Chickens scatter as Edward Tian, accompanied by half a dozen eager, youthful, short-sleeved and open-collared engineers, walks along a narrow hutong, or alleyway, in Gao Pei Dian, a village an hour from Beijing in Hebei Province. Wearing a gray suit and polished dress shoes, Tian, the 37-year-old CEO of China Netcom Communications (CNC), seems out of place in such quiet surroundings. Leading the group between rows of crumbling brick houses until they reach a small shop, he ducks inside and walks past a woman weighing mounds of powdery medicines made of antler, mushroom, and bark. She lifts her head just long enough to give the entourage an unimpressed glance and returns to her work.

The group follows Tian out the back door and into another alley leading to the unmarked doorway of a two-story building. They climb a flight of concrete stairs to a door. On the other side is the kind of nondescript room you might expect to see in an abandoned government building (which is exactly what this is). It's musty. Paint is peeling from the walls and 60-watt bulbs dangle from the ceiling. But appearances can be deceiving: This dingy room is ground zero of the next offensive in the Chinese Internet revolution.

Tian's group is here to inspect the room's contents: racks of Cisco, Huawei, and Lucent components, including DWDM repeaters complete with flashing diodes and blinking digital readouts, all connected by snaking yellow cables. It's the newest node in what will be - at 20,000 kilometers and 40-plus gigabytes per second - one of the longest and fastest high-bandwidth networks in the world. An ambitious project with numerous social as well as commercial implications, CNC's network promises to leapfrog China into the forefront of information technology. "Cities that never had phone service are being wired for broadband," Tian says. "We will basically wire the nation with fiber and bring limitless opportunity to the people."

The first phase of the broadband network has been built and CNC has signed up a long list of corporate customers that includes Unisys, Mastercard, China Mobile, ExxonMobil, and GE Medical. An army of 20,000 workers has dug 8,600 kilometers of trenches - a distance 1,400 kilometers longer than the Great Wall - and laid cables connecting all of China's 17 largest cities. There are 550 optical fiber distribution centers like the one in Gao Pei Dian, placed every 60 to 80 kilometers along the network to regenerate the light waves. (Currently located in old government buildings, these centers will soon be moved into ultraclean, temperature-controlled structures with redundant power supplies.)

Besides wholesale bandwidth, CNC's offerings include webhosting, data centers, voice over IP, and other Internet and intranet services. In less than two months, the company has rolled out the depth and breadth of services that it has taken WorldCom and Sprint years to build. With even more on the way, "CNC could become the largest and one of the most influential companies in the world," says Jay Chang, an analyst covering Asia for Credit Suisse First Boston.

But Tian won't stop there. He envisions his network giving birth to an economic boom in China unlike anything the country has ever seen - tens of thousands of startups, and services to modernize existing companies. He also foresees a social revolution that would result in radically improved education and a whole new health care system, made possible by unprecedented access to information.

With his neatly cut black hair, and serene, deep-brown eyes, Tian stands with arms folded, examining a piece of transparent fiber about as thin as a fishing line. "Is this G655 fiber?" he asks an engineer. Assured that it is, Tian turns toward me. "One pair of these fibers can carry 40 gigabytes of data," he says. How much is that? "Enough to carry all of the conversations on all of the AT&T lines at any single time in the United States," he answers - faster than the speed of WorldCom's UUNet network, state of the art in the US.

All told, there are 96 pairs of these wires in place and the potential for several hundred more. (New fiber can be blown through the cables to meet demand.) That equals ... Tian fills in the thought: "Limitless bandwidth. Imagine what limitless bandwidth can do for China." The thought of it is enough to inspire an impromptu motivational speech. "Think about bandwidth!" Tian says. "Enlightenment can flow through the taps like water."

The engineers seem moved. Afterward, one tells me, "Working with Mr. Tian, every day we feel as if we are writing China's history."

Of course, getting to the point where enlightenment flows like tap water is nowhere near as easy as Tian lets on. Internet use in China is doubling every six months and there are 20 million Chinese with Net access, but this is a tiny fraction of the nation's population of 1.3 billion. Implementing a world-class Internet infrastructure in China also means growing the market in a volatile economy, and dealing with a government that has a love/hate relationship with the Net - not to mention with its citizens.

But the obstacles don't faze Tian. As we're leaving the province of Hebei, we pass a gathering of bare-chested men sitting on boxes and smoking cigarettes on the side of the road. He points at them. Most observers consider the potential commercial Internet market in China to be the 250 or 300 million people in the large cities who earn at least three times as much as most rural Chinese, but Tian says the network will help the people on the side of the road, too. In other words, the other billion. They may not be using email or ecommerce anytime soon, but ubiquitous bandwidth will affect their lives. Tian says CNC will wire clinics, schools, and libraries, and a new China will emerge. "Since the broadband Internet can be picture- and voice-run, illiterate people can have access," he says. "Every school will one day be on the network."

In poor regions like Hebei, children often die because of inadequate medical training and resources. Tian envisions broadband Net access as a tool for teaching local doctors and connecting village clinics to medical centers in big cities. For more advanced consultations, specialists would be able to see patients and read their vital signs in real time. "Today," he says, "a child is born with something as simple as a cleft palate and nothing is done - he becomes a ghost." In the future, Tian hopes his network will help makes sure that no longer happens.

Edward Tian is known throughout China as the man who built the Internet. As a cofounder of AsiaInfo, the country's first homegrown Internet infrastructure company, Tian helped build the backbone of many of the nation's commercial networks. After its IPO on Nasdaq last year, AsiaInfo was valued at more than $1 billion - the first nongovernment Chinese company to reach that milestone.

Not that Tian was around to celebrate the momentous event. Eleven months earlier, he baffled many of his friends and angered AsiaInfo's board when he announced that he was leaving for a government-funded startup designed to compete with China Telecom, a 500,000-employee state-owned monopoly. It was as if Steve Jobs left Apple for a government-owned alternative to the post office. Only this was even more shocking: China's attitude toward the Net has been schizophrenic, at best.

__Tian's network - far longer than the Great Wall and growing - is one of the world's fastest. Today's target: 300 million urban Chinese. Tomorrow, the other billion. __

Ever since the public Internet was introduced to China in 1995, the nation's leaders have been in a quandary. The Net was essential to modernization and globalization, but it was also a threat to control. As a result, Beijing has by turn encouraged and discouraged investments in Internet technology, and it has installed and then dismantled regulations designed to monitor users and block access to certain content and communication.

But after a 1998 report written by Hou Ziqiang, a renowned physicist with the Chinese Academy of Science, Beijing seems to have realized that it can no longer halfheartedly support the Net. Widely respected for his work as a research fellow in the Institute of Acoustics at CAS, Hou studied the potential impact of Western IT companies - which would gain access to the Chinese market after China's admittance to the WTO - on homegrown technology. Would Chinese technology stand a chance?

Hou argued that the country could be competitive only if it launched a new initiative to construct a next-generation Internet infrastructure on a par with the most sophisticated networks in the world. He also concluded that China Telecom, a slow-moving behemoth rooted in old-fashioned wire telecommunications, could never pull it off. Nor could the government's smaller telcos, China Unicom and Jitong, which were focused on wireless and Internet applications and rife with managers from the ranks of China Telecom. So Hou offered a bold prescription. The government should create a company modeled on the best Silicon Valley success stories - an independent, entrepreneurial entity charged with building a new network. Essentially, he suggested that China do what the US government did when it broke up Ma Bell to promote competition - but instead of relying on the private sector to provide the competitors, the government would create the competition itself.

Beijing is often written off in the West as a single, intractable force. But Hou's boss, CAS vice chair Yan Yixun, found four influential supporters (who would become the company's board members): Sheng Guangzu, deputy governor of the Ministry of China Railway Communication; Zhang Haitao, vice governor of the State Administration of Radio, Film, and Television (SARFT); Yang Xiong, general manager of Shanghai Alliance Investment; and Jiang Mianheng, who represents the CAS. Jiang is the US-educated son of Chinese president Jiang Zemin and the supervisor of the Shanghai municipal government's Information Technology Office. Jiang, who has a PhD in physics from Drexel University and founded a government-backed VC firm, is widely respected for his political and financial support of China's most advanced technologies. The group championed Hou's controversial idea and arranged a meeting in February 1999 with Chinese premier Zhu Rongji. Hou acknowledged that the project would be controversial, but he ultimately convinced Zhu to support it.

With Premier Zhu, CASvice chair Yan, Jiang, and the other powerful government representatives behind it, the plan was approved over the objections of China Telecom's chiefs. But who could run the startup? Following Hou's advice, the directors looked to the private sector. Another supporter, vice minister of railways Peng Peng, says it was a bold departure for the government, but it was the only way to go. "The traditional route - hiring a manager from inside another government organization - would have failed," he says. "China's inevitable entry into the WTO required nothing less than new thinking."

Both Yan and Jiang had known Tian from their dealings with AsiaInfo, and they were sufficiently impressed to put him at the top of their list. Credit Suisse First Boston's Chang says Tian was an inspired choice. "They looked to Tian because he has passion - for technology and for China - and charisma. He has an enormous ability to keep people focused on the task at hand. He also has a willingness to learn. But his background was not in telecommunications, so his learning curve would initially be steep."

When Tian was offered the position in March 1999, he said no. "Everything about it seemed ludicrous," he remembers. AsiaInfo was preparing for its IPO. And the new company's ownership structure, which was hammered out in the high-level meetings in Beijing, seemed untenable. Four shareholders - the Ministry of Railways, the City of Shanghai, SARFT, and CAS - would each have a quarter interest in exchange for a $75 million investment. In addition, as a telco, the startup would be regulated by the Ministry of Information Industries, which not only oversees but also runs the company's biggest competitor, China Telecom. Translation: five demanding bosses whose interests could contradict and conflict with one another.

Still, Tian agreed to meet with the directors and quickly realized that each shareholder had something to offer. The Ministry of Railways could provide the right-of-way along its rail system and handle the physical work of building a network. SARFT had access to China's 80 million cable TV subscribers, an impressive initial user base by anyone's standards. The City of Shanghai's Jiang was clearly influential, as well as a valuable technical adviser. CAS could provide many of China's best scientists; it also had a powerful leader in Yan and a remarkable track record with Legend Computer Systems, a 1984 spin-off that has slowly wrested the leadership of the Chinese PC market from established foreign brands like Compaq and IBM. Legend now owns a 27.3 percent share of the Chinese computer market.

Beyond the practical considerations, though, the potential of the project began to keep Tian up at night. This new company could transform China. Chang puts it this way: "If somebody came along and said to you, 'I want you to create what could become China's largest company; you'll have all the relevant government support, and you can in the process contribute immeasurably to the people of China,' what would you say?"

Tian's desire to transform China can be traced to his boyhood. Born in 1963, he was a child of the Cultural Revolution, a tumultuous and brutal era launched by Mao Zedong in 1966 as a consolidation of his power. The main targets were intellectuals, including teachers, party officials, and other privileged Chinese. Before the revolution, says Tian's mother, Liu Shu, her son had a "normal life, playing under his grandmother's peach tree - free and joyful." After 1966, however, his childhood was characterized by a single, all-pervading force: hunger. "Hunger for food. Hunger for information," Tian says. "It's all I remember."

In 1968, Liu Shu and Tian's father, Tian Yuzhao, both Russian-educated researchers and teachers, were sentenced to hard labor in reeducation camps and factories. Edward stayed with his grandparents in Shenyang, but in 1969 the family's home was taken over by the Red Guard, Mao's citizen militia, which housed dozens of people there. Edward and his grandparents were confined to a small back room. At one point, the family was forced to surrender its collection of world literature to a Red Guard bonfire held in the yard. Panicked and tearful, Edward tried to rip a book out of the hands of a Red Guard officer, who seized the book and knocked Edward down.

__"If somebody came along and said to you, 'I want you to create China's largest company - you'll have all relevant government support,' what would you say?" __

The Cultural Revolution ended with Mao's death in 1976, and the second stage of Edward's life began. Deng Xiaoping "opened our eyes to the outside world," says Tian, referring to those years as the "wake up" period. "It was like seeing for the first time after living your life in darkness. The Chinese people, for the first time in my generation, dreamed."

Deng not only ceased the purges of intellectuals but created a new political and social climate. Finally, there was news from beyond the country's borders, and foreigners began to visit. Revitalized, Tian delved into his studies, and in 1980 he scored high enough on the nationwide college entrance examinations to gain admittance to the top local university, Liaoning. He majored in biology, and after graduation he entered the prestigious graduate school of the Chinese Academy of Science. While there, he married Jean Kong, whom he had met in a biology class at Liaoning. Soon after, a professor visiting from Texas Tech encouraged them to apply to school in the US; he left China in 1988 for Lubbock and followed Jean the next year. She worked on her master's degree in biology and Edward on his PhD.

The following May, the two found themselves in front of their TV each night, watching the news in their apartment. A growing pro-democracy demonstration in Tiananmen Square, 1 million students strong, had caught the world's attention. On June 4, government tanks and troops fired on the crowd, massacring as many as a thousand people (the exact number still isn't known). "Nothing would ever feel quite the same after that," Tian says. For months afterward, Edward and Jean tried to figure out what they should do. "We thought and thought: How could we sacrifice to help the people of our country? It became the central purpose in our life," he says. He wondered whether to abandon his education. "Pure science? I felt that a Chinese person did not have that luxury. I had to do something that would have an impact on China - but what?"

They were biding their time and continuing their studies while they tried to make a decision; Jean worked at a Mongolian barbecue, Edward translated books (including Lee Iacocca's autobiography) for his father-in-law's publishing company in China. He also did a lot of reading, gravitating toward material about Silicon Valley companies, and became inspired by the tale of Steve Jobs. Tian remembers, "He was the one who made me see a new way to change the world - entrepreneurism."

After he emerged from 14 years of solitary confinement in a PRC jail in 1993, dissident Wei Jingshen said, "People long for change, but they despair of it, so they go into business." Tian's turn to business had more to do with hope than despair. China, with a fifth of the world's population, was stuck in the past, while the rest of the world raced toward a new, undefined future. He saw that change depended on opportunity, information, communication, a progressive educational system, and a strong economy. "I became devoted to the notion that entrepreneurship could be an engine of a society's progress," he says.

Still in the US, Tian audited management classes at Texas Tech and contemplated the type of company that would have the largest impact on Chinese society. Then he discovered the answer. In mid-1992, a friend who was a campaign worker for the first Clinton-Gore campaign dragged him to an event in Lubbock, where Al Gore made a speech that stirred him. Tian had used Bitnet at Texas Tech, but Gore was describing the Internet in a way he had never imagined: as an "information superhighway" that could transform a nation through communication, knowledge, and education - exactly what China needed.

Tian finally knew what to do. He would return to China and work to connect the country to the information superhighway. This decision would lead to what Tian considers the third stage of modern China. "We were starved for information when we grew up," he says. "But because of this technology, China will soon be part of the open world and the children will become engines of change. We are at the new dawn - the stage of the information technology entrepreneur."

Returning to China, though, would mean a huge personal sacrifice for Tian. Jean, whose childhood trauma was similar to Edward's, was adamant about raising their children - they had a daughter, Stephanie - in the West. He would have to leave them behind. "I missed our daughter's first steps, her first words," he says. "Sometimes she gets angry at me for being away and she won't come to the telephone." Now, regular flights between the San Francisco Bay Area, where Jean and Stephanie live, and Beijing leave Tian with a constant case of jet lag, but he doesn't regret his decision. "There was no staying away. I returned to China to transform our country, and to give a voice to the Chinese people. China has struggled for two and a half centuries. My generation has that burden on our shoulders."

In 1993, the Net in China was a small system used by several thousand academics. While he was still in the US, Tian and another Chinese student, James Ding, started an email discussion group and bulletin board for Chinese expatriates, computer scientists, and environmentalists. He and Ding, an engineer who worked on Internet technology at UCLA, then launched an early Internet Web center with a search engine designed for academic research. By year's end, they founded the first incarnation of AsiaInfo with a news service called AsiaInfo Daily News, which translated Chinese news stories for an American audience.

Ding and Tian formed a consulting firm called Business Development International and opened an office in Beijing in 1994; they developed software for companies' information systems. Their first customer was the Shenzhen Stock Exchange. Next came Sprint International, an early Western presence in China that had been commissioned by China Telecom to build the nation's first commercial Internet. Sprint subcontracted with BDI, and the fledgling company completed the first phase of work on the backbone for a commercial Internet that was to become ChinaNet, the nation's main ISP. In 1995, in preparation for the second phase - the wiring of 30 Chinese cities - Sprint decided it didn't need a subcontractor and cut BDI out of the deal. Tian, Ding, and a third partner decided to create a new company, AsiaInfo, and compete for the contract, which they won. AsiaInfo went on to complete ChinaNet and to become the primary player in China's fast-growing systems integration business. Then Tian got the offer from Beijing that he wasn't sure he could refuse.

Though the government's vision of an IP- and fiber-based Chinese telco was exciting, Tian was cautious about the strings that might be attached. The negotiations were contentious, with the investors resisting some of his conditions. Many of Tian's demands challenged assumptions about Chinese bureaucracy: He asked for a hands-off policy - that is, he would be the commander and CEO, and his investors wouldn't second-guess his management decisions. Tian also wanted Western-style incentives for his employees - meaning shares of the company, a system rare in the Chinese private sector and unheard of in the public sector. As a sign of goodwill, Tian promised to give his own share to charity.

__A high-bandwidth network, Tian believes, will be the first step toward social revolution. "Enlightenment," he says, "can flow through the taps like water." __

Ultimately, the investors relented. Tian was given 1.5 percent of the company for employee stock options. (Since the CNC deal, Legend Computer has also offered its employees a stock-option plan.) Investors also agreed to give him autonomy - no second-guessing. Tian would have their support as well as the ability to introduce an IPO in the West (currently planned for 2001). That the government agreed to Tian's demands is remarkable, says Duncan Clark, a Beijing-based analyst with Internet consultant firm BDA China. "In China, there's not a long tradition of separation of shareholders from management," he says. "Too often a shareholder will want to show its power to other shareholders or to management, even hobbling or destroying a company in the process."

In spite of the government's assurances, Tian still anguished over the decision. Most of his friends thought he was crazy even to consider the offer. Ding, his cofounding partner at AsiaInfo, tried to dissuade him. "Working for the government, with its unwieldy bureaucracies, was diametrically opposed to everything we learned as entrepreneurs," Ding says.

"You cannot" Tian recalls AsiaInfo board member Daniel Auterbach saying. "Edward, you are ruining your reputation! The IPO is coming! It is professional suicide."

Tian's parents, on the other hand, told him he should serve his country. A family friend, Feng Zhi Jun - a senator in the People's Congress, the vice chair of China's Democratic League (one of the country's sanctioned minority parties), and the father of Bo Feng, an early AsiaInfo investor (see "He's Got Guanxi!" Wired 7.02, page 122) - was also in favor. "This is an historic calling," said the elder Feng.

Finally, in May 1999, Tian nervously decided to accept. Says CAS vice chair Yan, "Edward saw this as a chance to build technology that will help China catch up to and even surpass the West. He saw how it could change the entire country. And he saw how the WTO compelled us to build a competitive environment early so Chinese companies would be in a position to compete globally."

After winning Tian over, though, some government representatives had second thoughts. One questioned whether he was too young and whether it was safe to have someone who had been educated outside China in such a powerful position. His patriotism questioned, Tian was furious: It was patriotism that drew him to the job in the first place. Feng Zhi Jun assured Tian that these obstacles only demonstrated the importance of the calling. In addition to running the new company, he'd have to break down walls and change age-old paradigms that were hurting China's development.

Tian was finally appointed to the job, and he set up shop in a single room in Beijing's Friendship Hotel. "It was pathetic," he admits. "The fax didn't work. The printer didn't link to the computer. There was hardly any staff. The government's initial funding was slow to come in and I had to borrow 2 million yuan ($240,000) from James."

Ding, who had become the new CEO of AsiaInfo, says, "When I saw the office, I said, 'Edward! What are you doing? You left AsiaInfo for this?!'"

Tian made good on the loan from Ding in July, when $75 million came in from the funding shareholders. He also secured a $600 million line of low-interest credit from the Chinese Construction Bank; in exchange, CNC would build the bank a state-of-the-art network.

Next, Tian set about hiring a team. Recruiting was a challenge: Going to work for the Chinese government was considered an unenviable move at best. But Tian persisted and eventually landed many of China's best engineers, as well as Chinese executives from Motorola, Microsoft, and Intel in China and Europe. As a legendary figure in China, Tian himself was a selling point. Elaine Wu, a manager in charge of investor relations, says, "He is one of the most important heroes."

Tian promised that CNC's culture would be based on open communication among everyone, regardless of their level. He wanted his employees to work together toward common goals, but encouraged them to find their own paths to accomplish these goals. He initiated weekly rallies, at which he always used the word "we," never "I." He also developed an extensive training program, dubbed CNC Academy. Tian says, "Our managers all learn what we value here - to rule by trust, not fear."

It was a departure from Chinese business as usual. An engineer who came aboard from China Telecom told me, "There we had no rousing speeches, no feedback from our managers, no reviews or rewards, and no sense of a mission. We had jobs, plain old jobs, like everyone in China. Here we have an undertaking that can change China."

One early hire was Chareleson Zheng, who was convinced to leave his position as GM of Marconi, a British telco supplier, to become Tian's COO. "My friends thought I was crazy," he says. "But I'm not working for the government. I'm working with Edward. Edward sold me on his bandwidth dream." Says Liu Chuan Zhi, the chair and CEO of Legend Computer, "He assembled a handful of software and hardware geniuses - the most impressive minds in their fields, after many people said he would fail."

These days, CNC looks like a lot of US startups, complete with engineers working late hours, arsenals of Nerf guns, and an espresso bar. One of those present at one of the company's many midnight work sessions says that he didn't get a salary for his first six months because of the bureaucratic approval process, which Tian had yet to smooth out. (It has since been taken care of.) "But you don't worry about the money when you are on a mission that is as critical as this one," he says. Within two months, Tian had hired 250 employees. Now there are about a thousand.

__A Chinese expression, qihu manxia, translates, "Once you get on the back of a tiger it's hard to get off." Beijing's next-gen Net may have set an irreversible course. __

The team immediately began developing key strategies. The raison d'être: Build the network. Along with his top engineers, Tian set about designing, testing, and selecting technology from Lucent, Ericsson, Juniper, and others. He also negotiated for the laborers to dig the trenches and lay the fiber. A veritable army of construction workers, subcontracted through the Ministry of Railways, broke ground in February, digging and assembling the network - much of it along existing railway lines.

While this ambitious project was under way, Tian set several teams to work on near-term revenue-generating ventures, including a telephone card that allows calls over an IP network - much cheaper than calling via China Telecom or other providers' wires. It generated $48 million in its first year. Other IP products included an Internet credit card, a caller-ID system for businesses, and a technology called IP-800 that connects business Web sites with an automatic telephone callback system. Visitors to the China Hotel Web site, for example, can immediately get calls back from the company's reservation office to their computer's telephony system. Other teams worked to put companies online with myriad high-bandwidth enterprise solutions while others built and then began to lease servers at CNC's data centers. Ten of fifty centers are built. Now under construction: the massive, 500,000-square-foot Beijing Super Internet Center, complete with its own power plant. At the BSIC, some 100,000 servers will be housed on land that a year ago was a cornfield. Among the first to sign on was the Bank of China.

But the main effort remained the fiber buildout. "Fiber is like a road upon which everything else will go forward," Tian says. Fifty days after the groundbreaking, the first 8,500 kilometers were in and connected. It went live in August. In only 10 months, CNC built a network comparable to the ones built in the USby Qwest and Sprint in two to three years. The results astounded skeptics. Says Peng, "People who were opposed to this project now see what one man with a vision can accomplish. It is waking up people throughout the Chinese government."

In the big cities, the network was connected to high-bandwidth rings that link major buildings. CNC is in the process of wiring the enormous Oriental Plaza, a new office complex on Beijing's main thoroughfare, Xichanganji. At 8.5 million square feet, it's the largest business space in Asia. It's also a model for China's future development: Every office has a "smart desk" with a 100-Mbyte connection. On the first floor of the complex, CNC is working with Sony to create a showroom that will display the most futuristic Sony products linked to a high-bandwidth network.

Network offerings already being sold include a wide range of high-bandwidth IP-based services with voice, 3-D imaging, fax, videoconferencing, and data communications. Meanwhile, the network is being expanded to villages and ultimately to homes throughout China. With the help of SARFT, next year CNC will begin offering high-bandwidth access through service providers to the 80 million homes already wired with cable TV connections. CNC is also selling wholesale bandwidth to China Mobile and to ISPs. In the distant future, Tian plans to build a new broadband transcontinental submarine cable that will connect CNC's system with similar networks in North America and Japan. Meanwhile, Tian has set up a division to wire schools, libraries, and clinics in cities and remote villages.

So far, Tian's relationship with his government investors has gone smoothly. But he has to keep a number of officials happy and make sure the government shareholders remain apprised of his activities. "I view it as part of my job," he says. One of his first hires was Elaine Wu, liaison to the government shareholders, and he consults with them at frequent dinners and meetings. "We have open and frank communication," he says.

CNCis also making friends in the Chinese business community, thanks to the way it's cutting costs. "Fiber built here costs a twentieth of what it costs in the US," Tian says. CNC's is the cheapest bandwidth in the world. This has far-ranging implications, since bandwidth costs can easily eat up 60 percent of fixed expenses for a Chinese ecommerce firm. "With the network in place comes the environment for Internet companies to easily get started," he says. "As we get more of our network up and running and we offer more services, we can offer inexpensive access. As a result, the Internet economy booms. It's a snowball. Everything gets cheaper."

The number of individual Net users in China is small because of the high cost of going online. On average, Internet users in China spend 34.8 percent of their salaries on fees that average 1,000 yuan ($120) a month. CNC's monthly fee is 100 yuan for unlimited access. In addition, Internet telephony will be 60 percent cheaper than current phone rates. Tian says, "Everything comes from high bandwidth at low cost. No one else will get the right-of-way to dig up the land to lay down the cable. We got that free from our shareholders. Who can compete with that?"

And yet, things may not be so easy as Tian implies. Though it's difficult to imagine successful foreign competitors hurting CNC's main IT business in the foreseeable future, the company could face a challenge in China Telecom's nearly inexhaustible resources. If nothing else, China Telecom could make Tian's job difficult, thanks to the controls the telco has on rights of way in some provinces and wherever CNC's network needs to interconnect with existing China Telecom networks. On the other hand, as BDA analyst Clark says, "China Telecom doesn't have much goodwill with customers, to put it mildly. And I don't think China will allow the interference; I don't think it would risk hobbling an entire industry to protect a vested interest."

China Unicom is another threat; it has a great deal of capital after last year's IPO. Though Unicom is focused primarily on the wireless business, it is entering some areas that CNC has targeted. Another telco, Jitong, is also a competitor, but only in small segments of CNC's long-term business.

There are smaller competitors in China, including data center companies iAdvantage, PCCW, and CITIC, and a raft of niche players. And foreign competition will begin to flock in post-WTO: Look for Global Crossing, WorldCom, Exodus, AT&T, Qwest, NTT, and Vodafone, to name a few. However, says Credit Suisse First Boston's Chang, "by the time these players can enter the market with any significant strength, CNC will already have most of the important customers locked up, will have built a service organization into one that would be difficult to displace, and will have built the lowest-cost network in China with upgradable conduits - something that none of the international players have access to. It's why I'm so bullish on CNC."

CNC's biggest worry, though, is daunting: China itself. The fact that the government has waffled so dramatically on its Internet policies over the years makes it a dangerous, chaotic force. Clearly, those with an interest in maintaining the status quo do not share Tian's enthusiasm about a social revolution. As recently as November, new regulations were announced that restrict foreign news on portals such as Sina.com and Yahoo! China. Two sides of the Net-as-revolution debate are clearly emerging. "This is a horse race between encouraging information technologies and resisting them by people who want to control information," says Orville Schell, the dean of the Graduate School of Journalism at UC Berkeley and a longtime observer of China. "No one knows which horse will win."

James Mulvenon, who has been researching the Internet in China for the Rand Corporation, agrees. "Sure, the optimistic scenario says that the Internet's alternate channels of information begin to pluralize China. But the state can also use this technology for surveillance and to increase - not decrease - its control of its citizens," he says. "The young, urban wired in China think the world has changed, but just when they get too giddy, they could get squashed. Fundamentally, China is being altered by the information revolution. But we have to be careful to manage expectations."

Such talk doesn't worry Tian. As far as he's concerned, there's no sign that Beijing will withdraw support. "We're working together with the same goal," he says. "We are building our nation so that the people of China will have opportunities they never had before and our nation will thrive."

Indeed, by building a state-of-the-art Internet, Beijing may have set an irreversible course. "There's a Chinese expression, qihu manxia," says Schell. "It means, 'Once you get on the back of a tiger it's hard to get off.'"

Tian has moved CNC twice since he set up shop in that single hotel room. Now it's housed in a gleaming silver building, and his spartan corner office is decorated with photos of Jean, Stephanie, and the new baby, Frances, in small silver frames. There's a workstation on a rosewood desk from which he can see the Beijing financial district.

Early one morning, after the trip to Hebei, Tian is hard at work in his office. First he meets with representatives of some of the shareholder ministries, and then he's on the phone, checking in with the managers who oversee the building of CNC control centers, network hubs, and the laying of fiber in China's far-off outposts. From there, he goes to his email, planning a road show to the US that he hopes will attract a significant round of private funding. (It would turn out even better than he hoped: In October, the company announced $300 million in funding from a group of investors, including Michael Dell, Rupert Murdoch, and Goldman Sachs. The deal placed CNC's value at $2 billion in a single year - twice the milestone Tian reached at AsiaInfo.)

But all that was still to come. Right now, the most important thought in Tian's head is that it's late morning - bedtime for his children in California. He places a call to Jean, who tells him that the baby is asleep and then passes the phone to his 5-year-old daughter, Stephanie, who has much to say about her birthday party at the Discovery Museum in Sausalito.

After Tian hangs up, an assistant reminds him of a meeting in a corner conference room and he rushes in to greet representatives of Sina.com, China's largest Internet portal; among them is the company's founder and CEO, Wang Zhidong. Over the course of a month, Tian and Wang have ironed out an agreement that lead to CNC's hosting of Sina's exclusive coverage of the summer Olympic Games. In addition, Sina.com became the major online distributor of CNC's IP phone card and telephony service. After PowerPoint presentations about the partnership, the two Chinese IT stars shake hands and Tian returns to his office.

Several meetings later, Tian and I leave the CNC office and jump into the backseat of a car. We're heading to a restaurant behind a wall of bamboo, across the moat from the Forbidden Palace in old Beijing. The car flies past Tiananmen Square, where tourists follow guides, children play tag, vendors sell fruit ices, and newlyweds pose for photographs as an orange dusk settles on Beijing.

Driving past Tiananmen, or Heaven's Gate, with its enormous, three-story-high portrait of Chairman Mao, Edward has a rare wistful moment, reflecting on the call to Jean. He admits that the distance from his family remains the toughest part of his life - tougher than building CNC at this inhuman pace, tougher than negotiating the buildout of the network throughout the developed and undeveloped corners of China, tougher than managing his government partners - tougher even than starting over from scratch. In this solemn mood, he reveals a secret, personal motivation for his ultimate decision to accept the CNC job in spite of all the reasons not to. "Using videoconferencing with high-definition, large-screen monitors in Beijing and Marin, I could be in Stephanie's room in the evening when she goes to sleep," he says. "When the network is in place, I can be where I am compelled to be - here, in China - but simultaneously I can be in Stephanie's bedroom. Then, each evening, I can read to her until she falls asleep."