(FORTUNE Magazine) – IT WAS A BITTERLY COLD DECEMBER WEEKEND, WITH THE kind of frigid wind that would keep most sensible people indoors. But 33-year-old Jason Chen was going to throw his party anyway. He had rented a large room at the Beijing Library, a 15-minute cab ride west of Tiananmen Square, and had carefully arranged row upon row of personal computers for inspection by his guests.

As Intel's top sales guy in China, Chen had become quite expert at drawing crowds, sometimes hosting as many as 800 business types at his Meet Intel seminars. But this two-day session was to be a PC party, pitched directly at consumers, and after weeks of advertising in local newspapers, he was counting on as many as 7,000 showing up. His goal: to give Chinese consumers a taste of an American-style shopping experience, where you can handle the monitors, play around on the keyboards, maybe even buy. At the time, shoppers weren't permitted by local storeowners even to touch PCs before purchasing them.

The moment Chen threw open the doors on that morning in 1995, he understood that Intel's business in China was going to be a hit. There, in front of him, thousands were waiting patiently in line, and over the course of the weekend 45,000 people--enough to fill a baseball stadium--attended Chen's party. "It was incredible," says Chen, an intense man given to bursts of exuberance, who now, at 42, is one of Intel's youngest vice presidents, overseeing worldwide sales. "Almost everyone bought a PC that day, even though the room was so crowded many were forced to sit on the floor to fill in order forms."

The number of computers Chen sold astonished him. After all, even a no-name Chinese-made PC cost about $1,200--a year's wages for the average Chinese worker. Two weeks after the event, Chen's boss, Paul Otellini, then head of global sales and now Intel's president, visited China. "You are creating history--what do you need to do one of these a week?" he asked Chen.

The party was a watershed, not just for a rising young star but for Intel as a whole. Like many foreign companies, it had spent a frustrating decade trying to gain a foothold in the world's most populous country. But Intel's aggressive efforts in China would blossom into a huge foreign-investment success story. Today China is indispensable to the company's future--and to Otellini's success as its next CEO.

THE CHIP GIANT opened its first Beijing office in 1986, during the initial flurry of foreign investment in China. But the country's ambivalence toward capitalism and democracy, culminating in the Tiananmen Square massacre in 1989, put a damper on Intel's and the rest of corporate America's ambitions. Still, Intel--the company that made paranoia an asset, not a liability--always knew China would be important. Craig Barrett, Intel's CEO (and COO at the time), had been going to Asia since the 1970s and believed China would eventually understand that it would benefit from joining the world infotech community. "You had a government that was composed of a bunch of technocrats who recognized the importance of technology to the future of the country and the future of the economy," he says. "All of this pointed to bigtime infrastructure and bigtime investments in information tech."

By 1993, Barrett's hunches were proving correct. PC makers Compaq and AST were building their first China factories. So were U.S., German, and Japanese automakers. Motorola had just opened a plant in Tianjin to make pagers and was selling more than one million of them a year. AT&T signed a multibillion-dollar deal to rebuild China's telephone system. Meanwhile Chinese bureaucrats were buying PCs like crazy--about 200,000 were sold in China in 1992, 450,000 in 1993, and more than one million in 1994. In addition, U.S. technology export restrictions to China were easing. That meant that although Intel was still legally barred from selling its chips to China, it looked as if it might be able to soon. "I don't think it takes a rocket scientist to figure out what you ought to do next," Barrett says. He was particularly struck, he recalls, by the seemingly insatiable demand for Motorola pagers. If the appetite for them was so immense, he remembers thinking, "cellphones and computers are not far behind."

Intel ramped up its operations in a hurry. It sent Chen to Beijing in 1993 to build a sales force, opened a software lab in Shanghai in 1994, and in 1995 announced plans to build its first factory there to test and package chips. Today Intel isn't just inside computers in China; it is everywhere in China--now the third-largest market in the world for its chips. The company is one of China's largest foreign advertisers, with an annual budget of some $20 million. It's one of the largest venture capital players, having invested more than $200 million in 50 companies. And with more than 3,000 workers, it is one of the largest foreign employers, having committed or already spent nearly $1 billion on two plants--the one in Shanghai and a second, also for testing and packaging chips, in Chengdu. Indeed, today the Chinese don't know Intel as just the company with that "Intel Inside" sticker on their computers, as consumers do in the U.S. They know it as perhaps the most dominant high-tech brand in the country--a brand associated with progress itself.

Which is exactly the point. For all the money Intel saves by testing and packaging chips in China, what gets its executives really excited is China's potential as a marketplace. China generated $3.7 billion in revenue last year, or 12% of the company's total. (Intel's U.S. business, by comparison, is $7.4 billion, and its Taiwanese business $4.4 billion.) Intel won't break out separate profit numbers, but people familiar with its China business say it's one of Intel's most profitable. Demand for chips there is projected to expand at double-digit rates for the foreseeable future. In contrast, the U.S. market for chips has barely grown at all since the Internet bubble burst.

But Intel also understands that opportunity attracts a crowd. Competition from pesky rival Advanced Micro Devices is increasing. Meanwhile a future in which Chinese chip companies go head-to-head with Intel is arriving faster than anyone imagined even five years ago. What does that mean? It means that as hard as Intel has worked to become the dominant IT player in China and to navigate China's suffocating bureaucracy and labyrinthine political hierarchy, there are limits to its commitment. You won't find any Intel fabs--chipmaking plants--in China, for one thing. The nation's virtually nonexistent intellectual- property protections, among other factors, have seen to that.

YOU GAIN a vivid sense of the power and depth of Intel's marketing effort in China by getting on a plane and flying three hours from Beijing to Chengdu. There, on a recent summer day, Li Xiang Peng, Tan Jing Ge, and Feng Yi Ding try to program a Lego tricycle to follow a black line printed on a white mat. The sixth-graders had been coming after school for a month to a room in this rundown and largely vacant office building to work on the project, and now they were showing off the fruits of their labor. A crowd of their peers gathered to watch as Li crouched on the floor and readied the trike for its journey. The experiment started well, but it quickly went bust. The trike followed the line for two feet, but at the first 90-degree left turn, it went right and began spinning in circles.

The kids were frustrated that their creation didn't perform on cue. But they also knew how lucky they were to be working on a project like this. "The computers at school and the one I have at home are five or six years old," Li said through an interpreter, "and we don't get to work in teams like we do here." It's that way everywhere in China. Sales of 23 million PCs last year may sound like a lot, but in a country of 1.3 billion people, it means that more than 95% of homes still do not have one. Schools, if they use computers at all, have old models and typically can't afford up-to-date software. That makes this after-school project, where students can work on everything from programming to animation, a technology paradise. It even has a computerized sound-mixing studio with an electric piano where students can learn how to write, record, and mix their own music.

Chengdu is ground zero for Intel's latest foray in China, and the company is throwing money around all over town. It not only funds Li, Tan, and Feng's after-school program but also helps support the local I-café, one of the biggest gathering spots in the country for teens to play computer videogames. When Chengdu's government starts rolling out wireless broadband connections to homes soon, Intel will help support that too. It is spending millions to blanket the city with advertising. Intel ads are on local television, on billboards, on placards in hotels and at the airport, and on nearly every surface in the city's big computer store. And Intel is about half-finished with the first phase of building a $375 million chip-test plant in Chengdu--its sixth in the world. When the plant is done by the end of next year, Intel will be one of the largest private employers in town.

Chances are that most of the American companies schlepping into China these days have never even heard of Chengdu, which, with a population of six million or so, ranks well behind China's big, increasingly industrialized cities like Tianjin, Shenzhen, and Guangzhou, not to mention Shanghai or Beijing. But Chengdu is where Intel's 20 years in China are paying off, for the prices of land and labor are still low. It has one of the nation's newest airport, road, and telecom networks, thanks to a decade of government investment. And it is home to some of China's top engineering schools, making it a promising source of highly skilled labor--and of PC buyers. Intel knows that firsthand: In 1998 it held one of those PC parties on the streets of Chengdu, and the event turned out to be the biggest ever--more than 400,000 people showed up.

Perhaps most important, though, Chengdu is where China's national government asked Intel to invest. For the past four years a China concerned with spreading the benefits of economic expansion to all its population has been prodding foreign companies to "go west" (see "The New Wild West"). Right now average incomes in western China are half those in big cities like Beijing. And, says Jim Jarrett, who ran Intel's China business for four years until 2000 and is now its head of government relations, the one thing Intel has learned from its lengthy tenure is that when there is an easy way to curry favor with the government, you do it. Even today the government is involved in determining who succeeds and who fails in the business world. "Ergo, we go west," said Barrett when the deal was announced a year ago.

The payoff in Chengdu has been huge. Intel has its own liaison inside city hall to field problems and questions, and Chengdu party secretary Li Chuncheng talks about Intel executives with reverence. "For Intel this shows that the company is strategically farsighted," Li says, "and for Chengdu it is a demonstration that it is a suitable place to develop new and high technologies." That isn't just political blather. Li has completely re-architected the local government to thin its once crushing bureaucracy and more effectively attract foreign investment. Chengdu now has a website with detailed information for potential investors and a separate department, reporting to Li's closest advisors, that solicits and processes investment queries. Indeed, Li says that using the blueprint that Intel helped lay down for dealing with foreign companies, he is working on five other investment deals.

IF CHINA DEVELOPS as Barrett, Otellini, and the rest of top management hope, it will be not a moment too soon. Analysts predict that the U.S. market for PCs will grow by only about 5% a year for the foreseeable future. The company is feeling the impact of the U.S. slowdown acutely. Earlier this month it said that third-quarter revenues would be about 5% below expectations, and that gross margins would be two percentage points lower. Worse, many analysts believed Intel was sitting on too much chipmaking capacity even before the news. For most of the summer inventories have been at all-time highs. Not surprisingly, Intel's share price is down nearly 45% this year.

China, however, could fix Intel's problems in a hurry. Analysts project that PC manufacturing in China for domestic use and export is growing by roughly 20% a year. That's fast enough to make it Intel's No. 1 market for chips in the next five years, and Otellini says he thinks those projections are "very low," though he won't be specific. No wonder that while Intel executives aren't particularly happy to talk about their overall business these days, they will talk your ear off about China. "I come back from visiting China, and I feel as if I've visited the fountain of youth of computing," Otellini says.

China, for Intel, is not just about computer chips either. It is also Intel's most promising market for chips that run smart phones--handheld devices like the Handspring Treo 600--which combine a phone, address book, Internet access, and e-mail in one machine. This is only a four-year-old business for Intel, and it's still losing money on less than $1 billion a year in revenue. But strategically, building circuits for cellphones represents one of Intel's most important initiatives in a generation, and the company is banking on China, the world's largest market for cellphones, as its launching pad.

Barrett says the market is changing in ways that will enable Intel to grab a big piece of it. Right now the chips that go into phones come largely from Freescale Semiconductor, Texas Instruments, and Qualcomm. But Barrett says that as users want cellphones to do more things, the phones will require chips more akin to those in PCs--like Intel's. He also thinks wireless providers like China Telecom and China Unicom will increasingly try to bypass the big handset makers like Motorola, Samsung, and Nokia, hiring no-name contract manufacturers to build their phones. Barrett's bet is that the chips in those phones will be Intel's. "Shouldn't we expect what's happened in the PC market [where most manufacturing is outsourced] to replicate itself in the handset market?" he asks. That could lead to billions of dollars of new business for Intel.

It all sounds grand except for two things. Short term, Intel increasingly has to contend with its much smaller but nettlesome rival, AMD. Intel had the Chinese market to itself until three years ago, when AMD began moving in aggressively, grabbing market share and forcing Intel to cut prices on the chips it sells to PC makers. From almost nothing four years ago, AMD now has 16% of the China microprocessor market, compared with Intel's 84%, according to the Gartner Group. And AMD has lots of momentum. The market share numbers don't show the effect of the August deal the company signed with Lenovo, China's biggest PC manufacturer, to be the exclusive supplier of chips for Lenovo's low-end PCs. These days AMD marketing chief Henri Richard enjoys telling people that the company is saving the Chinese from the Intel monopoly. "Intel [in China] doesn't have customers but hostages," Richard says.

Long term, Intel has to worry about the Chinese developing their own chip industry. Few in the semiconductor business thought that was remotely possible five years ago. But since then the Chinese have harnessed some of the country's best minds and the government's deep pockets to make astounding progress. Researchers at Tsinghua University, China's top engineering school, have already designed a microprocessor that is as fast as the Pentium II, the chip Intel released in 1998. Meanwhile researchers at the Chinese Academy of Sciences, in cooperation with chip-design firm BLX IC Design, said earlier this year that they had designed a chip called the Godson 3 that is as fast as the version of the Pentium 3 introduced by Intel in 2000. AMD is happily lending BLX a hand through a partnership that the two signed last year.

China's accomplishments in chip manufacturing have been just as noteworthy. Companies like Shanghai's Semiconductor Manufacturing International Corp. (SMIC) have gone from nothing to almost $1 billion in revenues in four years. The company has built a $1 billion fab in Shanghai, acquired another in Tianjin, and is about to open a third in Beijing. In Shanghai alone, SMIC employs more than 4,000 people, and it has built housing and schools for 3,000. Most of its top 1,000 managers are from abroad, from places like Taiwan and Dallas, where chipmaker Texas Instruments is based. Sure, SMIC's manufacturing technology is still a generation behind that of Intel, IBM, and the dominant chipmakers of Japan and Taiwan. And as a so-called contract manufacturer, SMIC must rely on outsiders to bring it competitive chip designs, a setup that has produced lots of low- and mid-range semiconductors but has never produced a competitive microprocessor--a fancy kind of semiconductor. Intel, by comparison, both designs and makes its chips and understandably says this melding of design and manufacturing gives it both technological and manufacturing advantages unattainable by contract manufacturers.

But if any country has the intellectual talent, financial resources, and will to one day build microprocessors as good as Intel's, it is China. Otellini, whose job it will be as CEO to fend off the Chinese, sounds unfazed by any of this. Yet he doesn't make light of the Chinese threat either. Ask him about the possibility of an Intel of China, and you get: "Well, it's not one of the things that comes up in our annual strategic review. We tend to look at things in the five-year range in those reviews. But in the ten- to 15-year range, do I think that possibility exists? Yes, I do."

China graduates 250,000 engineers a year, compared with 50,000 in the U.S. It's attracting a slew of U.S.-trained executives, such as SMIC CEO Richard Chang, a Taiwanese national who has set up his company with foreign and Chinese backing. (He used to run Texas Instrument's semiconductor-manufacturing business.) And right now, at least, the government seems prepared to do whatever it takes to make sure the semiconductor business in China succeeds. "The Chinese government attaches great importance to the industry of semiconductors because it is the foundation or basic condition for the development of the IT industry," says Zhao Qizheng, China's director of the State Council Information Office. And as he and other leaders have made clear many times, China is banking on IT to help elevate its economy from the ranks of the world's developing countries into the ranks of the world's great industrialized powers.

China's single-mindedness is one reason Intel doesn't dare build a fab there. (U.S. technology export restrictions on high-end chipmaking equipment are another factor.) Chipmaking plants are where the bulk of the company's intellectual property lies, and as any businessman or intellectual-property lawyer in China will tell you, at the moment there is very little recourse against the theft of ideas. It's not only its chip designs that Intel is worried about but also the design and management of its manufacturing process. Although stealing from Intel isn't nearly as simple a matter as, say, ripping off a piece of software, Intel could still be vulnerable. China's leaders say they are working hard to write laws to protect intellectual property and to train judges to enforce them. But count on Intel to let other companies pave the way on that front.

Instead, says Otellini, Intel is going to do what it so far has done so well in China: work on making itself such a beloved brand --like Coke--that no matter what competitors come up with, consumers will choose PCs and other devices with Intel chips. The bigger Intel's brand and the deeper its connections with manufacturers, retailers, customers, and politicians, the harder it will be to push aside.

But will that persuade China to back away from its aggressive semiconductor push? Probably not. Consider the following: In April, because of a disagreement over wireless-networking standards, China nearly forced Intel to pull its wireless Centrino chip from the market. And until international pressure forced the government to relent in July, China's value-added tax on domestically manufactured chips was substantially less than those on imported chips--3%, compared with 17%. Just as Europe managed to set the world GSM standard for cellphones, few doubt that China will continue to try to dictate standards of its own.

Some critics believe Intel and its Western corporate compatriots will ultimately find themselves on the losing side. Says Handel Jones, founder and CEO of IBS, a Silicon Valley semiconductor-consulting firm: "What we see in China is potentially a huge trap for companies like Intel." Sure, the government is making grand pronouncements about its interest in foreign investment and going to great lengths to persuade U.S. corporations to build plants in China. But what the Chinese really want is to take American technological know-how, he says--just as they have stolen American music, movies, and software. "It's like Japan in the 1970s," Jones says, referring to the early years of Japan's technological development.

Otellini understands that better than anyone but says the more committed Intel is to China, the more likely it is to thrive. "Companies can be rewarded for symbiotic relationships with China" Otellini says. "If you're willing to invest there, they do give you the ability to get rewards." Will there be more head-butting with the Chinese? Probably, says Otellini: "What you are seeing is China going through a process of learning the requirements to play on a global stage in technology. There is always the easy desire to protect your local industries, but then you find very quickly that if you want to participate," you can't do that. "There isan analogy here relative to the Japanese 25 years ago," he adds, but the Japanese figured out how to be a fair-trade partner on the world stage. He's betting the Chinese will too.

FEEDBACK fvogelstein@fortunemail.com