The Rise of Huawei

In the autumn of 1987, in the half-built concrete shell of an apartment near Shenzhen Bay, a forty-three-year-old man with thinning hair and a square jaw signed the registration papers for a small electronics trading firm. Ren Zhengfei had pooled twenty-one thousand yuan, roughly five thousand dollars, with five other investors. He had recently been defrauded out of two million yuan at his previous job, divorced his first wife, and was sleeping in a ten-square-meter rental with relatives. The building where he registered the company, the Shenzhen Development Center, was a rotten-tail tower, construction abandoned mid-rise after a financier ran out of cash. The address sounded better than it lived. Ren named the company Huawei, hua for China, wei for achievement. A heart connected to China, making a difference. There was, in 1987, no plausible reading of the founding moment in which Huawei became the most consequential private company in the People’s Republic. There was scarcely a plausible reading in which it survived its first winter.

The man who signed those papers had been formed by every contortion modern Chinese history could perform on a person. He had been born in October 1944 in Zhenning County, a mountain town in southwestern Guizhou where his parents taught at a rural school. His father, Ren Moxun, had once worked as an accounts clerk in a Kuomintang-affiliated munitions plant, a fact that during the Cultural Revolution would mark the family for political abuse. The Great Leap Forward famine, which by some estimates killed a tenth of Guizhou’s population, taught Ren what hunger does to a household of nine; he later wrote about foraging for wild roots and castor beans. When the Red Guards came for his father in 1966, Ren was studying construction engineering at the Chongqing Institute of Architecture and Engineering. His father was beaten, paraded, and contemplated suicide before deciding that dying with his name still blackened would only deepen his children’s burden. Ren absorbed the lesson early. The defining feature of his subsequent personality, multiple interviewers and former colleagues have noted, was an unshakable, almost theological pessimism, an instinct that catastrophe was always one quarter away.

He joined the People’s Liberation Army’s Infrastructure Engineering Corps in 1974, not as a soldier but as a technical cadre, and was assigned to a chemical fiber plant being built on the Taizi River near Liaoyang in the freezing northeast. The Liaoyang complex was meant to give every Chinese citizen at least one decent set of synthetic clothes. Conditions were miserable, the equipment imported from France and unfamiliar to anyone present, and Ren worked his way up by reading manuals at night and solving automation problems other engineers could not. He joined the Communist Party in 1978 and attended the National Science Conference in Beijing where Deng Xiaoping declared that science and technology constituted the core of the productive forces. When the Engineering Corps was dissolved in 1982, Ren mustered out and followed his wife to the Shenzhen Special Economic Zone. He took a deputy general manager job at a state-run electronics outfit, and within a few years had been swindled, fired, and ostracized. Huawei was not the start of a career. It was the desperate restart of one.

The first product Huawei sold was a private branch exchange, a switch that let a hotel or small office route internal phone calls without paying the telecommunications ministry for every connection. Ren bought the boxes from a Hong Kong supplier called Hung Nien and resold them up-country to whichever Chinese hotel manager or factory boss could be coaxed into a contract. He was not the only person doing this in 1987; Shenzhen was thick with reseller hustles. What separated Ren from his competitors was the speed with which he saw that distribution alone was a doomed business model in a country where state-owned enterprises were already starting to manufacture their own switches and where foreign giants like Alcatel, Lucent, Siemens, and Northern Telecom were preparing to flood in. He pulled engineers off resale duties and put them on reverse-engineering. By 1990 Huawei had around six hundred R&D staff, an absurd ratio for a company that small, and was building its own switches. By 1993 it had launched the C&C08, a digital exchange whose name stood for City and Countryside. The product positioning was the strategy. Foreign firms were chasing the urban backbone in Beijing, Shanghai, and Guangzhou. Huawei would take the rural townships and the third-tier provincial markets where Ericsson and Alcatel could not be bothered to send sales engineers. From the periphery, Mao’s writings had taught Ren, one could surround the cities.

The strategy worked partly because rural China was hungry for telecom and partly because Beijing wanted it to work. By the mid-1990s the State Council was framing domestic telecom equipment as a national-security matter, on the theory that a country whose phone network ran on foreign switches could have its phone network turned off by foreigners. Jiang Zemin, then general secretary, met with Ren personally in 1994 and reportedly told him that a country without its own switches was like one without its own military. Procurement preferences flowed to Huawei from the People’s Liberation Army, the Ministry of Railways, and the public security organs. Cheap loans flowed from China Development Bank, eventually amounting to tens of billions of dollars in financing for Huawei and its overseas customers. None of this guaranteed success; other state-favored Chinese telecom firms of the period, including the once-dominant Great Dragon, withered. Huawei did not, because Ren was a paranoid, relentless operator who wrung work out of his employees in ways that would have been illegal in most of the developed world.

The discipline he built took its cues, openly, from Mao and from the army. In January 1996 Ren staged what became Huawei lore as the mass resignation ceremony, summoning every regional sales manager to headquarters and ordering each to submit a performance summary and a letter of resignation. He would decide which to accept. The exercise was theatrical, the resignations mostly refused, the message unmistakable. No position was permanent; every cadre served at the pleasure of the company. New hires went through three to six months of military-style boot camp before they could touch a customer. Mattresses lived under desks in the R&D labs; old Huawei engineers later described it as the mattress culture. Employees signed striver agreements waiving their right to overtime pay or unused vacation. The whole apparatus was called the wolf culture, after Ren’s metaphor that Huawei had to behave like a wolf pack: quick to scent prey, willing to suffer, willing to attack as a unit. In 1998 he commissioned a team of academic consultants to draft the Huawei Basic Law, an internal manifesto that codified the company’s values, its compensation philosophy, and its requirement that employees, not outside shareholders, would own the firm. Almost all of Huawei’s equity was held in a labyrinthine employee shareholding scheme administered through the company union. Ren himself never held more than about one percent. The structure made Huawei impossible to take public, impossible to acquire, and impossible for outsiders, including the Chinese government, to directly buy. It also tied tens of thousands of engineers’ personal wealth to the company’s results, which is how Huawei extracted twelve-hour days, six days a week, from a workforce that was nominally voluntary.

By the late 1990s Huawei had outgrown China. It opened a research center in Bangalore in 1999, signed a network contract with Hong Kong’s Hutchison the same year, and began the long slog of selling into Africa and the Middle East. In 2005 it secured Approved Supplier status from Vodafone, the first time a Chinese telecom firm had been admitted to a tier-one Western carrier’s procurement list. In 2003 Cisco had sued Huawei in a Texas federal court, alleging that whole sections of Huawei’s router software, including identical bugs, command-line interface text, and source-code comments, had been lifted from Cisco’s IOS. The case settled in 2004 after Huawei removed the contested code and admitted, in carefully drafted language, that some copying had occurred. The episode became a permanent stain in Western tech-policy circles, the first piece of documentary evidence that Huawei’s catch-up had not been entirely clean. It also motivated the company to professionalize, hiring IBM and Hay Group to install Western-style management systems, integrated product development, and HR frameworks. The internal Huawei joke was that the company spent more on consultants than on coffee.

In 2010 Huawei broke into the Fortune Global 500 with annual sales of twenty-one billion dollars. In mid-2012 it surpassed Sweden’s Ericsson to become the largest telecommunications equipment vendor in the world. Most of the world’s public did not notice. Carriers in Frankfurt, Lagos, Mexico City, and Riyadh did. Telecom networking is unglamorous, all racks and cabinets and rural cell sites, and Huawei had become very good at building and maintaining the unglamorous infrastructure on which the next decade of human communication would run. By the late 2010s its 5G base station equipment was deployed in more than half the markets globally that had begun 5G rollout, with Dell’Oro tracking Huawei at close to thirty percent of the worldwide mobile base station market, a share larger than Ericsson’s and Nokia’s combined in many quarters.

The pivot that made Huawei dangerous, in the eyes of Washington, was its move from selling other people’s silicon to designing its own. In 1991 Ren had quietly stood up an internal ASIC design center, initially intended to build the application-specific chips Huawei’s switches required. The ambition was modest. The skills were not. Through the 1990s and early 2000s, the design center grew into a small army of digital and analog engineers, mostly graduates of Tsinghua, Shanghai Jiao Tong, and the Chinese Academy of Sciences, who learned chip design by building modems, optical transport silicon, and basebands for Huawei’s own equipment. In April 2004 Huawei spun the unit out as Shenzhen HiSilicon Semiconductor Company. The Chinese name was haisi, sea of thought. Its first chips were unglamorous infrastructure parts, but it was a fabless design house in the architectural mold of Qualcomm or Broadcom, paying TSMC and other foundries to manufacture what its engineers drew.

HiSilicon’s leap into mobile happened slowly and then all at once. In 2009 it shipped a low-end smartphone applications processor called the K3V1 on a forty-five-nanometer process; the chip was a flop, prone to overheating, with weak GPU compatibility. In 2012 a redesigned K3V2 went into Huawei’s own Ascend D and Mate phones, and was again widely panned. By any external measure, HiSilicon was an ambitious also-ran, designing chips for a captive customer that often would have done better to buy a Qualcomm Snapdragon. The internal mood, by the accounts of engineers who later spoke to Caixin and the South China Morning Post, was different. They knew the first products were mediocre. They also knew Qualcomm and MediaTek had once been mediocre, and they were shipping in volume to a parent company that would buy whatever they made and pay them to keep iterating.

The redemption arrived with the Kirin line. The Kirin 910, in 2014, was the first HiSilicon SoC that could plausibly compete in midrange Android phones. The Kirin 950, in the Mate 8 in November 2015, was the first phone chip in the world to use ARM’s new Cortex-A72 high-performance core, beating Qualcomm and Samsung to that punch by a few months. The Kirin 970, unveiled at IFA in September 2017, included a dedicated neural processing unit co-developed with the Chinese AI startup Cambricon, making it among the first mobile SoCs with on-die AI acceleration. The Kirin 980, presented at IFA 2018 by Huawei consumer chief Richard Yu on the last day of August, was the world’s first commercially announced mobile chip on TSMC’s seven-nanometer process, narrowly preceding Apple’s A12 Bionic on September 12. None of these was a clean victory in benchmarks; Apple’s chips remained ahead in single-core performance, sometimes by a generation. But for a company that a decade earlier had been shipping reverse-engineered phone switches in a rotten-tail building, the gap to the world’s leading-edge mobile silicon had collapsed from a chasm to a stride.

The chip work paid off in handsets in ways that made Western executives nervous. Huawei sold over two hundred million phones in 2018. The consumer division overtook the carrier-equipment division as the largest single revenue line that year. Total revenue crossed seven hundred twenty billion yuan, roughly one hundred five billion dollars, larger than Microsoft the year before. R&D spending hit one hundred and one billion yuan, fourteen percent of the top line, putting Huawei among the top five corporate R&D spenders on earth. In the second quarter of 2020, in a market warped by the pandemic, Huawei briefly displaced Samsung as the world’s largest smartphone vendor by units shipped, with about fifty-five million phones in a single quarter. In the same quarter, HiSilicon broke into the top ten global semiconductor companies by revenue, the first Chinese mainland firm ever to do so. The company that had begun as a backroom design unit was now selling more silicon, in dollar terms, than Nvidia.

Washington had begun watching well before that. The first formal warning shot landed in October 2012, when the House Permanent Select Committee on Intelligence, after an eleven-month investigation led by Mike Rogers, issued a public report concluding that Huawei and ZTE could not be trusted to be free of foreign state influence and recommending that U.S. government systems exclude their equipment. The report was thin on smoking-gun evidence and Huawei executives ridiculed it, but its political effect was lasting. Through the Obama years, USTR repeatedly raised concerns about Huawei’s market-access tactics, and the FBI quietly investigated several Huawei employees in the United States for technology theft. Inside the Trump-era National Security Council, a small group of officials, including Matt Pottinger, the deputy national security adviser and a former Reuters reporter in Beijing, came to view Huawei’s combination of state-supported financing, telecom-equipment dominance, vertically integrated chip design, and aggressive 5G rollout abroad as a strategic challenge of a different order from any Chinese company before. The 5G networks Huawei was wiring into Europe, Africa, Latin America, and parts of Asia would, in the American reading, become the substrate for the next generation of intelligence collection and economic activity. A Chinese national champion that owned that substrate was, in the language Pottinger and others used in private memos, the digital equivalent of a foreign power building the world’s roads.

The threshold moment came on a Saturday morning in Vancouver. On December 1, 2018, Cathay Pacific Flight 838 from Hong Kong touched down at Vancouver International Airport at 11:30 a.m., on a stopover en route to Mexico City. Among the first-class passengers was a fifty-six-year-old woman traveling on a Hong Kong passport as Sabrina Meng, also known as Meng Wanzhou. She was Huawei’s chief financial officer. She had been born Ren Wanzhou in 1972 but had taken her mother Meng Jun’s surname after her parents’ early divorce, a Chinese tradition that allowed a wealthy daughter to walk through airports without being recognized as the founder’s child. Canada Border Services officers detained her at the gate, separated her from her luggage, and questioned her for three hours. They seized her phones and electronic passcodes. The Royal Canadian Mounted Police then arrested her on a provisional U.S. extradition request out of the Eastern District of New York. The charge, formalized in January 2019, was bank fraud and conspiracy: federal prosecutors alleged that in 2013 Meng had given an HSBC banker in a Hong Kong cafe a deceptive PowerPoint about Huawei’s relationship with a subsidiary, Skycom, that had done business in Iran in violation of U.S. sanctions, exposing HSBC to potential sanctions liability when it cleared dollar transactions. She was granted bail at ten million Canadian dollars, surrendered seven passports, and spent nearly three years under house arrest in two of the family’s Vancouver homes, wearing a GPS ankle monitor.

The arrest detonated. Within days China detained two Canadians, Michael Kovrig and Michael Spavor, on opaque espionage charges. Beijing’s foreign ministry condemned Canada in language usually reserved for hostile states. In Shenzhen, Ren Zhengfei did something almost no Chinese tech founder of his stature had ever done: he opened the doors of his headquarters to foreign journalists. Through 2019 he gave more than a hundred interviews to outlets including the Financial Times, the Wall Street Journal, the BBC, CNN, and the South China Morning Post, sometimes for hours at a time, sometimes alone in his office with a single reporter and a translator. He had been almost mute with the foreign press for thirty years. Now, with his daughter under house arrest, he spoke about his father’s beating during the Cultural Revolution, about reading Mao in his army barracks, about being prepared for Huawei to lose thirty billion dollars in revenue, about how the company would survive whatever Washington threw at it. He compared his company, in one Caixin interview, to a battered old plane that had taken so many bullets it would either fly or crash. He was, his interviewers wrote, calm, philosophical, and very tired.

The arrest of his daughter was the public marker of an inflection that had already taken place inside the U.S. government. The volcano under Huawei, in Ren’s word, was about to erupt. But in the late autumn of 2018, with HiSilicon designing some of the most advanced mobile silicon in the world, with Huawei’s 5G base stations going up across six continents, with the company on a trajectory to overtake Apple and Samsung in handsets, the rise was not yet over. It was at its peak. From a rotten-tail building in Shenzhen to a top-ten global semiconductor designer in thirty-three years was a sprint with no analogue in modern industrial history, and it had drawn the attention of a country that did not intend to let it continue.