One CEO's struggle against a culture of excess

Chip-maker Seeq Technology, launched in 1981 with bushels of venture capital, has had to learn the hard way that good business involves more than a parking lot full of German cars.

By the time the current CEO, Dan McCranie, took over, the money was gone and bankruptcy was imminent. His mission? Remake the way a whole company thinks about resources. -- E.O.W.

As the applause welled up around him and he surveyed the roomful of expectant faces, Dan McCranie said to himself: I'm not sure I'm ready for this.

He was -- had been -- a salesman. He knew nothing about finance. His engineering degree had collected 12 years of dust. And now they wanted him to run a company that made state-of-the-art semiconductors. This, he thought, was crazy.

J. Daniel McCranie had just been made the chief executive officer of Seeq Technology Inc., a San Jose, Calif., company of once-spectacular promise that now had come to rest at the brink of Chapter 11. Between 1981 and 1985, Seeq Technology had raised close to $70 million. But instead of saving Seeq when the company met the market, that lucre nearly killed it.

Gordon A. Campbell, Seeq's first CEO and a cofounder, was a charismatic leader who knew how to motivate people and run through cash. He put up a fancy building that looks more like an airport than a factory. Ray Charles entertained at the Christmas party one year. Beer busts, contests, and companywide celebrations were commonplace. So, too, were missed schedules.

Campbell's firing by his venture capitalists in October 1984 blew a hole in Seeq and depressurized the dream state in which the company languished. Campbell was followed by M. R. MacPherson, an engineer with no taste or flair for leadership. In nine months MacPherson was gone, to be followed by a three-man "office of the president," headed by one of the company's venture capitalists.

That was how Seeq had drifted down to this bittersweet day in March 1986, and Dan McCranie's ascension. To what? To the head of a comatose company that had used up $70 million in capital in just five years. A company that had proclaimed itself a leading-edge techno wonder but was now consigned to turning out commodity memory chips that cost more to make than they took in. In 1986, Seeq would lose $24 million on sales of $30 million.

Inside Seeq, the sense of destiny had died. McCranie knew that. He saw the people "sleepwalking toward oblivion." He spoke for 10 minutes that day. His message was charged yet simple; your basic Silicon Valley rallying cry. Seeq was a rocket at one end of a runway. At the other: a brick wall. The way to clear the wall was to stomp on the accelerator. If you hit the wall, fine. At least you'd go out in glory. His message, clichéd yet compelling, suggested at least that Seeq had a future, that redemption lay within its grasp. The people liked it.

The book on Dan McCranie had always been that he was a man of rough edges and thin skin. He swore too much to be a good CEO. When he got mad, it showed. Yet there were ambiguities about him, too. Yes, McCranie the salesman grabbed you by the throat and wouldn't let go -- but no one ever accused him of overselling. His self-confidence emerged in the way he spoke with punch and humor in front of a crowd. Those who knew him best remarked on his humility. McCranie was ambitious, but he had shied away from the CEO's job when it came open. It scared him.

Maybe in this newly anointed schizoid Seeq had found an idiot savant who could return it to glory. Dan McCranie wasn't sure. He saw his rise as an act of simple justice, a recognition that "one of Seeq's own, a worker," was getting a chance to lift the fallen angel. His appointment that day in March signaled a casting off, an exorcism even, of that which had come before: the reign of the profligate prince, the caretaker status of the well-meaning technocrat, the troika installed by an outside regime.

The company's fall from grace had come to serve as a parable of the place and time that spawned it. In the Silicon Valley of almost a decade ago, when Seeq was founded, the boulevards seemed lined with gold. Seeq's founders saw their path to riches assured even before they began. What they did instead was turn their company into an emblem of excess.

Such was the tattered legacy passed down to Dan McCranie.

To understand how Seeq ended up where it did in 1986, you have to return to 1981, when Seeq was founded by a group of Intel Corp. middle managers and engineers who claimed to have seen the future for memory chips. In the 1970s, with its development of the EPROM (erasable, programmable, read-only memory chip), Intel had come to own the program memory chip market. The natural follow-on to EPROMs were EEPROMS, electrically erasable memories. Dubbed E2 technology by the trade, it represented an elegant shortcut -- Seeq's founders believed -- that the world would want. Erasing and reprogramming the circuit on an EPROM required physically removing it from a circuit board and passing it for 20 minutes under an ultraviolet light. EEPROMs could be reprogrammed instantly in place with a five-volt power source, thus presenting the user with significant savings in time and labor.

On January 5, 1981, Campbell and his followers submitted their resignations, dismayed by Intel's lack of initiative in going after the E2 market. Seeq was incorporated a week later. That goosed Intel. Within a month, Intel sued.

Intel's aggressive move gave Seeq a charge. Gee, maybe we've really got something here. Seeq went on to raise a princely $50 million in venture capital and another $18 million from a public offering of stock. The reputations of its key backers, Kleiner Perkins Caufield & Byers and the Hillman Co., only heightened the blue-chip appeal. Gordon Campbell was Seeq's CEO. And Gordie, as everybody called him, was a real pied piper. People walked through walls to follow him.

When Seeq's founders went out to lease space, they found an adequate building, but a stockyard stood down the street. In the founders' minds, cow chips and the silicon variety didn't mix. So Seeq found more gently scented land, hired an architect, and raised a building with a swayback, cantilevered roof and soaring plate-glass windows. They filled it with modular furniture that cost three times as much as the basic bulletproof Steelcase stuff you usually saw around the Valley. Fussing over this monument cost Seeq six months getting to market. And it wasn't that long before the stockyard was gone.

The faithful at Seeq, meanwhile, buried a time capsule in the lawn out front. Among its contents was the prediction that by 1992 Seeq would be a $1-billion company. (Seeq did $45 million in sales last year.) Human-resources director R. Steven Gonia recalls Seeq's early days as a "glorious time. It was a high-profile place, intensely fast, very exciting." The best and the brightest wanted to work for Seeq, and Gonia was in hog heaven. "We could cherry-pick.'

Even outsiders caught the fever. In a national marketing campaign, the company offered a Porsche to the designer who could devise the most ingenious application of Seeq's E2 technology. Thousands responded. When informed that the ovens to bake the chips typically took about six weeks to install and get running, Seeq challenged the Thermco Systems Inc. crew: do the job in two weeks and we'll buy you dinner anywhere in the Bay Area. Soon the Thermco team was working nights and weekends. Recalls Gonia: "It worked so well the best Thermco guy became our maintenance manager. 'The hell with Thermco. I want to work for Seeq.' '

Dan McCranie didn't want to work for Seeq. The pied piper, however, had other plans. Campbell had sold chips alongside McCranie in Minnesota back in the mid-1970s. He contacted McCranie in early 1981, when Dan was working as a group vice-president of sales for Harris Semiconductor, in Melbourne, Fla. "I had barely heard of start-ups," recalls McCranie. Gordie to him was just another salesman. McCranie said no.

Campbell flew McCranie out to California to meet the team at Seeq and his lead venture capitalist, Frank Caufield. The answer, again, was no.

Weeks later, McCranie's phone rang -- Campbell again. He was flying out to find McCranie in Florida. McCranie kept shaking his head. Campbell was admiring a watercolor. Who painted it? McCranie's wife, Kathy, sighed: Dan did, back in the days when he had time to paint. No problem, said Gordie. If Dan comes to California, he'll have plenty of time to paint. Kathy perked up. Is that right?

In September 1981, McCranie was hired as Seeq's director of sales, with the express mission of cracking open the E2 market. He wondered what he'd done. He had taken a 40% pay cut for a minuscule equity position in a shaky start-up. But strangest of all, the company had no E2 chips for him to sell -- and wouldn't for another 18 months.

That didn't matter to Campbell. By 1983, E2 technology was the talk of the Valley. Sixteen companies had announced their intentions to build the chips. Dataquest, the leading research house in the Valley, fed the fever by issuing a report predicting that the E2 market would grow to $1 billion by 1987. But the technology was complex and held its secret close. An E2 still cost five times as much as an EPROM to make. The market for them remained small.

While Seeq was grappling with E2 technology, it needed to find a revenue stream. It fell back on EPROMs -- the passé product its founders had scorned Intel for dwelling on. EPROMs were fast turning into a commodity product, something best mass-produced by a giant like Intel, not a start-up like Seeq. Making EPROMS suggested a patricidal urge in the collective psyche of Seeq. Make one better than Intel. Blow Intel away and declare our independence. The time capsule, after all, had contained a prediction of what year Seeq would surpass Intel in sales.

Why not diversify the product line instead, find revenue in nonmemory technologies? That idea was debated within the company in 1983 and rejected. Campbell believed in betting on good people and letting their vote guide the group. The vote was to make memory chips. McCranie was torn. He was loyal to Campbell, but he was disgusted by the poor yields and ramp-up glitches the company was running into with EPROMs. This was supposed to be bread-and-butter stuff. If Seeq couldn't make a decent batch of EPROMs, how could it hope to execute with EEPROMs?

McCranie, meanwhile, was out on the road discovering the hard way how small the E2 market was. McCranie would jump on a jet and call on targeted customers. The typical response was: Who are you? What established companies are in this market? Good-bye. McCranie banged on Pitney Bowes, in Norwalk, Conn., for three years. (Today every Pitney Bowes postage meter has a Seeq chip in it.)

When McCranie dragged himself home to Seeq he found people living in a "fairyland." People didn't want to hear how tough the E2 market was to crack. They didn't want to confront the fact of the low yields on EPROMs. And the gilded life inside Seeq began to cloy. McCranie's father had been scarred by the Great Depression. He'd schooled his son in the virtues of thrift. At Seeq, thrift was a dirty word. "We were always rewarding ourselves for things that didn't count," recalls McCranie. "Every other week there was a champagne party for some weird thing -- the Tandem computer came in today. Well what does that mean? It means you've got to start paying the lease on it.'

While Silicon Valley toasted a booming EPROM market in 1984, Japan Inc. was cranking up to flood the world with chips. Then, just as suddenly, demand collapsed. Between the last quarter of 1984 and the third quarter of 1985, the price of a standard 128K EPROM went from $15 to $2. Seeq, reliant on EPROMS for 75% of its revenues, was making them for $5, and would struggle to get the cost down to $3. Life in fairyland was coming to an end.

When Seeq began to fray around the edges, the chairman of its board, Frank Caufield of Kleiner Perkins, asked for help. He invited one of his partners, E. Floyd Kvamme, a founder of National Semiconductor Corp., to attend a board meeting in September 1984 and critique Seeq's annual business plan. Kvamme, a tall, bespectacled man whose courtly manner and professorial presence veils a sharp technical mind, peered at Seeq's EPROM and saw an immediate flaw. The chip had a fancy quartz lid on it. Seeq had farmed this work out and thus lost leverage over 75% of the chip's cost. This kind of chip should have been made by a vertically integrated giant, not this gilt-edged boutique. The closer Kvamme looked, the more he saw the chip as emblematic of a top-heavy, misguided company. What was Seeq doing in a building big enough to block out the sun? Why did it think EPROM users would embrace the E2? These were two different markets. "The EPROM thing," concluded Kvamme, "was a disaster waiting to happen.'

But Gordie Campbell felt good in 1984. Seeq's sales were booming from $9 million to $43 million. The company would turn its first profit in the third quarter of the fiscal year. Campbell, for the first time in his career, felt as though he had arrived. He had a hot company and a hefty stock position. This was sweet.

Frank Caufield didn't think so. He wanted Seeq to stop burning money and start making it. From late '82 on he'd seen a succession of missed schedules, delays, and poor yields. Caufield knew where the company was weak -- operations. Come on, Gordie, let's change some of these people. Campbell dragged his feet.

Why was Caufield leaning on him? Maybe, thought Campbell, he had given his venture capitalists too much of the company. Why don't they go nurture their younger start-ups and stop tinkering with Seeq?

Campbell wanted the chairmanship. He saw his forte as reaching out to the larger world -- forging strategic alliances, raising even more money. He wanted to bring in an operations chief under him. OK, Caufield said, but he found Campbell's candidates potentially subservient.

Soon after the September meeting with Kvamme, Campbell played a higher card in his bid to be chairman. He submitted, then later withdrew, his resignation. Campbell had thrown down this gauntlet during a dicey time for high tech. A lot of high fliers that had gone public in '83 were going into the tank in '84. Kleiner Perkins had taken 10 companies public; not one had failed. With Seeq, the once brightest star, it wasn't really the money that mattered. Kleiner Perkins had already distributed its shares. The firm's name was on the line.

A few weeks later the board met at Seeq around noon. Campbell resurrected the resignation issue. It was accepted. By 2:00 p.m., on October 17, 1984, Campbell was out the door for good.

Seeq now drifted into the twilight zone. Campbell's successor, Bob MacPherson, Seeq's vice-president of operations, was a man who enjoyed poring over chip designs, not stirring up the troops.

For McCranie, who had come to see Seeq and Campbell as indivisible, Campbell's leaving cut the heart out of the place. But was that really true? McCranie himself had risen in an unspoken way to rival Campbell. It surfaced at company meetings in the form of lively repartee between the two men; Campbell challenging McCranie to sell every chip he made; McCranie chiding back that Campbell couldn't make enough. The employees warmed to this clash of egos. They found something to hew to in the form of the alternative set of values McCranie put forth. In McCranie, Gonia saw "a sense of reality." McCranie was tough but fair. He strived to meet schedules; he kept his word. He recognized individual effort. Campbell's was a more forgiving, elusive presence.

From January through June 1985, sales dropped to $17.2 million from $22.7 million for the same period the year before. Worse, losses rose to $7.8 million from $900,000. Layoffs ensued. They came in waves, only affirming the fear that the company was in a free-fall. People kept wondering when their numbers would be up. In July, MacPherson's was. He was let go and succeeded by a three-man office of the president that would operate on an interim basis: Floyd Kvamme, Gerald Robinson, and Dan McCranie.

The troika was an artifice, a stopgap imposed from outside while Seeq looked for a CEO. They spent a fortune on headhunters, but no one would touch this pariah.

Kvamme thought maybe McCranie could. The two had little in common; one the techno-philosopher, the other the gun for hire. Yet for all McCranie's bravura, a measure of doubt lay close to the surface. Kvamme liked that after Campbell's brazen assurance. The more he studied the man, the more he saw Seeq's next CEO.

McCranie's father's Depression experience had taught McCranie to play it safe, yet that was only one resonance from his past. McCranie's grandfather, a farmer, had traded commodities. Cotton futures. There was appetite for risk as well. McCranie's hunger grew when he confronted his private horror: ending up 60 years old, camped out in the lobby at Hughes Aircraft, trying to sell chips to "some 27-year-old snot-nosed engineer." It was time to jump at the job.

Dan McCranie is a dark-haired man of medium height and boyish charm, ardent about staying in shape. He walks on his hands to exercise his upper body, routinely covering 100 yards before the blood rushing to his head makes him dizzy.

When McCranie was thrust into the presidency of the company, all he had known was the business equivalent of putting one foot in front of the other -- sales. But selling Seeq out of its dilemma wasn't going to cut it. By May 1986, Seeq, the company that had raised nearly $70 million, had $13,000 in the bank. In its most recent quarter it had reported losses of $4 million on sales of just $8.5 million. Overhead was so high that for the year it would spend $54 million on the way to revenues of only $30 million. A rigorous restructuring was in order, but when McCranie looked in the mirror he saw a "fiscal midget." He needed to learn to walk on his hands again.

He closeted himself with books on finance and listened intently to the "tutorial" given him by the company's chief financial officer. McCranie got a few customers to pay their bills ahead of time to tide Seeq over, and then -- reflexively -- he hit the road, hat in hand, looking to sell as much of the company as people would pay cash to buy. "The Valley syndrome always had been: 'You get into trouble, you ask for more money.' " This time the ploy didn't work. "One investment banker told me my case was hopeless -- at any price." At least he was honest. A lot of the money men would string McCranie along, tell him maybe and tomorrow when they really meant no. That made him mad; they were wasting his time. He was more comfortable with customers. At least they spoke his language. Or used to. . . .

McCranie was livid upon receipt of an internal Xerox memo that discussed eliminating Seeq chips from Xerox products. He felt betrayed. He flew to Xerox and confronted management -- to no avail. After a similar Hewlett-Packard memo surfaced, McCranie fired off a letter to company president John Young, earning him an audience with HP's top engineers and a reprieve. McCranie had spent years cultivating these people, soliciting their loyalty, and now they were ducking him.

One of his best customers, Howard Charney, a vice-president of 3Com Corp., a computer networking systems company, saw McCranie's books every month. He was worried Seeq was going under and wanted to be the first to know. Charney saw the Valley as littered with victims -- people unable to guide their fates -- but McCranie was never among them. "No matter how bad things got, Dan always had a plan. 'OK, here's where we stand today on this. Here's what we're going to do about it.' '

What appeared to Charney as coherent action was McCranie's "dither theory" at work, by which McCranie proposed as many credible plans as entered his head. These ranged from selling more chips to new customers this quarter to selling the company. "It was manic depressive," he recalls about the summer of '86. "I'd have a victory and couldn't stand how happy it made me feel. Then something would go wrong and I'd wonder, why am I doing this? At one of these places I gave my pitch, and in the middle of it one of these guys was giving me a funny look. I said, 'What's wrong?' and he replied, 'To tell you the truth, Dan, we're not interested in buying your company. We never have been. We just wanted to see how you were holding up.''

Dan McCranie's bias all along had been to preserve Seeq's customer base at any cost; keep moving EPROMs. If he abandoned them, then there was no honor. Floyd Kvamme believed Seeq's EPROM customers couldn't have cared less about the company. To them, Seeq was no better than their number-four supplier of a commodity chip. What Seeq did have that was unique was its E2 technology. For this, people would pay.

At a key meeting back in mid-1985, Kvamme had said they weren't going to cut R&D; they were going to focus on getting out the 256K CMOS E2 chip. There were easier targets -- another type of technology or a lower-density chip. But Seeq needed a talisman that would signal revival.

In 1985, McCranie, then still just one-third of Seeq's interim "office of the president," hadn't seen what Kvamme was getting at. After all, one investor had come in, put his feet up on the conference table, and said that the only way to save the company was to cut R&D. But a year later, CEO McCranie shared Kvamme's vision. The 256K E2 was almost done. It looked like a winner. Close on its heels came a new family of breakthrough chips, the 128K and 512K Flash E2. These were single-transistor E2 chips, costing one-quarter of what the full-featured E2 cost.

But just when it looked as though the technology was coming around, the company was running out of money. Before the July Fourth weekend, after only three months at the helm, McCranie gave "another fire-and-brimstone speech on how to lower expenses and raise revenue" to his executive staff. But this time, like some candidate on the stump too long, he only "half believed" the words coming out of his mouth. McCranie thought Seeq was through. The vultures would descend and rip the ripening technology from the body.

As McCranie walked out of that meeting, his controller, Ralph Harms, followed him. Harms was a shy man, but now he spoke boldly. I don't think we've done enough, Dan. We have to do something really tough. What do we make money on? What don't we make money on? We have to think of this company with half as many people as before, he said. McCranie went home for the weekend and heeded this other quiet but insistent voice.

At its peak, Seeq had employed 700 people. By mid-1986 that number was down to 400 -- rock bottom, according to company dogma. The dogma also had it that Seeq could sell itself out of the EPROM disaster. But McCranie came back from the holiday and said the company would lay off another 130 people and get out of EPROMs. He told his managers how many people to cut and told them to do so this afternoon, not next week.

In the first couple of months after the layoffs, meetings were held to analyze why Seeq had screwed up so badly. Some turned tearful, others turned into shouting matches. When people said they had done their best, McCranie replied I know that, but that's not what I want to hear. I want to know how we can stem these losses.

In time, his people turned on him, accusing him of poor sales turns and being a lousy market forecaster. He protested. Then he realized he was reacting as defensively as everyone else in the room.

The rap on Campbell had been that he wasn't tough enough, and there was resistance on the board to giving McCranie the job for the same reason. He was a salesman, after all -- a glad hand for everyone. McCranie knew he had to expunge that perception; he had to atone for the sins of another's past. Besides, he was convinced Seeq needed the "hard negative feedback" it had never received; it needed to stop being a place where "people were afraid to criticize." That era began the day an operations supervisor stopped McCranie in the hall and said, "Boy, I never realized what an asshole you could be.'

McCranie thought, that makes two of us.

McCranie's commitment to being tough was connected to something more than Seeq's needs -- it was connected to his ego. When the money men snubbed him, he was hurt. He wanted to show them that though they didn't care about his company, he did. He would do whatever it took to avoid the mark of Cain he would carry through the Valley should Seeq fail. There goes Dan McCranie, the guy whose company died under him.

Restructuring the company also turned into a game that appealed to his tactician's instinct. Before him he saw a succession of wins and losses. How would he artfully play them off one another so as to ride a middle course? The July layoffs had cut Seeq's weekly payroll from $310,000 to less than $200,000. McCranie was able to keep his best people. That was nice. But morale took a hit. Experienced research-and-development technicians were relegated to rote work in wafer fabrication. As a result, productivity jumped. McCranie told these employees their sacrifice meant that the company would survive.

Others had told him the company would now crack. You can't squeeze any more blood from this stone, Dan. That intrigued him. McCranie spent a lot more time in the manufacturing area to better understand what he, the impatient and unschooled CEO, could demand of production. The move made him visible among the rank and file. Tough, yes, but maybe also caring. The company stopped buying new wafers for machine setup and replaced them with used ones. Maintenance on machines less crucial to the process was scaled back; the maintenance crew was reduced to working odd, staggered shifts that blurred the line between night and day. The employee badge numbers -- emblems of seniority -- were collapsed, eliminating this vestige of those who had come and gone. Here stood the hard core, the Seeq survivors.

These small changes began to offer the look and feel of a different company. That spilled over into McCranie's perception as well. The past and all its rosy optimism no longer mattered. What McCranie sought to do was "zero base" the operation. He called it "washing your brain." Each month McCranie told his people: we want to end up with more money than we started with. What can we bring in this month in revenue?

In mid-1986, McCranie figured the company could do $7.5 million per quarter in sales. And at that level Seeq could have positive cash flow, provided he could get overhead under control. That would take a big harpoon. Campbell's legacy was the rent. Leasing plant and capital equipment accounted for 35% of Seeq's total outlays, about $1 million a month. Assuming sales of $7.5 million a quarter, McCranie calculated that even if he got direct labor and materials free, Seeq would still lose money at the rate of $4 million a year. He needed a break from his lessors. Otherwise, everything else would mean nothing.

Many of Seeq's leases had been signed when interest rates were touching 20%. Some were for as long as seven years. Yet the customer base for E2 and Ethernet chips was solid. Flash was coming on. The industry was in a slump. Good timing. If lessors yanked their equipment out of Seeq, what were they going to do with the stuff? Sell it to the navy for anchors? McCranie figured they'd be lucky to get 10¢ on the dollar. They decided to ask for an extension of the leases, effectively reducing payments by 50%.

Seeq had six major lessors, some of them more sympathetic to the company's plight than others. Any one of them could scuttle the deal. McCranie decided to work the group, rounding up the more compliant ones and moving them in the direction of the more recalcitrant.

A series of individual meetings in the summer led to a final meeting with all the lessors, most of whom by then were on board. At that meeting McCranie arrived with two sidemen, Seeq's corporate counsel and its "junkyard dog" insolvency counsel. One bore the bankruptcy papers, the other the restructured lease agreements. McCranie's lessors were free to choose which set of documents they wanted to sign.

McCranie knew that to pull this deal off, they had to come up with a carrot. That they did, in the form of warrants allowing the lessors to buy stock in the company. McCranie reasoned that that was only fair. Seeq had signed those leases and was asking for forgiveness.

By mid-October the lessors had fallen into line. That made other pieces fit. Seeq's early investors, General Electric Ventures, Kleiner Perkins, and the Hillman Co., agreed to give Seeq an additional $2 million convertible into preferred stock, in the hope that the company could restructure successfully. In November, Monolithic Memories Inc., another chip maker, bought approximately 14% of Seeq for $4 million and agreed to a four-year joint-technology program. This gave Seeq not only cash but a way to diversify the E2 technology. (A similar deal was made last year in which Seeq exchanged the Flash technology for access to National Semiconductor's manufacturing muscle. Seeq started to ship Flash chips in December, and it currently has about a two-year lead over its competitors in the Flash market.) The CMOS 256K E2 chip started selling briskly in the last quarter of the fiscal year, which ended September 30. In fiscal year 1987, it accounted for 25% of Seeq sales.

In late 1986, McCranie told his board he expected the company to turn a profit in the third quarter of calendar 1987. It started making money in the first quarter and has done so ever since. In fiscal year 1987, Seeq earned 22¢ a share, compared with a loss of $9.48 a share the previous year. It now has $33 million in the bank.

The chairs in the Seeq boardroom are a soft, glovelike leather. Spaced around the expansive slab of a table are enough of them, it seems, to seat the entire U.S. Cabinet. Late on a Friday afternoon Dan McCranie sits alone at the table, dwarfed by the space. "The biggest worry I have is how many guys with a sales and marketing background are at the top of very large, successful semiconductor companies. This is a fast-moving situation. There are new processes coming along all the time. If something downstairs doesn't work, I'm useless. I'm useless, other than marshaling the forces to fix it."

McCranie's wife calls, wondering when he is going to be home. He tells her soon, says "I love you, baby," and hangs up. I tell McCranie I'm out of questions, giving him an out. He seems to want to stay and talk. He is leaving for Hawaii tomorrow for a week, his first family vacation since the turnaround two years ago. He says he intends to beat his daughter every morning at tennis. He has arranged to be part of a conference call on Friday. The man won't let go. Or can't.

Faster and faster he must now run. In its first seven years, Seeq built 12 basic chip designs. In the next two, it will introduce 18. There now arise again the expectations of investors, employees, customers. There a