its reputation for zaniness on other fronts, Sun has always been a tightly managed company that meets its financial objectives. In 17 years, Sun has had only one losing quarter--in 1989, an important turning point for the company. Read more about this in Chapter 5.
Thanks to Bechtolsheim's work at Stanford, Sun was already known within the scientific/technical community, which would prove to be the basis of word-of-mouth recommendations. "The SUN workstation has been reviewed and acclaimed by the scientific community," the original business plan stated. "Literally hundreds of inquiries have been received without any active marketing." The Sun founders had stumbled upon a community of supporters who would not only buy Sun's technology, but help it to evolve. No hardware or software vendor, no matter how big or powerful, survives without this web of external support. The web includes customers whose needs define how a technology evolves, and software developers who write programs for it. By targeting technical computer users first, Sun gained entree to sophisticated, independent buyers who generally made their own decisions about technology. This was not a new strategy: competitors Apollo, Digital Equipment, and Hewlett-Packard had all used it; Sun, however, would take it to the extreme, becoming a company of engineers selling to other engineers. Sun thus achieved a camaraderie with its customers, who propelled sales with enthusiastic reports. This early model gave the company a solid base upon which to expand.
MEN ON BOARD
Flush with seed capital, Khosla and Bechtolsheim could now turn to the problem of staffing their start-up. Khosla thought of a friend he knew from Stanford, someone who was running operations at another Unix start-up, Onyx Systems, in San Jose. This was Scott McNealy, who had grown up in a wealthy Detroit suburb, courtesy of his father's executive position at American Motors, and developed a reputation at college as a party-going jock. Though he attended Harvard and Stanford, McNealy was hardly known for academic excellence. Even today, he'll tell you he majored in "beer and golf." With his toothy grin and self-deprecating manner, the personable McNealy was highly regarded by his college friends, including Khosla.
After Sackman committed the money, Khosla took McNealy to celebrate at the local McDonald's, which suited Khosla's wallet and McNealy's fondness for fast food. Khosla said to his friend, "So when are you quitting your job?" McNealy replied, "You haven't made me an offer." Khosla panicked for a moment, concerned that McNealy would puncture his dream of the two of them working together. Then McNealy, who had the natural jokester's knack for timing, told Khosla: "I will take the job." McNealy, Khosla, and Bechtolsheim were all baby boomers born in the mid-1950s. "We were twenty-something-year-olds running a company," recalls Bechtolsheim, "and we had just met, but we certainly shared the passion" of starting a company.
McNealy--young, unattached, and eager for adventure--joined Sun as a designated cofounder and head of manufacturing. One of Sun's investors, Doug Broyles, was also the CEO of Onyx. Broyles had recruited McNealy there at the suggestion of a mutual friend, Bill Raduchel, who had known McNealy at Harvard and now worked for a customer of Onyx. "Scott doesn't know anything about computers, but he's a real good operations guy," Raduchel confided to Broyles. "He's going to make someone a lot of money someday." Broyles, who would soon become an investor in Sun and join its board, was favorably impressed with McNealy's energy and willingness to learn. "Scott grew up in a manufacturing environment because he'd worked summers for his father," Broyles recalls. "We brought him on as director of operations, and put manufacturing and purchasing under him. Within a couple of weeks, Scott had the 50-year-old manufacturing guy's respect and they were working as a team. Scott went out on the line and talked to people." McNealy's ability to relate to different kinds of people would prove critical to Sun, which has attracted an astonishing array of talent over the years, including many managers and engineers who went on to head companies themselves.
The trio of Khosla, Bechtolsheim, `and McNealy worked feverishly over the next three months to develop a prototype of the first product, the Sun-1 workstation, but they weren't completely happy with the Unix variant they were using as the operating system. Fortunately, just across the San Francisco Bay was one of the recognized experts on Unix, Bill Joy, a graduate student at the University of California at Berkeley. The team went over to woo him.
Joy is a bushy-haired, bespectacled genius who still delights in dressing in the psychedelic colors of Berkeley in the 1970s and 1980s. When he's discussing technology, he usually goes off into the stratosphere, way over everybody's head. But Bechtolsheim, who has the same knowledge of hardware that Joy does with software, hit it off instantly with the prospective recruit. He was impressed when Joy walked over to the Digital VAX minicomputer that he was using and casually turned it off. "He was shocked because you couldn't just shut computers down back then," Joy recalls. The tweaks he had made to the operating system--a version called Berkeley Unix, which Joy practically owned--allowed the quick shutdown. Joy was about the same age as the other three, so he shared the optimism of youth. He was also getting fed up with fighting for space and resources at Berkeley. After dinner with the trio, during which he spent most of the time talking shop with fellow tech-head Bechtolsheim, Joy signed on as the fourth cofounder. "I said, `what the hell'; these guys seem young and naive," Joy recalls, "but I was frustrated at Berkeley and ready to take a chance."
A Team Is Born
The key people were now in place for Sun to build a successful product: Khosla with the ambition and the vision, McNealy with practical manufacturing and personal skills, Bechtolsheim and Joy with the technical expertise. (McNealy and Joy remain at Sun, while Khosla is now with the prestigious venture capital firm Kleiner Perkins Caufield & Byers of Menlo Park, and Bechtolsheim is an engineering vice president at Cisco Systems of San Jose.) As the driving force, Khosla had grandiose ambitions for what was now called Sun Microsystems Incorporated. (The first name, SUN Workstation, was discarded as too narrow.) "I am a very competitive person and I didn't want to build a small company," he says. "There are companies when they start that are zero million dollars and there are companies that are zero billion dollars. Sun was a zero-billion-dollar company." The analogy he likes to use was that this start-up was a rocket ship that needed to reach orbital velocity. "The moon or bust was our motto," he says. In contrast to Khosla's intensity, McNealy was more lighthearted, joking that Sun "will be the biggest belly flop." You could see the difference already between their respective styles. Joy and Bechtolsheim kept their heads down and concentrated on technology.
But, what Sun lacked was someone to sell the product that Joy and Bechtolsheim were avidly designing. Joy turned to an older, more experienced friend of his at Berkeley, John Gage, then doing consulting for cable companies. Gage, today a grizzled visionary whose eyes still light up over the prospects that technology offers, is another noteworthy character in the company menagerie. A former hippie who helped organize anti-Vietnam War marches in Berkeley and Washington, D.C., Gage can wax poetic on everything from fluid flow dynamics to ethics. His sprawling house, located in an older section of the city of Berkeley, is plastered with old protest posters. He's a bright man who majored in mathematics, but whose real genius lies in his ability to assimilate information from diverse spheres and bring it together. Several years older than the founding team, he brought real-life experience to the table. Says Joy: "I brought John down because Scott and Vinod didn't understand marketing and sales. He was in a different zone," a description of the enigmatic Gage that rings true today.
Now a chief scientist at Sun, Gage "basically sold every machine in 1983," Joy says. He remembers how Gage would divide up the phone messages piled on his desk into various time zones. Gage "would start with Europe in the morning and, as the day progressed, he would move on to different time zones. Once in a while he shoveled everything back into a pile because he didn't have enough time." Adds Gage: "The first year all the sales came from me talking on the phone, saying, `How many do you want?'" When he joined Sun, says Gage, McNealy and Khosla were a bit clueless about the business world. "Scott and Vinod were trying to hire a vice president of marketing and sales by raiding Digital and Data General--talking to mature people with families and country club homes." Snorts Gage: "They wasted a lot of time." In addition to marketing and sales, he wound up handling technical support and answering all the outside calls. Gage took orders, and Sun built the machines and shipped them.
Sun was typical of many start-ups, chaotic and unstructured, but out of these early days constructs emerged that would solidify and guide the company's evolution. The first was that the company ran lean and mean, and employees worked long hours, usually with far fewer resources than their competitors. "I remember my wife calling about midnight to see when I'd be coming home, and I'd say it's going to be another couple of hours. Then I'd show up at 4 A.M.," recalls Crawford Beveridge, vice president of corporate resources from 1985 to 1991. Khosla and McNealy were both frugal, particularly Khosla. He figured out how many engineers he needed by looking at the smallest viable player (next to Sun) in the market, which he identified as Hewlett-Packard. "HP had 300 engineers" in the division selling workstations, recalls Khosla. "I said I'm twice as good as HP--I need 150 engineers." He adds that the notion of open systems allowed Sun to leverage innovations developed elsewhere, rather than rely solely on proprietary, in-house technology.
The second construct was that non-product-related functions, such as marketing and customer support, were dealt with on a rather ad hoc basis. They were afterthoughts to the engineering-driven technical aspects of the product. As it moved up-channel, Sun would be forced to revise these priorities, but they held for a good chunk of the company's history.
The third construct was that McNealy was the one who talked to employees and dealt with morale issues, which meant that he was emerging as the de facto leader, even though he was vice president of manufacturing and Khosla was the CEO. But Khosla was too driven, and Joy and Bechtolsheim too buried in technology, to relate well to personnel and cultural needs.
THE FIRST CRISIS
Sun introduced its first real product, the Sun-2, in late 1982. (The Sun-1 was essentially a prototype.) The Sun-2 incorporated elements that the engineering market was looking for: it ran Joy's Berkeley Unix, which was considered the coolest, hippest operating system of its day, on top of Bechtolsheim's sleek hardware design. So, it was fast, for its time, and relatively easy to use because most engineers are proficient in Unix. The best way to instill customer confidence "is to design products around accepted industrywide standards at all the interfaces where one technology will need to talk to another," McNealy explained in a viewpoint article in the December 16, 1985, issue of Computerworld . Unix "is one example of such a standard that can be shared by all the classes of computer equipment from supercomputer to desktop...." The Sun-2 had one more thing high on the target user's priority list: a big 19-inch monitor that allowed for leading-edge graphics capability.
The trouble was that the Sun-2's new monitor, obtained from an outside supplier, emitted so much static electricity that it could short out the whole system. Dave Cardinal, an early hire at Sun in product support, remembers the problem well. At the 1983 Comdex trade show, he was assigned to get nine Sun-2 machines set up for product demonstrations at Sun's booth. "We packed 20 machines to have nine that worked," he recalls. "The whole truck was full of spare material." When someone tried to use the machine, it would often reboot itself due to the buildup of excess electricity. "Out in front John Gage would ad lib his 'reboot speech,'" Cardinal recalls, while in back he and the rest of the small Comdex team worked frantically to get the machine running again. Because the failure rate was so high, Sun had to offer an extended warranty promising replacement of monitors to reassure customers.
That same year Sun recruited Bernie Lacroute, a 14-year veteran from Digital and a French immigrant, to head engineering. One of his first tasks was to fix the monitor problem. "It was a life-threatening situation," recalls Lacroute. He discovered that the monitor would discharge electricity through the connecting cables and zap the machine. "It was like a lightning bolt hitting the CPU (central processing unit) board." Sun couldn't immediately find another supplier, so in the interim Lacroute and his team tried a series of cables that might contain the electrical discharge from the monitor. "Every evening we'd try a new cable out on units on the manufacturing floor," says Lacroute. "We shipped the darn things knowing we'd have to replace the monitors. If we couldn't ship we were dead." The Sun-2 crisis took almost a year to really clean up. "We ended up hiring a couple of guys from HP who finally did it," Lacroute says.
The Sun-2 eventually became a relative success for Sun, selling in the thousands of units and paving the way for the next-generation workstation, which would propel the company to billion-dollar status. Despite its problems, the Sun-2 enabled Sun to develop a manufacturing strategy of building machines quickly to meet specific customer demand. Ship at all costs became the motto, which meant the four founders and everyone else in the company would congregate on the manufacturing floor the last day of a quarter to assemble machines for shipping. Finally, the Sun-2 was the product that Sun used to make a critical deal, one which forced the leading workstation vendors to sit up and take notice of the little upstart in California.
GRAPPLING WITH COMPUTERVISION
In the early 1980s, Computervision, based outside Boston in Bedford, Massachusetts, was the leading supplier of complete systems for the CAD market. In late 1982, Computervision set off a free-for-all among workstation companies by announcing that it would shift its reliance on minicomputers to the cheaper workstations, which were now powerful enough to meet the demands of CAD. The choice came down to the new kid on the block, Sun, and the established vendor, Apollo, which had the additional advantage of being in Computervision's backyard. Sun bid with its new Sun-2 workstations, which were lauded by Computervision's engineers, but Apollo landed the contract. Khosla got the bad news in a phone call from a Computervision purchasing agent.
In a move that has achieved legendary status among Sun old-timers, Khosla and McNealy grabbed a red-eye flight from San Francisco to Boston and showed up uninvited at Computervision. This moment was vividly captured in Sunburst , a book about Sun's early years. The two young Sun cofounders "planted themselves in the plush lobby of Computervision's headquarters. From there they called everyone they knew inside Computervision, asking for another chance, demanding an opportunity to revise their bid." It was a classic example of the win-at-all-costs West Coast approach versus the more polished, mannerly East Coast mentality. Apollo and Computervision figured they had a deal; Computervision staff members tried to shoo Khosla and McNealy out of the lobby. Nothing doing. Finally, a Computervision vice president cut a deal with Khosla and McNealy: Get out of the lobby and Computervision President James Barret will call you at the local sales office.
According to Sunburst : "When the call came, Khosla knew it was his last opportunity to sell Sun. So he sold hard. Winning Computervision ... would prove to industry watchers that Sun was real ..." Sun virtually gave the machines away at cost, a number of people recall, but it won a three-year, $40 million deal with Computervision that established the company as a player. Khosla sprang for a celebration cruise on the San Francisco Bay for the entire company, which consisted of 40 people. The essence of the deal was not the price Computervision paid for the workstations, Khosla emphasizes, but the fact that it would also manufacture and resell them. That meant Sun captured a portion of the gross margin on Computervision's sales with no manufacturing costs on its part. Sun had also effectively blocked its biggest rival, Apollo, from making the deal that might have established its technology as the preeminent standard for CAD/CAM.
Todd Basche, who would later join Sun as a director of desktop systems, was an engineer at Apollo when the Computervision deal was sealed. "I remember being in Apollo meetings where people would laugh at Sun and say, `It's a bunch of kids at Stanford,'" he recalls. After Computervision, "they changed their tune." He adds, "Basically, Sun sold its soul to get the deal. But we didn't laugh at Sun any more." Basche believes Apollo's machines at the time were more reliable and superior to Sun's. Sun won, he says, for two reasons. First, Apollo's traditional management team was unprepared for Sun's aggressive, unruly approach, and didn't know how to react when Sun snatched the deal away. Apollo's executives could only fume that their rival hadn't played fair. Second, Sun's workstations used Unix, which was becoming a standard in the engineering world. That meant less training and enhanced collaboration with colleagues who were likely to know Unix but would have to learn a proprietary system like Apollo's. If Apollo had been willing to move to Unix, Basche suggests, "it would have been a different story."
Because sales were never huge, the Computervision deal was not a big moneymaker for Sun, but it did something much more important in opening the door for other significant contracts. A year or so later, for example, Sun bid on a contract with the National Security Agency, which had already bought Apollo workstations in the past. "I thought Sun might get 40 percent of the contract and Apollo 60 percent," recalls Gage, who handled the negotiations. "We got 100 percent." The NSA deal underscored the importance of Sun's use of a standard operating system, because federal agencies needed to exchange information and were moving to Unix as the means. Sun was well on its way to a goal no one would have believed possible even months before: overtaking Apollo. Or, as McNealy was wont to say, " killing Apollo." Khosla still has a company T-shirt from the time with the word Sunburn stamped across Apollo, as well as Digital Equipment and IBM.
TROUBLE AT THE TOP
Sun soon learned that for every silver lining like the Computervision deal, there was often a cloud. In this case it was a major thundercloud: a raging dispute between Khosla and Sun's new president, Owen Brown, that was exacerbated by the twists and turns of the Computervision negotiations. The board of directors of Sun, which now included four investors, Sackman, Broyles, John Doerr from Kleiner Perkins, and David Marquardt from Technology Venture Investors (TVI), wanted more experience at the company. In early 1983, Khosla and Broyles flew to Massachusetts to recruit Brown from Digital, where he had nine years of experience as a sales manager. He also had an engineering degree and had served a stint in the Navy, where he was still on reserve duty. Brown, who graduated from Auburn University in 1964, was 42 years old, a decade and a half older than the Sun founders. On paper his credentials looked like a perfect fit with what Sun needed, and Khosla lobbied hard to bring Brown on as president, over some objections from the board. But soon, a rift opened between Khosla and Brown.
"The power struggle with Vinod [Khosla] just got worse and worse," recalls Brown, who now heads his own management consulting firm in Silicon Valley. "Vinod was CEO and I was president. I was going to focus outwardly and he was going to focus inwardly. But Vinod could never just focus inwardly, so there was a constant struggle." In a small company, it was hard to hide that constant struggle. Khosla says another point of tension was that Brown "wasn't particularly action-oriented and not very inclusive of Scott in the decision-making."
To top it off, Brown was away from Sun, on a two-week duty in Norfolk, Virginia, for the Naval Reserve, when the Computervision deal seemed ready to go Apollo's way. Khosla blew his stack, pinning the blame on Brown for what could have been a near-death blow to the young company. Brown says he was involved with the Computervision negotiations and helped cement the deal. "It just so happened that [the two weeks] was when we got word that Computervision had chosen Apollo," he says. Regardless, Brown had violated the Sun ethic, which was that the company came first. Sun has always demanded an intense loyalty and a willingness to do whatever it took to make the company successful. Not everyone wants to work in an atmosphere so prone to overload and burnout, and that was the conclusion Brown drew himself.
Not long after the Computervision deal, Brown went to Honolulu to serve another week of reserve duty. One day when he called Sun to check his messages, he had one from Lacroute, who broke the bad news. Khosla and McNealy had gone to the board and insisted that Brown had to go, or they would leave. "I had made no secret of the fact that Vinod and I were having difficulties," says Brown. It was obvious that the board had to act. "I could have put up a big fight over this and taken a lot of people down, but I decided to resign," he says, leaving Sun in February 1984.
GIANT STEPS FOR A COMPANY
Brown was gone, but his legacy remained. From Digital, he had helped to recruit other senior executives for Sun, like Bernie Lacroute as executive vice president and Carol Bartz as vice president of marketing. Meanwhile, Khosla's own credibility was severely damaged in the prolonged power struggle. The board, unwilling to cede all the executive authority to the CEO, elevated McNealy to president. The exacting Khosla, who had the drive and vision that had brought Sun this far, would be offset by the more personable McNealy. The two were friends, former classmates, and even roommates--with McNealy sharing a townhouse with Khosla and his wife. As Sun entered its third year, the power-sharing arrangement seemed like an ideal solution.
In spite of the technical glitches with the Sun-2 and the personality issues with Brown and Khosla, Sun in early 1984 could look back on its first two years with a sense of solid achievement. It was closing in on $39 million in sales in only its second fiscal year of operation. It had won over Computervision, opening the door to a number of other big deals. Its success had forced the rest of the minicomputer/workstation industry, which at the time was concentrated in the Massachusetts area, to take these kids from California seriously.
More than that, Sun had a strategy that would take it a lot further than anyone ever thought it would go. When Carl Swirsding joined in 1983 to handle creative services, he interviewed with Khosla, who, characteristically, took him to a hot dog place for lunch. "We sat outside and I asked him, `How big do you think this company is going to be?'" Swirsding recalls. Khosla replied with total conviction: "We'll be $100 million in a few years." This convinced Swirsding that the company had a future, and he signed on. (What Sun actually reached in five years was $537 million.) "I had a sense that these guys had a vision of where they were headed, a perspective of what they would add and how it was unique," says Swirsding.
Sun's business strategy, developed informally and piecemeal and, in some cases, out of sheer necessity, turned out to be right on. First, it put a computer on every user's desk. This gave engineers a freedom they had seldom experienced. Rather than standing in line to use a centrally located minicomputer or mainframe, they could sit at their desk and fiddle with whatever they were designing, and the workstation would immediately show them the new iteration. Apollo invented the workstation, but Sun improved on it by giving engineers a universal operating system that they loved and a hardware design that could be easily reconfigured to meet a fast-moving market. Second, it gave each user a large monitor with a graphical user interface (GUI). Although Apple would make the interface a major selling point in the PC world, computers of the time, particularly those for professionals, relied on long lines of code typed into the machine. Sun's GUI was radical in simplifying the user's interaction with the computer. Third, Sun networked every computer, allowing users to collaborate from their own workspaces. Fourth, it used industry standard parts to build the machine, undercutting Apollo's pricing by half while enabling rapid evolution of the product. Fifth, it pinpointed where it could truly add value--for example, in tuning Berkeley Unix to a higher level.
Finally, Sun would do anything to win. With Computervision, it gave the larger company manufacturing rights to products that would be developed jointly. This was rare in the early 1980s, when the monolithic, vertically integrated model of IBM and Digital was the norm. From its infancy, Sun was a disruptive force that would break any rules it had to to succeed. This kept competitors and allies alike off-guard, while it gave employees a clear sense of where Sun stood in comparison with the competition. "Sun did the big things right," sums up Bechtolsheim. "We had the right business model--the open system approach and open attitude toward our customers. Once you get that right, you can afford mistakes." Fortunately for Sun, the company had gotten two big things right--its business model and its new president, because there was more turmoil lurking around the corner.
Copyright © 1999 Upside Magazine. All rights reserved.
Need Help? Copyright © 1999-2025
