Share Your Thoughts


Eleven months ago, my brother became a citizen of the United States. The oath­taking ceremony was held at the Masonic Auditorium in San Francisco. I remember being stunned by the sheer diversity of the freshly-minted citizens. Among the 2,000 hope-filled faces was that of a lawyer from Iran, a businesswoman from Ghana, a barkeep from Ireland, an au pair from Germany. Call it a melting-pot, a salad bowl. a bouillabaisse, it was the most visceral experience I’d yet had, illustrating the success with which America blends in new immigrants.

I would have expected that the constant stream of fresh blood would mix seamlessly into the body politic. But at every turn, I found this not to be the case. Immigrants clumped into professions formed thriving little communities that specialized in certain businesses. Every taxi-driver that ever ferried me to San Francisco airport seemed to be from Brazil. On a road­trip in the American Southwest, I heard only Gujarati spoken at each motel check-in desk. There seemed to be not a single doughnut shop on the California coast that wasn’t run by a Cambodian. And an investment in a San Francisco deli led me from Feroz to Hassan to Abdul, every last owner from Palestine. This phenomenon isn’t merely the stuff of anecdote. The canonical work in this field is “Race and Culture,” a detailed study of immigration and assimi­lation written by Hoover Institute scholar Thomas Sowell.

Chapter by chapter, he makes a convincing case that the primary determinant of success of an ethnic group is its “cul­ture’-its attitudes towards work, thrift, and education. Groups tend to enter a new country with skills that help them spe­cialize in a niche. They expand within that niche by financing new stores or outlets in those businesses that they understand, and by staffing those same businesses with their relatives and friends. This is why the grocery business in Compton is dominated by Koreans and why newspaper stands up and down the New York subway lines are run by Indians.

The conclusions reached by this relent­lessly detailed scientific examination are echoed by a business bestseller written in 1993. The book is called “Tribes,” in which Los Angeles journalist Joel Kotkin argues that the new commercial world order will be based on enterprising networks of ethnic groups connected to each other by jetliners, fax machines, and the Internet. It is hard to dis­pute this when reading about Hong Kong Chinese ferrying money from their friends in Vancouver to open assembly lines in Shanghai and then selling the products to tourists in Jakarta.

One of the most successful examples of the ethnic diaspora is the Indian community in Silicon Valley. Over 30 years, students from every corner of India have followed a well­-worn path from village and city to IITs (Indian Institutes of Technology) and RECs (Regional Engineering Colleges), and thence to a hun­dred M.S. programs at U.S. college towns with names like Stillwater and Gainesville and Chapel Hill.

The generation in the 1960s came from these universities to the Bay Area, attracted by the vast number of jobs available at the defense contractors, companies like Lock­heed and Ford Aerospace. The generation after came just on the cusp of the computer revolution, to design microprocessors at Intel and graphics cards at Apple. The current gen­eration seems to eschew all need for security. They bypass the seven-year stretch of accu­mulating a 401(k) and buying a 3-br 2-ba in Fremont, and plunge straight into building a company at age 26, companies with names like Exodus and Hotmail.

Over these 30 years, the ethos of the incoming Silicon Valley Indian has changed. Where there was once the ‘need for security and a long climb up a corporate ladder, there is now the urge to create an enter­prise uniquely one’s own. The quest for a window office is being replaced by the hunt for a

warehouse for the startup. The time that used to be spent shopping for a new Honda Accord is now spent at consignment offices finding used office furniture and in registering new Internet domain names.

What had been missing from the Indian community for much of these 30 years was a sense of community, not social community—there were enough rummy parties and Saturday idli brunches to fill a year’s worth of social engagements—but business community. The forums where an entrepreneur that had survived a harrowing year-long search for financing could tell 10 others of the pitfalls along the way and of the strategy that had finally worked; the cocktail hours where com­patriots across different companies could congregate and talk about the trends that they perceived, the sale that had been unforeseen, the business prospect that the other should talk to; the scheduled occasion where the young hopeful could stand in the same room as a Suhas Patil or a Vinod Khosla and ask them what exactly they had done in founding a Cirrus Logic or a Sun Microsys­tems.

The first steps towards changing that occurred in late 1992 when a group of Indian businessmen found themselves cooling their heels at the Santa Clara Marriott, waiting for an Indian government official whose flight had been delayed. They redefined the agenda to be a “get-acquaint­ed” meeting. From that group was born The Indus Entrepreneur lTiE), an organization whose members now number  over  400. The members meet every month, sometimes just to network, and sometimes to hear about a recent company or business plan.

TiE has so far kept determinedly away from being any kind of venture capital firm. Its goal is to be a forum for all business people from the South Asian subcontinent to get acquainted with each other. Membership is open to anyone, and details are available at

Starting in 1993, TiE has organized an annual weekend-long conference called TiEcon. The success of the conference has been remarkable. It is co-sponsored by 17 of Silicon Valley’s biggest venture capital and law firms.

The roster of speakers—Larry Ellison, Joe Graziano, Gordon Bell—reads like it was taken off the list of attendees at a Fortune magazine business executive summit. Last year, the San Jose Fairmont ballroom was packed to the rafters and more than a hun­dred people had to be turned away.

India Currents this year, in a section ori­ented around TiEcon, is profiling three busi­nesspeople who have been instrumental in TiE’s success.

Suhas Patil (‘The Builder”) is the founder of Cirrus Logic, a billion-dollar company. He is one of the founding fathers of the electronic design automation industry, and a man whose ideas have led to seminal changes in areas such as fabless chip design and compact computer storage.

Raj Jaswa (“The Entrepreneur”) is the embodiment of the drive and doggedness that it takes to get your product in the mar­ketplace. Six months after a bittersweet part­ing from Opti, the first company that he founded, he was back starting Selectica in an office park in Santa Clara.

Finally, Kanwal Rekhi (“The Mentor”) had a stellar career at Excelan and Novell, only to come back with twice the enthusiasm after retirement. In the last two years, he has advised, helped fund, and guide the fortunes of no less than 10 new companies.

Business Focus: Suhar Patil – The Builder

Suhas Patil is ensconced in a leather couch in his eyrie high above Cupertino. An exquis­itely laid-out garden can be seen through the huge bay windows behind him. Spring showers have just passed, scattering raindrops on the leaves. The sun has started to come out, and a gentle breeze is beginning to stir. Inside there is absolute quiet. The floor is spotless, the brass fit­tings gleam, the object d’arts in the living room are few but tasteful. The whole scene is out of a glossy maga­zine, perhaps one on refined living. It is hard to imagine the path he has traversed over 35 years, beginning from a store full of broken radios in Jamshedpur.

Oh, how he loved that store! His father ran a business repairing electronic equip­ment, and helped Suhas graduate from Meccano sets to saws to screwdrivers. The boy that took apart toy trains soon learned to put together bicycles and eventually to fix the fabric-fronted Bush sets with their huge valves so that they picked up the latest Hindi film songs on the Binaca Geet Mala show. It was the start of his lifelong love of electron­ics.

There was only one place In India in 1961 that offered an electronics program of any distinction. It was the Indian Institute of Technology at Kharagpur, funded by UNESCO and modeled closely after MIT, right down to the road that separated the academic depart­ments from the residence halls. There was no question that was where he was going to go. He graduated first in his class and, to his delight, was accepted to the doctoral pro­gram at MIT.

Boston in 1965 was the center of the elec­tronic universe. IBM had donated a massive mainframe to the university for advanced research. The 7094 ran one of the very first time-sharing systems, called CTSS.

The pioneers in artificial intelligence were starting the legendary AI lab, while at the same time the fundamental blocks of com­puter architecture were being laid by his advi­sor, Professor Dennis, among others. Patil joined the research group on multi-access computers, famous nowadays as Project MAC. The subject of his thesis was the coordination of events in parallel computers so that the results were provably correct.

While his undergraduate years had tested his ability to ingest and regurgitate facts, the Ph.D. program was all uncharted territory. He was faced not just with having to stake out new horizons of knowledge but also with dealing with the disappointment when a direction proved fruitless. In an age where young company founders do not have the patience for school, he holds the opposite opinion. Serious graduate work, he believes, is the best possible training for the drive and persistence needed by entrepreneurs in later life.

He loved MIT, and had no second thoughts about accepting an offer to join the staff there after gradua­tion. To his surprise, teaching was but a part of his responsibilities. He had to rein in runaway budgets, get lab heads to work in unison, and, most onerous of duties, be responsible for office allocation. Distinguished academics became children when the issue was a window or a view. Because his powers were manifold, he learned that a carelessly uttered word could cause turmoil. Even now, he speaks with great delib­eration, perhaps a habit learned from those days. His mind is always weighing the differ­ent ways his sentences can be construed before he permits himself to utter them.

He worked there for five years, and was promoted to Associate Professor. It was an exciting time in his field. The huge radio valves that he remembered from the equip­ment in his childhood had been replaced by tiny transistors and now researchers all over the world were racing to etch more and more transistors onto a single chip. The electronics had gotten so minute that people were talking about radios that you could slip into your shirt pocket or wear on your wrist. Very Large Scale Integration, it was called. Carver Mead down at Cal tech was working on software that would take the input and output that you needed from a component and actually gen­erate the logic diagram to fabricate the chip.

He was building up a reputation and he got an offer from the University of Utah. The department there had been funded by the National Science Foundation and had had a whole chip fabrication facility donated to it by General Instruments. Patil would have the lib­erty of testing out a new two-dimensional lan­guage he had developed. It described a chip’s architecture and allowed it to be translated directly to photo-masks for etching the cir­cuitry in silicon. The opportunity was too tempting and he moved across the country for the new job.

General Instruments loved his work, and saw great potential to use it in the design of the circuits for their cable television set-lap boxes. Within a year, they had asked him to work full-time for them on the project. He presented them with the idea of starting new company to do this. They immediately agreed. He cast about for a name for the company. High technology companies seemed to allhave names with many z’s and x’s: Multex, Zantex, Genzan. His lawyer advised him that if he thought the company was going to succeed, he should name it after himself. And so Patil Systems was born. A number of his colleagues at the department at Utah came with him.

General Instruments was a supplier to Atari. The whole computer game market cratered in 1982, as kids cross the world got tired of playing Pong and Asteroids. Atari began what was to be a long fall to extinction. Faced with losing one of its biggest cus­tomers, General Instruments started cutting back on all new projects. The one involving Patil Systems was an early casualty. All the promise with which Patil had embarked on the new venture began to evaporate.

Startups are a triumph of hope against reality. One hears mostly of the Netscapes and the Yahoos, the dozens of overnight mil­lionaires. Very rarely does one learn of the drudgery and toil, the hundreds of nights spent staring drearily at a computer screen, the fear of whether you’re going to make your rent check this month. For two years, Patil Systems stumbled along, supported on the loans made by family members and college friends. At one point, Patil was forced to take aside Dr. Ravindra, one of the other faculty members who had left with him, for a heart­-to-heart talk. Funds were running terribly low, he told him, and he wasn’t sure it was the right company any more for a person with a young wife, expecting their first child. Not to worry, was the reply, he was going to gut it out no matter what the pain. Fifteen years later, Patil still chokes up when he remembers that conversation.

One fact was slowly becoming clear. Most of their customers came from a 40 square mile area In Silicon Valley. The mountain had to move to Muhammad. In late 1983, four of them moved from Utah to Northern California. They incorporated there on February 3, 1984.

Regis McKenna, their consultant, advised them to pick a corporate name that had pos­itive connotations. They spent a day writing up all plausible choices on a board. The winner was the name of an airy cloud that signified clear skies and sunny weather. Cirrus logic.

They worked furi­ously on the design of an Integrated chip for Jer­rold Systems that would drastically bring down the cost of set-top boxes for the selection of cable channels. They came up with a design that worked perfectly but found that they had no money to actually get the chip fabricated. Swallowing hard, Patil gave the design to the customer for nothing more than the $100,000 engineering fee.

It was a tough financial decision, but it was the right one because it gave his compa­ny cachet and credibility. Other customers came calling. News of their success spread, and they were courted by a host of venture capitalists. In nine months they were finally well funded, with names like Nazall, Brentwood, lVl, and Robertson Stephens all on board. With this war chest backing his cut­ting-edge technology, Suhas Patil was able to mount an all-out campaign to bring on one of the most highly regarded managers in elec­tronic design as CEO. Mike Hackworth was a marketing and sales wizard who was in charge of 2,300 people at Signetics. He signed on with Cirrus, which at that time had a total of 12 employees.

Patil and Hackworth complemented each other so well that they became known as one of the best manage­ment teams in Silicon Valley. Patil would emerge from the labs after much thought and careful design with a product that was aimed dead square at a consumer need. His partner would then set the compa­ny strategy to market and sell that product for all it was worth.

Cirrus’s next innovation was to bundle the controllers for disk drives with the actual drives themselves rather than treat them as separate bulky entitles. This made the drives much more portable and spawned an entire industry in Integrated Drive Elec­tronics. In 1989, they introduced a controller chip for the IBM VGA graph­ics standard that allowed identical pro­grams to run on laptops and desktop computers for the very first time. They went public that year.

Each year seemed to bring a new hit.

One year. It was a way to connect CD-ROM drives directly to PCs without needing costly adapters. The next, it was helping define credit-card­sized PCMCIA cards that are used in almost all the portable computers sold today. Fax modem chips came shortly thereafter, followed by chips for the wireless communi­cation market. From 1992 to 1996, the company’s annual revenue grew sixfold, from $1 71 million to $1,147 million. Along the way, the company pioneered a way of working called fabless design. The engineers designed all the circuitry, and sent the blue­prints to be fabricated at other company’s foundries. That way, there was no necessity to invest hundreds of millions of dollars in capi­tal equipment. It also resulted in more time and freedom to attack new markets quickly and to be nimble in delivery.

Investors took notice of Cirrus’s success.

During three giddy quarters in 1995, the stock rose from $12 to $62. A billion-dollar compa­ny with that kind of visibility attracts a lot of competitors, especially in the jungle of high technology. And come out of the bushes they did, Oak Technologies in CD-ROMs, Sierra In modem chips, S3 in graphics. Adroitness and instant response to market changes became the company watchword. For every PC sold, Cirrus sold about a dozen support chips worth around $200, generating revenue per PC second only to Intel’s $300 per micro­processor. For better or worse, the company became known as a bellwether for the health of the PC industry. A good quarter for Cirrus meant that PCs were selling well.

In early 1995, with the order queues being full, the company made a decision to cut back its customer list to only a dozen or so of its biggest customers. In order to ensure supply, some of the largest customers like Packard­Bell placed their orders way in advance. But they forecast wrongly since customers bypassed the older computers in favor of faster ones Introduced in late 1995. Frantically, Packard-Bell and others cut back on their earlier orders. Cirrus’s sales in the last quarter of 1995 were disappointing. The stock got slaughtered, and retraced Its way back from $60 to the $20s. Momentum Investors tend to be a nervous lot.

Talk to Suhas Patil now about those days, though, and you are met with total equanimi­ty. He is unperturbed by the cavalier way that Investors treated the company. He has reason to be sanguine. The company has had a recent $300 million convertible offering. Inventory levels are substantially lower than a year ago. The PC business continues to boom they have introduced new high-perfor­mance audio cards, graphics processors, and mass storage controllers. There are plans for personal digital assistants, digital video disks, and network computers.

There is still the air of the inquisitive child about him, the person who cares deeply about how something works. He is the rare personality who blends both the analytic abil­ity of an MIT professor and the tinkering itch of a backyard inventor. A recent issue of ‘Byte” magazine hazarded guesses about what the computer of the next millennium would look like. There were wild postulations about holographic imaging, terabyte thumb­nail drives, speech-driven commands. Whatever the reality is, it is easy to visualize one image. It is of that computer taken apart, all the little pieces being subject to micro­scopic scrutiny, on Suhas Patil’s workshop bench.


Carl Gustav Jung once wrote that life is lived forward but can only be understood backward. When faced with a decision, humans often make their choices blindly. They accept the job offer or agree to move cities, or vol-unteer for service in the army. Each choice throws them in new situations and makes them confront yet more decisions. At the very end of life, they reminisce and relate their stories as if they were pushed forward at each step by pure rationality rather than by happenstance and impulse. Once they become the storytellers, they polish and smooth the telling of each stage as the child in their lap looks at them with awe, marveling at how smart grandpa was when he was young.

The mythologist Joseph Campbell called this “The Hero’s Journey.” Each person fashions the myth of their lives, making up the story into which each episode and situation can be neatly fit. The simplest synopsis of this on paper is called a resume. There is the half-hour version, which you relate to a seatmate on the plane, and the years-long version which you share gently and grad­ually with your spouse or your best friend. The paradox of this myth is that it is remarkably similar among the vast majority of people in the broad brush strokes; and yet it can be remarkably different in the fine lines and sub­tleties. Most everybody has the same experi­ences of childhood, school, adolescence, marriage, middle age. And yet there are those that get particularly good at one aspect of liv­ing that others merely pay lip service to. These are the people that you read newspa­per articles about or the ones who inspire others. These are the people like Raj Jaswa.

Jaswa’s early life reads much like the story of any other student in India who dreams of coming to the United States: the boy who was smartest in his class, the math and science whiz, the determined struggle with the Agrawal Joint Entrance Examination training classes, the admission into lIT Bombay and into graduate school in electrical engineering at Toronto. Hidden in the description of those fledgling years is the will that made him choose not to take the easy path of partnering with his family members in their thriving businesses in real estate, jewelry, textiles, and su­gar. Also hid­den is his steely efficien­cy. He gradu­ated on a Monday, got his green card that Tuesday, set up a job interview for We­dnesday, and had started work on Friday. He started as a junior design engineer at RF’ Communications at a then-princely salary of $16,500 a year. The city was Rochester, New York, often confused with being a suburb of New York City but in reality 400 miles north- west, not far from the Canadian border. Winter lasted from November to April, a peri. od during which icy winds gusted off the frigid waters of Lake Ontario. Eight months was enough for him. He started interviewing all over the Sun Belt and ended up accepting a job at General Electric in Daytona Beach, Florida. Some day in the future, when schol­ars study why the South and California became magnets for industry in the late 20th century, weather will show up as the number one reason.

IIT had given Jaswa a broad-based educa­tion, with exposure to the liberal arts. He added to his academic credentials by getting an MBA by night at Stetson University. He decided to move to marketing and became a product manager at lntel in Santa Clara, California. He was given responsibility for the compiler software for the 8086 chip. Today, 95% of Intel’s business comes from micro­processors for the PC. In those days, the fig­ure was under 10%. The majority of Intel’s business came from the embedded market, where its chips were used to control the path of machine tools, or in electronic fuel injec­tions for Ford cars, or within the coneheads of Cruise missiles.

As it is now, Intel in 1981 was a very well­-run company. Top management came from the engineering ranks, and was intimately familiar with issues that would have been too detailed for executives at other companies. It had a stellar training department, where employees could sign up for a course as often as they wished to. There was a healthy amount of creative tension, with organization charts shuffled around almost annually in order to expose employees to as many different aspects of the business as possi­ble.

Jaswa thrived in the rough and tumble of the new Environment, but he cried halt when his group got moved to Oregon. He had begun to belie­ve that most im­portant advances in technology happened in Silicon Valley because it was one of the very few places in the world that had hit critical mass in engineering talent, venture capital, and manufacturing capability. It would be like leav­ing Florence during the Renaissance. He decided to stay, and took a job marketing chip sets that were peripheral to the main micro­processors but which were nevertheless needed for everything from graphics to net­working.

Even the best-managed companies suffer during an industry downturn. In 1986, just such a slump hit the entire chip industry, Intel, as the bellwether, suffered the most. Products were axed, assembly lines were shut down, and the corn­pany underwent nine separate rounds of lay­offs. In early ’87, Jaswa switched over to Chips & Technologies, Intel’s principal competitor in the chipset business.

It was the meeting of the man and the moment. The IBM PC AT had come along, and was revolutionizing business. It used the 386 chip from Intel and the BIOS software from Phoenix Technologies to support MS-DOS. The third and final piece of the puzzle was the integrated chipset made by Chips & Technologies to allow the PC to use a modem or hook up to other external devices. Jaswa launched the NEAT chipset, which stood for NExt generation AT. He traveled the globe, to France, to Korea, to Taiwan, to England, pros­elytizing all the while. In two years, he took the product line from $30 million to $200 mil­lion in annual sales. The company recorded profits of over $100 million, with gross mar­gins in the astronomical 70 range.

It should have been the zenith of his pro­fessional career. He was called in for a meeting with senior management. Surely, he thought, this must be where they rewarded him for the startling success of his product. But no, they were hauling him in to tell him that they didn’t think he had the right stuff to manage a business unit with revenues as large as $200 million, he who had been largely responsible for its success in the first place!

Here’s where this particular myth diverges from the myth of the majority of people in this situation. Given his title and high pay, most people would have swallowed their pride and rationalized that it was possible that the senior management was right. Instead. Raj laswa quit.

Three of his fellow workers had come to him with an idea so crazy that it was just remotely possible that it would work. They were proposing to design a chipset that would pair the new 486 microprocessor with the connector or bus for the older IBM AT architecture. This would mean pairing a 32-bit processor with a 16-bit bus. It seemed almost retrograde, a path to technology that was behind the times. The only other feature was an integrated cache controller, a component others deemed so complex that it had to be independently designed. On blind faith, and not drawing a salary, the four of them rented a 4.000 square foot office on Augustine Drive. His mother-in-law had joked that he was an incorrigible optimist. So they named the com­pany Opti.

The idea turned out to be a stroke of genius. IBM was pushing the MicroChannel bus, which was proprietary and which they thought would help them wrest back control of the PC from Intel and Compaq and Microsoft; nine other manufacturers were arguing for a new and still largely untested bus called EISA. But people were very com­fortable with the AT bus. And Opti was the only game in town.

It is extremely rare for startups to be prof­itable in their first year. Opti was cash flow positive in six months. In 1990. the company recorded $25 million in revenues. At a head­count of just 20 people, it was bringing in more than a million dollars a person. Chips & Technologies, their old company, fought them in court for 18 months, then settled out of court.

In 1991, Opti’s revenues soared to $65 million and the year after, to $98 million. They were outselling Texas Instruments. Western Digital, and a slew of Japanese competitors, companies with ten times the number of engi­neers that they had. The company ran like a sports car with two wheels constantly off the road. In order to get ahead in a manufactur­er’s allocation queue, they placed orders for wafers before their chip design test results came back. They worked obsessively to make sure that there was never a quality glitch that would delay their getting product to market. They also made it very easy for PC manufac­turers to use their products, going so far as to design the entire motherboard for them.

Opti went public in 1993 at a valuation of $180 million. Revenues kept increasing over the next three years. IBM signed on as a cus­tomer. So did Olivetti and Hewlett-Packard. In 1995, the company was sitting pretty with an annualized revenue stream of $150 million. It was the darling of Wall Street analysts. The stock began a steady climb to the high $20s.

Raj laswa was identified as one of the 20 best-compensated executives in Silicon Valley. His net worth rose to the tens of mil­lions.

One of the great world myths is that of Icarus flying too high and getting his wings melted by the sun. The fiery body in this case was Intel, which had surged from strength to strength in the mid-nineties. To ensure continuing growth, that company started taking over the design and manufacture of increasingly large pieces of the motherboard. One of the pieces it decided it wanted was chipsets. That put Opti in its crosshairs.

Opti found its customers being siphoned off by Intel. In the space of one week, it lost Packard-Bell, which was responsible for $40 million in annual revenue. In the third quarter of 1995, when the company had a shortfall, the stock began a freefall. Jaswa came back from Christmas vacation in India to find out that the rest of the management team had executed a coup d’etat. He was without a job.

Again the myth diverges from the path that would have been followed by the majori­ty of people. Early forties, net worth in the millions-it is an invitation to step back and enjoy all the wealth. Jaswa did do that for a few months. He took his family to Africa, vis­ited Victoria Falls. But an itch gnaws at true entrepreneurs that most others do not even comprehend. He had to get back in the game.

He spent many months talking to differ­ent people about ideas that they had, ideas that could be translated into products and sold in the marketplace. He talked to dozens of inventors. He knew that what he enjoyed most was growing small companies. He want­ed to get into the brand-new economic world being created by the Internet. And he wanted to have a good portion of his new company be based in India.

The person he found was Sanjay Mitral, a scientist who had worked for many years at the famous Xerox Palo Alto Research Center. The fruit of his work was an artificial-intelli­gence engine that lent itself to being imple­mented under lava, the hottest programming language on the planet. The engine was ver­satile and compact enough to be used by Web sites so that customers could configure prod­ucts with very little assistance.

For instance, if they went to the BMW home-page, they could start with the basic model 3251, and configure it as they desired. As each option was added, the picture of the finished car would change on the fly, and a lit­tIe bill on the side would show what the new total cost would be. The best part of the soft­ware was that it would allow rules to be encapsulated. For instance, someone choos­ing the 3.2 liter version of the car might not be able to choose the turbocharged option. Using Mittal’s software, they would be auto­matically prevented from doing so. That option would be whited out from their menu of choices.

Jaswa sits today in a small office in Santa Clara. His new company. Selectica, was founded in tune 1996. He has programmers working frantically to finish the release of the SRx product. There are dozens of phone calls to return, a JavaOne booth to plan, financing to arrange, people to hire. His plans call for 40 employees by the end of 1997, half in India. He seems energized by all the bustle around him. Icarus has found his wings again. This man is fixing to fly.

Business Focus: Kanwal Rekhi – The Mentor 

I was born in Rawalpindi in the pre-Partition days. We left Pakistan when I was two years old, with nothing but the clothes on our backs. We were completely unencumbered when we settled in Kanpur. Being uprooted every so often is good because inse­curity is the most necessary thing for growth. That’s why immigrants suc­ceed in their adopted countries at a higher percentage than natives.

“We were an army family. My grandfather was a soldier, my father was in the army, his brothers all served in the armed forces. Both my brothers went to the National Defence Academy in Khadakvasla. Nobody knew that I had applied to the Indian Institute of Technology. I didn’t take any special courses to prepare for the entrance examination. I’d always been good in mathematics and physics, and I just got into the institute at Powai (Bombay). My father did not even know what the lIT was. He’d always wanted me to become a chartered accountant.

“I studied Electrical Engineering and did well enough to be admitted to the graduate program at Michigan Tech. Have you heard of it? I didn’t think so. It is in this tiny town called Houghton in Michigan. Midwestern, real heartland values. I love the place. They have just invited me to be on their board.

“It was hard to get summer jobs then.

That first summer, I took a bus to Chicago. It was 1968, the year of the famous Democratic National Convention where the police where fighting with students all over the streets. I had $20 to my name when I got off the bus. The boarding house cost $4 a night, and din­ner cost a dollar. So I had only four days before my money was going to run out. There was a restaurant that had a sign that said ‘Help Wanted: I walked in and they offered me $3 an hour to wash dishes. I put on an apron and didn’t have to starve for those three months.

“Those were tough years to be graduating as an engineer. The space program was wind­ing down and the Vietnam War was past its peak. Thousands of engineers from these two efforts were entering the work force. I must have applied for a hundred jobs. I got an offer from IBM in Kingston. The second offer came from EAI, which made analog computers, which you are too young to remember. I thought I’d learn more from a small compa­ny. I turned down the mighty IBM.

“It was a bad choice. I got laid off within a year. I went back and re-applied to IBM. But they had long memories. They had actually kept my file and noted that I had rejected their first job offer. They said that I had given them insufficient cause! They refused to take me and this sent me into a depression for many months. They had 80 percent of the computer market. But, looking back, it was one of the best things that could have hap­pened because I was stopped from being an IBM lifer. I’d probably be unemployed and selling Amway by now.

“My next job was at Systems Engineering Labs, for which I had to move from New Jersey down to Fort Lauderdale in Florida. They made real-time computers. The company got in trouble with their business with the space program, and I was out on the streets again in eight months.

“I moved up the coast to Palm Beach to start work at RCA. You probably don’t know that they used to be in the computer business. That’s because they pulled out of the computer business eight months after I joined. So you can see that my career did not start out with what you would call much success. The best thing about Florida is that it was where I first met the woman who became my wife.

“We moved completely across the country to Northern California. My new company was called Singer-Link, and it was a defense con­tractor that made flight simulators to train pilots. This was in 1971. Silicon Valley must have had at most 200 Indians. There were no Indian restaurants in the South Bay, and we had to drive 40 miles to Berkeley to eat at Pasand. When we couldn’t drive that far, we’d find grocery stores nearby that catered to Arabs. Baklava and halvah were close enough to our Punjabi desserts. We used to substitute Syrian bread for chapatis. Now, of course, everybody calls them pitas.

After three layoffs in two years, I had become hardened. I constantly interviewed with new companies to make sure that I had jobs handy, and I constantly turned each job down. I began to see jobs as a purely eco­nomic arrangement. Nobody was ever going to take me my surprise again. And it was nobody else’s responsibility to see that I was happy.

“I was at Singer-Link for nine years, and I rose to the top technical rank in the compa­ny, which was called Principal Engineer. I’m a very hands-on person, and I learned every­thing about the company’s products. Even so, I felt that I had peaked in my profession at age 34. r was earning $65,000 a year, which is the equivalent of an annual salary of $200,000 now. I looked around me and saw that all the other Principal Engineers were running small businesses on the side. Outside my industry, there wasn’t much respect for personnel who worked in defense companies.

“Not one of the 30 or so commercial companies that I interviewed with would hire me. One of them actually asked me if I was so smart, why I had worked at Singer-Link for so many years. I ended up taking a 30% pay cut just to be able to join a non-defense-related company. It was called Zilog. They made microprocessors and also dabbled in distrib­uted networking with a product called Z-net. At Zilog, I learned that I had a great aptitude for logic design.

“This was now 1981. There were a lot of companies beginning to build integrated computer workstations. These machines were tied together with what we now call ethernet. In those days, the term was DIX, which stood for Digital Intel Xerox, the companies that proposed the original standard. These work­stations all needed different protocols to interface with them. Multibus, VMEbus, Unibus, Q-bus, you name it, there was a com­pany that supported it.

“I teamed up with Indermohan Singh and Navin Jain to start a new company called Excelan. It was going to produce networking products to connect all these workstations together. We each kicked in $10,000 for the initial financing and organized a Friends and Family program to raise another $50,000. The only other Indian I knew of at the time who was an entrepreneur was Jaggi Tandon, down In Simi Valley, CA.

“We made very reliable products-both the adapter cards that fit into the back of the workstations and the software to make those workstations talk to each other. We re-sold transceivers that were made by an English company. These are connectors to tie togeth­er short pieces of cable. With these three offerings, we were the only company that could tie your mainframes, your PCs, and your workstations together.

“lnder was the CEO. He had gone to the Doon School and had a Ph.D. from Yale. He was sophisticated and impressed everybody who met him. I suffered from a speech imped­iment and was not good at public speaking at all. I was happy being the Vice President of Engineering and Operations.

“lnder was quite unfocused as a leader. He constantly churned out ideas and wanted to go after new markets before we even had a chance to get a dominant market share in an old one. We had a code phrase for his enthu­siasm. YAFO, we’d call it. Yet Another Fabulous Opportunity. Of course, the F didn’t stand for Fabulous.

“In the first year, we were supposed to bring in $3 mil­lion in revenue. We did half a million dollars. In the sec­ond year, the expectation was $12 million. We did $5 million. The venture capital­ists that had helped fund us were angry. They removed Inder from his post. He wanted both of us to resign in protest, but we were hav­ing none of it. I was made the acting President.

“We had 75 employees. We were down to our last million-and-a-half, and we were losing money at the rate of $200,000 a month. The venture capitalists told me to sit tight and not spend any money until they found a real President for the company. But I have never been one to sit on my hands. “We tried partnering with Microsoft from 1986 to 1988 but that relationship was one­sided and disastrous. Novell was the champion of networking then. I went to see Ray Noorda, the CEO, in early ’88 about a partnership. He said no because he didn’t see any customer need. But his cus­tomers started clamoring for our software by the end of 1988 and he came back. asking for a merger. We agreed. at a $200 million valua­tion. The merger was a rousing success Holders of EXLN shares got an eight-fold return in two years. I was made Chief Technical Officer at Novell, reporting to the President of the company. I was also given a board seat.

“I managed the release of the Netware release 3.11, which was the crown jewel in the company’s product line.

Those years until 1991 were good years. Steve Jobs, Scott McNealy, they all made a beeline to our head­quarters. We had a couple of terrific years. In 1992, Ray split the company in two, one piece concentrat­ing on Unix, and the other on Netware. He seemed to go back to his Mormon roots, wanting to help the state of Utah more than he wanted to help Novell shareholders. That was the beginning of a decline in the company’s fortunes.

‘There were three im­portant migrations happen­ing at the time-towards client-server, towards Inter­net-based software, and towards wide-area (instead of local-area) networks. Novell didn’t take any of these into consideration when designing Netware 4.0. Microsoft came in with Windows NT and began to take over.

“Ray Noorda retired, and he brought in Bob Frankenburg to replace him. Bob had had a successful stint making Hewlett-Packard a powerhouse in the PC business. Bob came from the consumer side of the business, and he concentrated on it to the detriment of the much larger corporate market where Novell had been so successful. He actually moved to Orem In Utah to be close to the miniscule WordPerfect business unit. Novell continued to decline.

“l had been traveling 200,000 miles a year for three years. Novell business meetings had consumed my life. I was physically exhausted and mentally burned out. I took a three­- month leave of absence in 1994 to get some long-delayed polyp surgery. I took my family to the south of France to recover my spirit. When I went back to work, it was with the conviction that I was ready to leave. Bob con­vinced me to stay on for six more months as a transition period. I left Novell in the first week of 1995.

“I plunged into a Bay Area organization called The Indus Entrepreneur. It was formed to help people from the Indian subcontinent get off the ground if they decide to found a company. I told members that I would offer two hours of my time to review any business plan shown to me in 1995. I must have looked at couple of hundred different business plans that year. I should have made my life easier by not getting involved in a company unless I also invested in it. You have to have some skin in the game.

‘1 oversaw the organization of the TiE conference in 1996 and was elected President of the organization this year. During this time, I have learned some things while talking to all those would-be entrepreneurs and looking at all those perfect-bound dreams. Some of the documents were brilliantly conceived but many, many of them had little hope of suc­cess. It is not enough to want something very badly. You must also be realistic.

“Entrepreneurs are born. You have to have that drive, that fire, within you. Don’t be disappointed if you aren’t the leader. Every locomotive has to have 10 grade cars and a caboose. Where would the world be without the grade cars and the caboose?

‘It’s a game for young people. I feel that I slowed down in my late forties. Things are moving so fast in high-technology nowadays that it will be a miracle if the current genera­tion can sustain the pace past their thirties.

‘What advice would I offer young people?

Learn to speak in public. I’ve got my son in debate class right in his early teenage years. Study economics and history they are the great educators. And always ask yourself what value you are adding to any situation. If you cannot honestly say that you are, then get cracking. Life is Darwinian. Step out of your comfort zone. You either grow or you die:­ByKanwal R£khi as told to Sukumar Ramanathan