Alpha Chips
In January 2019, Robert Swan was allowed to drop “interim” from his title—he became the permanent CEO at Intel. In April, only a few months after that announcement, he flew to China, where he first used his official Chinese name, Si Ruibo. He met with the vice minister of industry and information technology, a tradition for Intel executives (whenever they visit China, they schedule time with government officials, even if the meetings seem symbolic). And he convened Intel’s Chinese employees, who were eager to hear their new boss’s plans for their home country. Ian Yang, Intel China president, recalled: “Our CEO told us to dive deep in the demand of Chinese market.”1 Yang interpreted this as a charge to create a customized strategy for China and not just copy Intel’s global strategy.
If you are wondering why one country, even one as large as China, deserves special attention from the new CEO of one of the world’s leading semiconductor makers, consider this: Despite trade tensions between China and the United States, China still delivered $20.16 billion in annual revenue for Intel in 2019, more than one-quarter of the company’s $72 billion total.2 That poses an obvious question: How did one of the oldest, most respected firms in Silicon Valley thrive in a country that seems to favor agile upstarts?
Intel opened its first China office in Beijing in 1985, amid an early wave of foreign investment. Like many of its peers, the chip giant didn’t make much progress there at first. China’s economy had only recently opened to the rest of the world, and Western investments were typically small. Mostly, Intel waited and watched from its office at the National Palace Hotel.
In 1992, Deng Xiaoping, the Chinese leader who had initiated his country’s economic reforms, took a tour of the southern cities that were thriving thanks to their burgeoning economies—Shenzhen, Zhuhai, and Shanghai. Throughout the trip, Deng boasted of the region’s economic accomplishments while deriding his critics. The publicity around the trip and Deng’s pointed remarks reinvigorated economic reform in China, which had been imperiled by the 1989 political turmoil. Foreign companies, especially IT firms, entered a new phase of growth in China. US PC makers AST and Compaq built factories. IBM added Shanghai and Guangzhou branches in addition to its Beijing headquarters. Hewlett-Packard started transforming its Chinese joint venture, introducing a more Western-style operational model. And Motorola invested $120 million in a plant in Tianjian, as its pagers became a hit among China’s emerging consumer class.3 That year, China attracted more foreign direct investment than any other developing country.4
Andy Grove, then Intel’s CEO, visited China the following year. After his trip, he and the rest of Intel’s management team concluded that Intel would, in time, thrive in China. “Everybody we met at government was single-minded and determined and generally very well informed on technology,” Sean Maloney, then Grove’s technical assistant, recalled about the trip. “Many governments elsewhere were ambivalent or uninformed. Here was a series of local and central officials who had a single purpose, which was to develop the country around a strong technology base.”5 And China was eager to welcome Western technology giants like Intel. It had begun to try to build its own semiconductor industry in the early 1990s but had not yet made real progress.
Intel redoubled its efforts in China. It appointed Wee Theng Tan, its director of business operations for Asia Pacific, who was ethnically Chinese and had grown up in Singapore, as its country manager and sent Jason Chen, a Taiwanese sales manager, to build out its distribution network. It also dispatched Jim Jarrett, an executive at Intel’s headquarters, in Santa Clara, California, to be the company’s first China president. Tan recalled how Andy Grove and Paul Otellini, then the CEO and senior vice president, respectively, had persuaded him to take his China post: “China is going to be the next big thing for Intel. If you want to do something about it, come over to China.”6 That was 1996, and Intel China’s revenue was only about $75 million, less than 1% of the company’s total. Tan wasn’t sure, but his bosses were adamant.
In the middle of the 1990s, personal computers were luxury items in China, often priced around 15,000 RMB ($1,800)—more than a year’s salary for someone in a top-tier city like Beijing and Shanghai. Foreign brands, including IBM, Compaq, and Hewlett-Packard, dominated the market, and their computers were Intel’s main source of revenue. Lenovo, now the world’s largest PC maker, and other Chinese manufacturers accounted for only 10% of PC sales. Though Intel’s top management believed PCs would come to be used by everyone in China, the high price made that vision initially unrealistic. Intel’s executives in China realized that, to make their bosses’ vision a reality, they needed to make PCs affordable for more of the Chinese people. “If you tell consumers to buy an IBM PC, the price is too high, and the entry barrier for consumers is too steep,” Tan said. “The locals are able to make PCs with the same functionality at less than half of the price.”
Since the sky-high prices were the biggest hurdle for PC demand and foreign brands were often sold at even higher prices than local brands, Intel China predicted that the local manufacturers would come to dominate the industry in China. So it decided to help along the development of local manufacturers, providing central processing units (CPUs) at reasonable prices, which would enable Chinese manufacturers to compete with foreign brands and consumers could buy PCs at reasonable prices. That was a bold move, given that the most profitable local PC maker, Lenovo, sold only an average of 100,000 units each year. Intel executives at headquarters were initially hesitant, but they gave the China team the go-ahead.
In China, the person tasked to implement the plan was Ian Yang, who had become Intel China’s president. Originally from Chongqing, Yang had attended college in the United States in the 1980s. As part of his program, he’d interned at Intel and never left. Upon returning to China in 1995, he wasted no time in cultivating relationships in the fledgling computer industry. At one client meeting, he met Yuanqing Yang, the general manager of Lenovo’s PC division. The two hit it off, with Ian frequently visiting Yuanqing at his office.
Intel launched its Pentium series in 1993 and, with the help of Microsoft’s Windows operating systems, inaugurated the “Wintel” era. Yet by 1995, all the PC makers in China, foreign and local, were still selling computers with Intel’s less powerful 386 or 486 microprocessors. The thinking was that those were good enough for an immature market like China. Ian was determined to persuade Yuanqing to adopt the Pentium for Lenovo’s computers. But, after many attempts, his pitch wasn’t working. Then one day Yuanqing unexpectedly called to say he was running out of 486 chips and wanted to upgrade. One of the main reasons was Intel had cut the prices of its Pentium chips recently, and Yuanqing saw the opportunity to improve his PCs but still make a profit.7 The order came with a condition—he wanted a big discount. Ian and Tan had to fly to US headquarters for approval.
In 1996, Lenovo released PCs with Intel’s Pentium chips, priced at 10,000 RMB ($1,200). That brought an enthusiastic consumer response: Lenovo sold 240,000 units that year and surpassed Compaq and IBM, becoming the leading PC maker in China, with a 10% market share.8 Inspired by Lenovo’s success, other local PC makers followed the same strategy. With Intel’s support, more local manufacturers sprung up, competing head-to-head with foreign brands. The competition culminated in 2005, when Lenovo acquired IBM’s PC division and the famous ThinkPad brand.
Intel’s alliance with Lenovo and other Chinese manufacturers went far beyond providing chips. As a well-known multinational, with the admired Grove as its chairman, Intel was often asked by its young Chinese partners for advice. Tan recalled giving talks on handling vendor relationships at Lenovo. He and his colleagues were willing to help out because they believed that building guanxi could only bring benefits.
The emergence of local PC makers and their affordable computers helped increase the Chinese appetite for PCs. Intel, one of a handful of B2B companies that had managed to make itself a household name in the United States, had well-honed skills for pitching to consumers directly and brought those to bear. It threw Chinese “PC parties,” marketing events intended to amplify demand.
To educate tech-naive Chinese consumers, Intel would hold the events every weekend, inviting people to play with and learn about computers. Of course, alerting them to “Intel inside”—the company’s storied marketing slogan—was an indispensable part of those “educational” sessions. Intel also arranged for local PC manufacturers to sell their products at the events at deep discounts, sometimes half of the market price. The parties drew crowds; one held at the Beijing library attracted over 40,000 people. Another, which took place on one of Chengdu’s electronics streets, was attended by over 200,000.9 Over several years, Intel hosted around 1,000 PC parties in multiple cities.
Tan credited the success to the team’s execution—“especially Jason Chen.” According to Tan, Chen didn’t miss a major PC party. He traveled around China and even to Tibet and stayed in cheap hotels to ensure the parties and other marketing campaigns came off as planned. With his efforts, channel sales grew to be one-third of Intel’s revenue. Chen later became a vice president of Intel, overseeing global sales and marketing, and is now CEO of Acer, a Taiwanese electronics company.
Intel had little difficulty attracting talent in China. A Shanghai newspaper once rated it as the most desirable foreign employer in the city. That resulted in a government affairs team that other multinationals envied. Navigating China’s nontransparent, unpredictable regulations wasn’t easy, said Wangli Moser, a former head of the team, but the one lodestar was to “align your strategy with China’s national agenda.”10
China had focused economic development on coastal cities, leaving underdeveloped its vast western regions. In 2000, the central government issued a “go west” policy, encouraging domestic and international enterprises to invest in such hinterland hubs as Chengdu, Chognqing, and Xi’an. The majority of multinational companies hugged the urbanized East Coast at the time.
Intel became the first to heed the government’s directive. It was already planning an assembly and testing plant in China. It had previously set up a lab and a testing and packaging facility in Shanghai, so Shanghai would’ve made sense as the location for the plant. But Intel chose Chengdu, 1,000 miles west of Shanghai and with much less developed infrastructure and transportation.
Chengdu was better known as the home of China’s famed pandas (the country’s research center for panda breeding is there) than as a center of commerce. Intel initially invested $375 million in the plant and then added $150 million and an additional $75 million. When completed in 2005, the plant was the biggest operation by a multinational in Chengdu. Later, it became one of Intel’s largest assembly and testing sites—one of every two laptops in the world was configured with a chip processed by this plant. Intel’s suppliers, partners, and other multinationals (for example, Lenovo, IBM, Dell, Siemens, Texas Instruments, Philips) also flocked to Chengdu. Even rival Advanced Micro Devices created a research and sales center in the city. These days, Chengdu aspires to be China’s Silicon Valley, and its former mayor, Honglin Ge, considers Intel’s arrival the impetus of that dream.11
Picking Chengdu wasn’t just smart politics for Intel; there were financial payoffs of cheap land, lower wages, and a tax reduction. And the political gratitude kept coming. A former Intel China supply-chain executive recalled her negotiations with China Telecom. The state-owned monopoly was known for its bruising negotiations. But the former executive said it took a different approach with Intel. China Telecom asked her what terms Intel wanted and agreed to them—treatment she had never received in other deals she’d handled.
Intel had figured out what the Chinese government valued most—the country’s long-term growth—and demonstrated its commitment to that. Intel was rewarded accordingly. In 2015, when its rival, Qualcomm, was fined $975 million for violating China’s antimonopoly law,12 Intel came away untouched, despite having dominated the Chinese semiconductor industry for decades.
Getting Intel’s top executives in California to agree to Chengdu wasn’t a sure thing; the company was also considering options in Thailand, Vietnam, and the Philippines. But Tan, by then Intel China’s president, kept lobbying in Santa Clara. Often people thousands of miles away can’t really understand the on-the-ground reality of a distant and culturally distinct market like China. That can lead to conflicts. Intel was no exception. Tan recalled being challenged or even embarrassed many times. This was, after all, a company whose chairman, Grove, had titled his memoir Only the Paranoid Survive. Skeptical questioning was expected. “But you can’t be afraid to take the risk, address the senior management team and give them the rationale,” Tan said.
Fortunately for Intel, Tan’s bosses, especially Grove, retained their faith in China’s potential. Tan first met Grove while working as Intel’s general manager for Southeast Asia. Shortly afterward, he found himself exchanging emails with Grove, which stunned him after spending 10 years at IBM, a company notorious for its hierarchical reporting lines. Grove encouraged him to be brash. Once Tan was interrupted and ignored while giving a presentation at a senior management meeting at headquarters, and Grove urged him to assert himself, to make sure the view from China was heard. What’s more, Grove, who was CEO until 1998 and chairman until 2004, insisted that every board member visit China at least once a year. Though Tan groused about becoming a “travel agency,” he knew how valuable it was to have those kinds of relationships. Craig Barrett, Grove’s successor as CEO, witnessed a PC party in Chengdu with 200,000 attendees, and Paul Otellini, who took over from Barrett in 2005, became convinced that China was critical to Intel, insulating it against slowdowns in the US market.
China didn’t disappoint. The market’s growth ratified Intel’s bet. In the mid-1990s, the entire country had only about 600,000 PCs. By 2000, that number had jumped to more than 7 million. Then in 2012, China topped the United States as the largest PC market worldwide, with over 66 million shipments. Intel’s business in China swelled accordingly, growing from $75 million in 1996 to $20.16 billion in 2019.13
Of course, China’s potential was no secret; it appealed to Intel’s rivals too. Advanced Micro Devices, better known as AMD, entered China in 1993 but hedged its commitment. Its management team was primarily based in Hong Kong, leaving only the sales force in mainland China. In 2002, AMD changed strategy, establishing a China headquarters in Beijing and assembling a local team, including a Chinese CEO. It started a price war, forcing Intel to cut prices as much as 30%. It even broke into Intel’s longtime partnerships with international brands such as Dell and Hewlett-Packard. They both launched PCs in China with AMD microprocessors. In just a few years, AMD managed to seize 15% of the PC market.
While scrapping with AMD, Intel also found itself in a spat with the Chinese government. In 2004, China announced a homegrown standard for wireless technology: the wired authentication and privacy infrastructure (WAPI). All international companies that sold computers, microprocessors, and other wireless products in China would have to adhere to it and work with Chinese companies selected by the government to implement the standard. “That amounts to opening your chip, and all your intellectual property would be gone,” Tan said. Intel opted to resist, announcing it wouldn’t provide a WAPI-compliant chipset. Tan and his team spent a year and a half fighting the new standard, and the wrangle eventually landed at the WTO, which adjudicates trade disputes. China ended up dropping the standard. Though Intel won, a rupture had opened between the company and Beijing. To close it, Tan’s bosses in California decided to “give China the fab they always wanted.”
A “fab” is a semiconductor fabrication plant, and it houses a chip maker’s intellectual property. Given China’s lack of protection for intellectual property at the time, none of the foreign chip makers had set up fabs there. Intel’s fabs were mainly located in Western countries, and a fab required a hefty investment. China had been asking Intel to construct one since the 1990s in hopes of boosting its semiconductor industry, but Intel had held back. In March 2007, it announced it would spend $2.5 billion to build a chip-manufacturing plant in China, then the country’s single largest foreign investment to date. Intel again displayed its facility with corporate diplomacy by choosing to locate the plant in Dalian, a northeastern city where the central government had called for investment.
The plant opened three years later with fanfare, but its reality didn’t match the public relations. Concerned about its technologies and facing US restrictions on technology exports, Intel had brought to China a technology two generations behind its production processes elsewhere. Its Western factories were producing its most advanced 22-nanometer chips, while the Chinese plant could make only 65-nanometer chips, mainly used for the low end of the market. As Intel upgraded its technology elsewhere, the chips produced in Dalian became obsolete, losing appeal for potential customers. As a result, the plant’s output was low, and the large investment didn’t yield a return.
Several years later, Intel converted the plant to the production of nonvolatile memory chips, which don’t require power to retain data. These chips were widely used in smartphones, tablets, and data centers, a rapidly growing segment in China. As memory chips didn’t require the same level of sophistication as microprocessors, the US government didn’t restrict technology transfers to the plant. Intel now could take more advanced technologies, such as 3D XPoint, to Dalian and thus compete with Samsung, SK Hynix, and Toshiba, the main players in this field. Nonvolatile memory chips were a small part of Intel’s portfolio, accounting for about 6% of total revenue, but the business grew at over 20% in both 2017 and 2018.14
Tan left Intel in 2008 after the original Dalian deal. During his 12 years in China, Intel had become one of the most prominent tech companies there. It had committed $4.3 billion to the country, and that investment yielded significant returns. Annual China revenue reached $7.5 billion in 2008, contributing 15%–20% of global revenue. Intel China also employed over 7,000 people in 16 locations. And it was a major venture capital investor, having staked 50 companies, including sohu.com, one of the earliest internet companies in China to go public on NASDAQ. More importantly, it dominated the semiconductor industry, with an approximately 85% market share. The role of AMD, Tan had joked, was to save Intel from accusations of a monopoly. Though AMD had good technology, it hadn’t achieved the market acceptance in China that Intel had. Intel had built a brand in China. “When you build brand acceptance, it’s not easy for a newcomer to come to your field to knock you off,” Tan said.
Intel did make smart choices, but its successes during this period also related to China’s stumbles in developing its own semiconductor industry. China’s efforts dated back to the 1950s, when the technology was invented in the United States. But in the 1960s and 1970s, the political turmoil known as the Cultural Revolution waylaid the country. Only in the 1990s did the central government begin to try to revitalize the semiconductor industry. The results were lackluster.
Unlike in more basic industries, such as clothing or home appliances, where China quickly caught up, semiconductors required advanced technological capabilities that China didn’t have. And it couldn’t acquire them from the United States, South Korea, or Japan, as these countries blocked access to their know-how and thwarted Chinese acquisition attempts. A shortage of talent was another problem. China was graduating 500,000 engineering students a year, but few had the background to work in semiconductors. Becoming an expert in semiconductor engineering is a long-term commitment, and China’s best engineers saw more immediate returns in the booming internet economy. Making chips also required a massive up-front investment and the financial wherewithal to wait several years to see a return. Chinese companies didn’t then have that capacity, and investment from the central government was insufficient and fragmented. At one point, the government had invested in 130 fabrication sites across more than 15 provinces, none of which worked out.15
More recently, the Chinese semiconductor industry has burgeoned, creating challenges for Intel. Local companies have grown from start-up to sturdy in just a few years. Semiconductor Manufacturing International Corp. (SMIC) in Shanghai has become the world’s fourth-largest chip maker in just a few years. Spreadtrum and Hisilicon have risen quickly in chip design. And Chinese demand for chips has swelled, consuming more than half of the world’s semiconductors annually. That growth has brought even more competition in the form of such firms as Taiwan Semiconductor Manufacturing Co., a maker of chips for the iPhone and iPad, and Nvidia, a leading chip maker for artificial intelligence.
The Chinese government also has increased its investment in the industry, aiming to create national champions. In 2014, it set up a $21.9 billion investment fund to boost design and manufacturing capabilities. In 2018, after the Trump administration banned US companies from selling components to ZTE, a Chinese telecommunications giant, the fund received $47.4 billion for a second round of investment. It has also taken capital from the private sector and has been run with a market-based approach, with a goal of making profits. China has also been able to attract talent from Taiwan. From 2014 through 2018, nearly 1,000 Taiwanese engineers moved to mainland China.16 “It’s relatively easier to get talent from Taiwan, as we share the same language and culture,” said an Intel R&D leader, who spoke on the condition of anonymity.17
Intel largely missed out on the mobile revolution globally; Qualcomm still dominates China’s mobile market. But Intel has diversified its operations in other areas, including the rapidly growing data-center demand from such leading Chinese internet companies as Alibaba, Tencent, Baidu, and JD.com. The first three companies, along with Facebook, Google, Microsoft, and Amazon, are Intel’s so-called Super 7, customers so large and important that they get access to chips before the designs are released to other customers.18 Data-center business accounts for over 30% of Intel’s total revenue, and, according to the R&D leader, China probably accounts for 30% of that, second only to the United States. “Chinese internet companies’ tech know-how still lags behind that of the US companies,” the same R&D executive said. “So Intel has more influence on Alibaba and Tencent than on Google or Amazon. Future opportunities lie here.”
Beyond the data-center business, Intel keeps investing in China. In 2014, it spent $1.5 billion to buy 20% of Tsinghua Unigroup, a state-affiliated company that owned China’s top chip designers, Spreadtrum Communications and RDA Microelectronics. Intel also injected $100 million into Tsinghua University and Montage, a subsidiary of China’s state-owned tech company, CEC, to form a joint venture on chip design. In 2018, after breaking up with Micron, its partner for developing nonvolatile memory chips, Intel partnered with Tsinghua Unigroup, providing memory chips for products like microSD cards and solid-state drives. And China remains the world’s largest PC market, continuing to generate a big chunk of revenue for Intel.
Whether its success can continue is, of course, to be determined. Tan, for his part, isn’t sure. “Nobody can dominate a market forever,” he said. “Intel has done that for a long time. I do believe ultimately the Chinese challenge would come up somewhere along the line. It’s just a matter of time.”
Intel was one of the earliest Western companies to venture into China and has realized significant success by any measure. These factors were critical to the outcome:
These elements and other factors in the framework are noted in Table 8.1, along with our subjective assessment of the decisive factors.
|
Factor |
Explanation |
|
|---|---|---|
|
Demand |
▲▲▲ |
The computing revolution occurred in China, as in the rest of the world, fueling strong demand for microprocessors. |
|
Access to market |
▲ |
The Chinese government provided unfettered access to its markets on account of Intel’s willingness to align with national priorities. |
|
Advantage |
▲▲▲ |
Intel’s chips and production processes provided a wide moat, but Intel bolstered its competitive advantage by building its brand in China. |
|
Commitment |
▲▲ |
Intel has demonstrated a strong commitment to China, born from Andy Grove’s vision. As a pioneer, the company endured many unprofitable years, yet it kept investing as China grew into the world’s largest semiconductor consumer. |
|
Governance |
▲ |
Intel China was tightly integrated with headquarters because the technical requirements are too rarefied and the cost of fabs too great to have more than one global hub. |
|
Leadership |
Intel’s local leadership showed savvy about China’s politics and regulations as well as the diplomatic skills to succeed in a large, complex corporation. Both Wee Theng Tan and Ian Yang had influence in California. Tan enjoyed Grove’s and other top managers’ support, and Yang had Otellini’s ear, later rising to be a member of Intel’s executive management team. |
|
|
Strategy |
▲ |
Intel brought to China unparalleled technology, aligned its expansion plans with the government’s national agenda, and built up its local brand. |
|
Product |
The technical complexity of the product and its manufacture does not allow for significant localization. However, the technical requirements for microprocessors are relatively uniform globally. |
|
|
Agility |
Intel proved that if local leaders are sufficiently skilled and persistent, a company can be agile even with global coordination. Tan reported to the general counsel, but he had a second reporting line to CEO Paul Otellini, whose influence he leveraged when needed. |
|
|
Luck |
▲ |
Intel was propelled by a multidecade megatrend favoring computing technology. |
|
Note: Upward-pointing triangles indicate a positive factor. Downward-pointing triangles indicate a negative factor. The number of triangles is our subjective assessment of the relative importance of the factor. We omit indicators for those factors that we do not believe were significant for this case. |
||
We now turn to the entirely different success story of Ermenegildo Zegna, a brand born in the Italian Alps more than a century ago.