John Edmond, one of the six founders of Durham, North Carolina, based Cree Research, recalled that it was on December 12, 1993, when he and his colleagues saw the article in Electronic Engineering Times reporting the sensational news that Nichia had achieved a candela-class bright blue LED. “We were like, Holy shit!” Edmond said, sarcastically, “this is a great Christmas present!”
As by far the world's leading maker of blue LEDs, Cree was the company most affected by the Nichia announcement. In times of crisis, you have to move fast. Two days later, Edmond and his colleagues were sitting in a conference room in a Tokyo hotel. Facing them across the table were Nichia's top management and Shuji Nakamura. Edmond had managed to organize an audience with Nichia on short notice because he knew Nakamura from going to the same conferences together. The Americans were eager to see Shuji's new device. When one was produced and illuminated, the visitors gasped in astonishment. They were hardly able believe their eyes. “It was fifty times brighter than our brightest stuff.”
Edmond and company countered by proposing a collaboration with their Japanese counterparts. Cree was also working on gallium nitride. Within six months they would have a production-grade device that would be just as bright. It made sense to work together. The Americans could supply Nichia with silicon carbide substrates. But the Nichia folks were simply not interested. “They were like, gaijin [foreigner], go home. It was just, How many LEDs do you want to buy?” None, the Americans spluttered, we don't want to buy any. “But they didn't care what we said, they didn't care that we were working on nitrides.”
The Cree team returned home stunned by what they had seen and heard. At the same time, however, the flat refusal on the part of the Japanese to collaborate also served to pump the Americans up, to stiffen their resolve. “So it was a shock on the one hand, but it was like, we're going in the right direction, we're going to crush them in the end, we've just got to keep working,” Edmond recalled. “We were young and stupid…and very confident.” A resilient ego was ever a necessary precondition for successful entrepreneurs.
Edmond had predicted that Cree would come up with a competitive LED in six months. In fact, it took the Americans a year and six months to release their first gallium nitride product. “That was in June ’95, and it was a crappy device. It was maybe half the brightness of the Nichia one, and Nichia had made improvements since then. But at least half the brightness was better than one hundredth or whatever it was.”
Meanwhile, in their determination to catch up with and crush Nichia, the Americans had also devised another strategy. They would try to poach Nichia's star researcher. At the end of 1994, Edmond sent his Japanese counterpart a letter inviting him to come and join them. In it, he wrote that Cree would double Shuji's salary. More importantly, the US company, which had gone public the previous year, also offered him generous stock options. Had Nakamura accepted the options then, they would in due course have made him a very rich man.
Trouble was, Shuji did not know what a stock option was. This is not so surprising, given that stock options would be illegal in Japan until 1997. He went home and asked his wife whether she had any idea what “stock option” meant, but Hiroko didn't know either. At the time Nakamura was so busy working on developing a blue laser that he neglected to look up the meaning of the term. When Edmond called from the United States to see whether Nakamura was interested in joining Cree, Shuji politely declined his offer.
Over the next six years, Cree would make repeated attempts to persuade Nakamura to leave Nichia and join them. Eventually, at least to some extent, they would succeed.
Cree Research was founded by two brothers, Eric and Neal Hunter. They hailed from the small town of Boone, which is named after the famous frontiersman Daniel Boone. The town is located in the beautiful Blue Ridge Mountains of northwestern North Carolina, a range that includes the highest peaks east of the Rockies. The brothers attended Watauga High School, where they both competed on the same swim team.
On first hearing the name Cree, I imagined that it derived from the Native American tribe of the same name. Many years later I discovered that the Cree live in Alberta, Canada, which is a long way from North Carolina. In fact, Cree is an abbreviated form of the Scottish family name McCree, also spelled McCrae. It happened to be the middle name of the Hunter brothers’ late father, a doctor in whose memory they named their venture.
The Hunters both attended North Carolina State University, one of the three schools that form the triangle that gives Raleigh-Durham's Research Triangle Park its name. Neal was a mechanical engineer who, on graduation, got a job as a salesman with a big control systems company. Described by a high school friend as “very smart, very athletic, and very competitive,” Neal quickly discovered that he did not like working for other people: at heart he was an entrepreneur.
In his older brother, Eric, the entrepreneurial tendency was even more pronounced. Bob Davis, his professor at NCSU's graduate school, observed that Eric had wanted to form a company from the day he arrived. “You could tell he was an entrepreneur,” John Edmond recalled. “He was always on the phone in the lab, talking to brokers, buying and selling stock. There'd be a crystal growth run going on over here, and I'm like, What the hell is he doing—doesn't he have better things to do?”
The specialty of the Davis lab was silicon carbide, a compound semiconductor material. Silicon carbide does not exist in nature; it was first synthesized in 1808. For most of its history this substance, which is made by reacting sand and coke in a furnace, had been used in abrasives. As we have seen, it was the first material in which electroluminescence was observed. Then, in the 1950s, it was realized that in the form of electronic devices like transistors, silicon carbide might be able to go places where plain old silicon could not, withstanding higher temperatures, or operating at higher powers and frequencies.
But how to grow bulk crystals of the recalcitrant material, which you could slice up into wafers and fabricate devices with? This presented a formidable challenge. For one thing, you have to heat the stuff up to searing temperatures. For another, when you heat silicon carbide, unlike most materials, it doesn't melt. It sublimes. In other words, the compound jumps straight from being a solid to a gas, just like dry ice. Then, when it crystallizes, silicon carbide has 177 different structures that it wants to form. Controlling the growth process so that the crazy compound crystallizes into the structure you want is extremely difficult.
Nonetheless, that is what over ten years Bob Davis and his students at NCSU, notably Calvin Carter, managed to do. They became the first group to grow bulk silicon carbide, ahead of several large corporations, including Westinghouse. The crystal material they made was sliced into tiny, fingernail-sized wafers on which devices could be grown. “We would take the wafers and divvy them up between us,” Warren Weeks, who worked at the Davis lab in the early 1990s, recalled. “They were as precious as you could get, we'd break them up into little bitty pieces and try to make them last as long as we could.”
In July 1987, after a meeting at a biscuits restaurant in Boone, Eric and Neal Hunter decided to form Cree Research to commercialize the work. They licensed ten patents from the university. In return Cree issued the university shares in the company. The Hunters took out second mortgages on their homes and raised $25,000. As their corporate headquarters, they moved two desks into a poky eight-by-thirteen-foot office.
John Edmond, a member of the Davis group whose specialty was ion implantation—a high-tech method for adding dopants to semiconductors—was the first full-time employee the Hunters hired. “They said, We're going to give you an advance,” Edmond recalled. “I said, Man, that's good! Because being a graduate student, I didn't have any money. So they went and maxed out their credit cards and gave me a check for 4.2 months of salary. And that's why I came on board.” Two other members of the Davis lab, Calvin Carter and John Palmour, joined shortly afterward.
The rangy, intense Edmond was the only one of the group who was not local. Indeed, so far from being a Tar Heel, he was actually a Northerner, from Cohocton, a town of nine hundred residents in western New York State. A self-confessed propeller head and science nut, Edmond went to the New York State College of Ceramics at Alfred University. Neal Hunter would tease him by referring to his alma mater as “the fighting cement mixers of Alfred.” As the winner of a scholarship from the Office of Naval Research, Edmond could have gone to any graduate school in the country, but, attracted by Davis's reputation, he ended up at NCSU. “I was interested in a ceramic material that could be used in electronics. They were working on silicon carbide. I thought, Silicon carbide? That's a ceramic, they use it for abrasives, grinding wheels, sandpapers; if it's a semiconductor, that's perfect, that's what I want to work on.”
In 1987 Edmond was twenty-six years old, Neal Hunter just twenty-five. Chances of success in founding a start-up straight out of university are not good. But, as Hunter later put it, “I was too young and too stupid to know.” As with every start-up, securing funding proved a major initial obstacle. Cree Research began with four contracts from the Defense Department's Small Business Innovative Research program. But just weeks after Cree's formation came Black Monday, October 19, 1987, the largest one-day crash in the US stock market's history. Venture capital virtually dried up. But the company was rescued from oblivion by an unlikely savior. The Hunters tapped a local McDonald's franchisee who was dating a cousin of theirs. The franchisee brought in some of his friends from McDonald's as angel investors. “They were a great group of guys; they believed in us,” Edmond said. “They put the money there. They made gazillions of dollars on Cree, because they put a lot into the company early on.”
With cash in hand, the question for the fledgling firm now was, What to make using their silicon carbide expertise? Fancy devices for use in high-power switching and high-frequency communications would take time to develop. In the short term Cree needed a cash cow, a product for which there was an existing market. The Hunters approached Edmond. “They said, OK John, we want you to make a blue LED, that's going to be our first product. I said, What the hell's a blue LED?” Eric Hunter told him not to worry. “It's easy,” he reassured Edmond. “It's like making donuts!”
Unlike Nakamura, Cree had chosen to do blue not because silicon carbide was the best material for the application, but because silicon carbide was what they knew how to make. At that time, the reputation of blue light emitting diodes among LED makers was, to put it mildly, not good. “I always considered blue LEDs a black hole,” former Hewlett-Packard R&D manager Roland Haitz told me. “Around the late eighties I could be quoted as having said that I was proud that HP had not spent a single dollar on blue LED development.” As far as Haitz was concerned, for a light emitter, blue was an “unnatural act.”
With word about recent Japanese developments in gallium nitride beginning to filter through, however, that negative attitude was changing, albeit slowly. Edmond recalled visiting George Craford at HP around 1989. “They had one guy who had been working a little bit with blue LEDs, but they were like, These things will never work, you know. And obviously they knew that silicon carbide wouldn't work.”
Edmond began his attempt to figure out how to go about making blue silicon carbide LEDs by combing through the literature, to find out who had done the best work. The answer turned out to be the giant German firm, Siemens. To make their devices, Siemens employed a primitive process. From the blowholes of the giant furnaces they used to make silicon carbide abrasives, they would go in and chisel out tiny crystals of dark black material. These crystals would be the basis for making LEDs using the traditional method of vapor phase epitaxy. The LEDs thus produced shone blue, but they were so dim you could barely see them. The devices were sold individually, for scientific purposes, as a spectral reference for blue light. Produced in minute quantities, they were very expensive. Edmond remembered ordering two Siemens LEDs from Hallmark Electronics for seventy-five dollars apiece. In 1988 the Japanese consumer electronics firm Sanyo came out with a silicon carbide LED that was brighter than the Siemens device. It sold for about ten dollars. “We looked at what they were doing, and it was a vapor phase epitaxy process also,” Edmond recalled. “We said, This is just stupid, we can't do that—let's do something different.”
Cree's experience was with the new high-precision growth method MOCVD, the very same technique that Shuji Nakamura was coincidently around this time learning nearby at the University of Florida. One of the company's first employees, H. S. Kong, a native of Shanghai, had mastered MOCVD while working in the Davis lab at NCSU. In fact, since the precursor materials he was using (silane and methane) do not count as metal organics, the method was really just a variant of vapor phase epitaxy. It required extremely high temperatures of around 1,700 degrees Celsius, higher than the melting point of quartz. Somehow Cree's crystal growers figured out a way to handle this insane process.
The Cree team mustered in-house all the other skills required to succeed in the LED business. Tommy Coleman, one of the company's six founders, was a hands-on equipment builder. Carter was the bulk silicon carbide crystal growth expert, Edmond and Palmour the device fabricators, Neal Hunter the salesman. Starting from scratch, it was a case of believing in themselves, believing that they could make blue silicon carbide LEDs, and then doing what they knew how to do. But it was a bumpy ride. “We had several months, and many times when we couldn't grow a crystal,” Edmond recalled. “Then we'd say, OK Calvin—you're not in charge of crystal growth anymore, Neal's in charge! So Neal would go in and he would do just this wild stuff. I mean, we've got pictures of crystals he grew that were great looking; they were dog-shit wafers, but they looked cool. For example, the old egg crystal, it looked like an egg, like it had a yolk. It was unbelievable.”
“We would challenge each other. I worked in the polishing and cutting and grinding of the crystals, all sorts of things…we did everything. We would interact, we had a lot of passion, which I think is the key to success: we loved what we did, we believed in what we could do, and you'd just go in there and do it. In crystal growth there's a lot of black magic—I mean, just weird things going on.”
There would be dark days when, in desperation, they would be forced to fall back on their last resort—the magic power of a rubber chicken. When all else failed, they would grab the dummy bird by its neck and swing it in an incantatory gesture over the machine that they were having trouble with, all the while chanting the mantra, “C'mon big money!” The demons having thus been cast out, the thing would start working again. Or at any rate, the cause of the problem would have been identified. Such fun rituals became part of the close-knit Cree culture.
The team worked together and played together. Out back of their building was a basketball court where they would come off work and shoot hoops or play pick-up games. Here, too, their fiercely competitive nature would be in evidence. “Those guys really want to win,” said Warren Weeks, the crystal grower from the Davis group, who worked for Cree in the mid-1990s. They drove themselves hard. Their employees, too: “I was only at Cree for two years,” Weeks recalled, “but they were really dog years. It felt like fourteen!”
Working close together under great pressure saw lasting bonds of friendship and respect develop between the founders. “We were like brothers, a really good family,” Edmond recalled. “You'd get here early in the morning, you'd eat lunch together, you'd have dinner together, you'd come back here and work till eight or nine o'clock, then you'd go home, and you'd do it seven days a week. There was no day off, for years.” Like Nakamura, they were driven. “You've got to be self-motivated; you've got to love what you do.”
This relationship would stand the test of time. When I met Edmond at Cree in late 2005, he was just about to go on vacation with some of the other founding members. He had recently bought land to build a house out at Colvard Farms, a rustic six-hundred-acre parcel of land nearby that Neal Hunter had converted into an upscale subdivision, characteristically naming the development after his maternal grandfather. “Neal's out there, Calvin's out there, I'm out there. We really are a close group. We were in school together, we grew the company together, we're all going to be living next to one another—it's going to be kind of fun.”
In October 1989 Cree introduced its first commercial blue LED. “It was much better than the Siemens and Sanyo devices, and we were selling it for a dollar,” said Edmond. By this time, however, the company was almost broke, with just a month's worth of cash left. They were “about ready to starve,” as Edmond put it, when an unexpected savior arrived on their doorstep in the shape of representatives from the giant Osaka-based trading company Sumitomo Corporation.
Japan's sogo shosha—general trading companies—comb the world looking for unique new products to introduce to the Japanese market via their extensive distribution networks back home. Blue LEDs were not exactly unique, but as we have seen, there were only two rival products, both of which were much more expensive than Cree's dollar-a-pop devices and not nearly as bright. In 1990 Sumitomo signed up the tiny US firm—Cree then had only about twenty employees—to a Japanese distributorship agreement worth one million dollars. It was a match made in heaven, a relationship that flourishes to this day. Sumitomo would become Cree's biggest customer, accounting for about a third of the company's overall sales. That initial agreement would be followed by a series of others, the most recent at the time of writing being a 2005 purchase of a massive $200 million worth of bright blue LED chips.
Some huge ironies here. As it happened, in 1990, the year that Cree and Sumitomo got together, I visited Research Triangle Park. I was there researching an article on what former Harvard professor Robert Reich had called “the rise of techno-nationalism.” That is, a desire to exclude Japan as punishment for what critics saw as Japanese companies having taken a free ride on US science and technology. The Japanese were manufacturing high-tech products, such as computer memory chips, undercutting American firms and driving them out of business. US R&D consortia based on the supposed Japanese model of government-industry collaboration were springing up. One was Sematech, a chip-manufacturing equipment consortium based outside Austin, Texas. This nationalistic organization would not even entertain visits by Japanese journalists, let alone membership by Japanese companies. Predictably, however, techno-nationalism was most prevalent in Washington, DC.
What I discovered at Research Triangle Park, only a hundred or so miles south of the nation's capital, was a very different story. There, in a state that then ranked thirty-fourth in terms of per capita income, the Japanese were welcomed with open arms as a source of much-needed investment and jobs. Several Japanese firms had set up shop in the park. They included Mitsubishi Semiconductor, a chipmaker, and Sumitomo Electric, another branch of the conglomerate, which produced optic fiber. I also found other forms of benign Japanese interest. They included the sponsorship by Kobe Steel of a chair at North Carolina State University. The chair was then occupied by a professor named . . . Bob Davis.
In those days, there was not yet much evidence at Research Triangle Park of an entrepreneurial culture. Cree would be North Carolina's first great start-up success. And it was Sumitomo, a Japanese firm, which initially underwrote that success.
A final irony, for those disposed to notice such things, was the fact that in the United States, the land of the rugged individual, Cree was as we have seen very much a team outfit, with members dividing up various tasks among themselves. Whereas Japan, a notoriously group-oriented country, had produced Shuji Nakamura, a lone researcher working independently, doing almost everything by himself. So much for national stereotypes.
The initial customers for Cree's blue LEDs in Japan were mostly themselves LED makers. For example, Stanley Electric built the first large full-color LED display screen, which it unveiled at the 1992 Electronics Show in Tokyo. The screen was not very bright, but it was good enough for indoor applications. This display was in some sense an early vindication for Jim Tietjen's vision at RCA back in the late 1960s of an LED television that could be hung on a wall. A few such screens were sold, one of them to the new international airport in Seoul.
Pachinko parlors, those gilded palaces of (venal) sin that are to pinball-loving Japanese what slot-machine casinos are to Westerners, would become one of the largest markets for Cree's chips as the signage on the parlors transitioned from gaudy neon to even brighter colored LEDs. But that would come later.
The reason Cree was able to crack the Japanese market in the first instance was that, Sanyo having dropped out, there was simply no other source of blue LEDs. The American devices were it. In April 1992, however, Edmond received his first intimation that this happy monopoly was not indefinitely sustainable. He was one of the handful of attendees at the first gallium nitride workshop in St. Louis. This, as we have seen, served as Nakamura's US coming-out party. “Shuji showed his LEDs and he compared them to ours, because there wasn't anything better. His were still way down here [in terms of brightness], and ours were way up there, and it was so funny because our LEDs were the benchmark, an indirect-bandgap semiconductor.”
It was not so funny the following year, when Edmond and Nakamura met up again, at another conference. “He said, I've got something that I can't talk about because my boss won't let me, but it'll blow your stuff away. And I go, Yeah, sure Shuji. And he goes, I'm telling you.”
Edmond took the hint. Especially after he bought some Akasaki-style gallium nitride MIS devices made by Toyoda Gosei that were several times brighter than Cree's best blue LEDs. A grant from the Defense Department to work on nitrides for blue lasers helped kick-start gallium nitride research at Cree. “We were working hot and heavy on gallium nitride, because it's a direct-bandgap material, so it's got to be better.” Then, at the end of 1993, came Nakamura's shattering announcement of candela-class bright blue LEDs.
Luckily for Cree, there would be a period of grace in which to sell its dim-but-cheap blue silicon carbide LEDs while Nichia ramped up production volumes. During this time, at five dollars each, bright gallium nitride devices were still comparatively much more expensive. In 1994 the US firm shipped seven million silicon carbide devices, accounting for 99.9 percent of the market. Applications ranged from Stanley's indoor displays to the indicator lights on Crown audio amplifiers.
By June 1995 Cree had developed its first GaN devices. In the rush to meet the challenge from Nichia, it was suggested they abandon their trademark silicon carbide in favor of the sapphire substrates that everyone else was using. Edmond was determined to stay with SiC. It was partly a matter of pride, partly the not-invented-here syndrome. He would be damned if Cree's devices were just like Nichia's. Grafting the two materials onto each other turned out to be a nightmare—though the lattice match between gallium nitride and silicon carbide is much closer than between gallium nitride and sapphire, GaN will not grow on SiC substrates—but eventually Cree's researchers discovered a wonderful trick that saved the day. For years afterward Cree would continue to describe its products as “silicon carbide LEDs.” In fact, however, they were really gallium nitride devices grown on silicon carbide substrates. Cree was the first, and would remain the largest, US manufacturer of gallium nitride devices.
Cree's first-generation gallium nitride devices had an unfortunate tendency to burn out. By 1996 Edmond had fixed the problem. In that year the company got a really big break. The president of the German carmaker Volkswagen decided that, in order to differentiate its vehicles, the dashboard lighting on them should be blue. He approached Siemens, who in turn tapped Cree, who the Germans knew through the silicon carbide connection.
In addition to differentiation, LEDs offer other advantages for car interiors. They let carmakers shrink the thickness of the instrument cluster in the dashboard. Incandescent bulbs are intrinsically big; they also emit a lot of heat. Thus, if positioned too close to the plastic parts they are illuminating, over time the plastic distorts and melts. Some space must be left between bulbs and instruments. But as carmakers try to cram in ever more features, airbags and what have you, space becomes a premium. Saving an inch or so in the thickness of the dashboard is thus a significant benefit.
Interior lighting for cars was a huge application; in fact, it drove the market. By 1997 Cree was shipping 75 percent of its LED production to Siemens for use in VWs. The devices, which Cree still makes today, were simple gallium nitride blue light emitters, blue being VW's signature color. Cree focused on producing these low-brightness devices because that was what their customer wanted, and the US firm was keen to comply with VW's requirement. Meanwhile, in Japan, Nichia had begun manufacturing complex high-brightness blue devices with indium gallium nitride quantum wells. Edmond tried desperately to catch up with Cree's Japanese rival, but to no avail.
“They were always a step ahead. We'd come up with an indium gallium nitride device that was about the same. Then they'd come up with one that was brighter again. Back in the late nineties I gave several presentations at LED conferences, and I always followed Shuji. It really pissed me off; it became a running joke because [the organizers] would always put me right after Shuji. So he would come up with these new devices that would, like, blind you. And I'm sitting back there saying, Oh God, I've got to follow him. And he did it to me over and over again.”
During these years, Edmond had many interactions with Nakamura. He made several attempts to hire Nichia's star researcher. But Shuji would refuse, always having an excuse. For example, his wife's mother was elderly, and they couldn't leave her behind in Japan. Then, in October 1999, Nakamura happened to attend a conference at Research Triangle Park. On a visit to Cree, Shuji made it clear that he was unhappy at Nichia. “We were like, Hmm, that could be interesting,” Edmond recalled. “Then we started talking seriously about getting him out of there.”
How things panned out in Shuji's departure from Japan and the subsequent, dramatic developments is the subject of part 3. Then, in part 4, we investigate how the transition from conventional to solid-state lighting will take place, how LEDs will replace the lightbulb.
This is a process in which Edmond is determined to participate. “Ten years from now I want to see the incandescent gone. It's going to have to be cheap…but we can get there, it can be done.” Now multimillionaires, he and his colleagues no longer have to work, but nonetheless they still do. “Why? Because we love it. I'm more excited about Cree than I ever have been. I want to replace these,” Edmond told me, pointing at the fluorescent strip lights in the ceiling of the conference room at Cree's headquarters in Durham.
For that to happen, considerable increases are required in the conversion efficiency of LEDs, up from around 50 percent today. “I believe 75 percent efficiency will happen in the next five years. We've already demonstrated 100 lumens per watt on a smaller chip.” Lumens per watt being the lighting industry's equivalent of miles per gallon.1 Today, the vast majority of Cree's revenue comes from such small chips used for applications like cell phone backlights. But the company's goal is to move toward larger “power” chips for use in general illumination. At 70 lumens per watt, LEDs can compete directly with compact fluorescent lamps. At 100 lumens, they can take on fluorescent tubes. “My objective is to get to 100 lumens per watt on a power chip. I think that would set things on fire, personally. That's where it has to be to really git things crankin’.”
“It's going to be amazing, what's going to happen next—with headlights in cars, backlights in LCDs, general illumination, I mean, it's just the beginning. So I want to see this continue, our baby [Cree] is only eighteen years old as far as I'm concerned. I've got several years left in me. I'm forty-four, not ready to retire yet. I want to see us at a point where you can go into Lowe's and buy a light that's all LEDs, and it's putting out 1,500 lumens and it's using 15 watts of energy. When that happens, I probably will take some time off.”
In 2004 Cree announced that it had officially entered the race to replace the lightbulb. That year, the company posted annual sales of $390 million, with LEDs accounting for 82 percent of the total. The company was worth $1.7 billion and employed over 1,300 people. But it was still true to its roots. “Even though they have 1,400 people over there, they still operate like a start-up,” said Bernd Keller, manager of the company's Santa Barbara Technology Center. “They're very opportunistic, very aggressive, very high speed; that has sustained the company and made it what it is today.”
In April 2005 Neal Hunter, at forty-three, resigned as president of Cree, “to pursue other opportunities.” These, as we have seen, included property development. But building and selling houses was never going to be enough to satisfy Neal Hunter's entrepreneurial urges. Six months later, in October 2005, he and three former Cree cohorts announced the launch of a new company, named LED Lighting Fixtures. The start-up's immodest goal: to take on what it saw as the big names in the lighting industry, GE, Philips, and Osram, for a chunk of the $40 billion global lighting market.
“I think it's time to challenge them,” said Neal Hunter, who is not known for being the most patient of people. Such firms had no incentive to innovate so long as consumers keep buying lightbulbs. The consumer market for LED lighting was not moving fast enough for his liking. Hunter's new company would help accelerate the transition from lightbulb to light emitting diode. Needless to say, its lighting fixtures would use LEDs supplied by Cree.
In April 2006 the start-up announced its first lighting fixture, a recessed can luminaire that produced, according to independent tests, 73 lumens per watt. The fixture operated using less than 15 percent of the power of a comparable incandescent bulb and 50 percent of a compact fluorescent lamp. Drawing fewer than 10 watts, the lamp was cool to the touch. The company claimed that it could provide fifty thousand hours of light—enough for twenty years under normal usage of five to six hours per day—compared to two thousand by incandescents. By replacing incandescents with LEDs, consumers would save seven dollars per fixture in annual energy costs. The fixtures would pay for themselves in less than two years.
“LED Lighting Fixtures’ technological advances shatter conventional thinking regarding the projected timeline and quality of light offered by LEDs for general lighting applications,” the announcement claimed. “This revolutionary reinvention of light could have global implications for decades to come.” Production and distribution of the new fixture, which would be manufactured in China, was slated to begin by the end of 2006.
1. A lumen equals the amount of light given out by a one-candela source radiating equally in all directions.