From Windfalls to Pitfalls: Qualcomm’s China Conundrum - MacroPolo From Windfalls to Pitfalls: Qualcomm’s China Conundrum - MacroPolo
From Windfalls to Pitfalls: Qualcomm’s China Conundrum

For many multinational businesses, the launch of China’s Reform and Opening forty years ago seemed nothing short of an opportunity of the century. The potential to tap a market of nearly one billion customers, whose rising incomes invariably meant they would be able to afford more foreign products, was mind boggling. For KFC and other first movers like it, the bet on Chinese consumers paid off handsomely. Yet for other multinationals, the opportunity of the century turned out to have numerous strings attached, particularly in the realm of technology and market access.

That’s because the opening up of China’s market was never meant to be a free-for-all—it was a gambit always firmly anchored in the country’s own economic and political priorities and interests. From Beijing’s vantage point, it made perfect sense to gradually attach conditions to foreign direct investment (FDI) over time, because the underlying purpose of attracting FDI was to use it as an accelerant to catch up to advanced economies. And of course, the conditions attached to FDI changed as the Chinese economy, and its attendant priorities, evolved.

Today China is no longer a capital scarce country but a market that craves technology. This should hardly come as a surprise because it reflects dramatic changes in the Chinese economy. Of the three main inputs to economic growth—labor, capital, and technology—China has been deploying the first two to great effect. The country’s tremendous growth over the past forty years has built up an enormous capital base, even as its labor market is tightening on the back of demographic shifts. So the dividends from capital and labor are diminishing, leaving China little choice but to rely more and more on technology inputs to improve efficiency and economic productivity.

While capital and labor were comparative advantages for China, technology was, and continues to be, its notable deficiency. Recognizing its technological gap with industrialized economies, the Chinese government’s attitude on FDI increasingly turned to demanding technology and intellectual property (IP) transfers to support the next stage of growth. Such a shift has been exemplified by the insistence on joint ventures, in which the foreign partner usually had a minority stake and was expected to transfer, at a minimum, know-how and expertise that could strengthen domestic industry.

Of course, the strategy of “you can continue to profit in our market if you help us gain a technological edge” has not sat well with multinationals from the United States to Europe. When boiled down, transferring critical technology is tantamount to creating your own future competitors, and no company would willingly do that. Needless to say, China’s latest turn to focus on technology has become highly controversial and is at the heart of tensions between the United States and China.

In August 2017, the US Trade Representative Office launched a year-long investigation into China’s practices in technology transfer and IP theft. What’s more, the trade war the Trump administration launched against China is widely considered part of a strategy to get Beijing to modify its behavior on technology transfers.

These issues have seemingly come to a boil overnight. But in fact, they have percolated beneath the surface for years, if not decades. China’s fixation on gaining technological leadership is hardly new and has always been a main purpose of Reform and Opening. In fact, as early as 1987, a concerned US congress had demanded extensive studies on China’s technology transfer practices. What has changed are China’s own capabilities and its goals for technology acquisition. Not only are Chinese companies now capable of developing their own leading technologies, they are also increasingly demanding the crown jewels of foreign technology firms.

On the flip side, these longstanding concerns didn’t obscure the vast opportunities China’s Reform and Opening brought to multinationals, including technology-intensive firms. For a certain set of technology companies in particular, it wasn’t simply about the revenue potential of selling to a market that was a quarter of humanity. They also saw China as a unique opportunity to cement their technology standards to dominate global market share. These companies were playing a long game, with China being the focal point of the strategy.

One American technology giant that’s emblematic of both the enormous windfalls and eventual pitfalls of operating in China is Qualcomm. In the early days, Qualcomm had pushed its products, technologies, and standards into the China market, at times against the government’s economic agenda. Its efforts yielded tremendous commercial success and allowed the company to gain dominance in global telecommunication standards for decades.

Qualcomm’s very success, however, was also partly responsible for its own loss of momentum in the China market. It is tempting to blame Qualcomm’s recent troubles—from fighting off a hostile takeover from Singapore-based Broadcom to scrapping its attempted acquisition of NXP because the Chinese authority blocked it—as simply collateral from the ongoing US-China trade war. But that would be overly simplistic and skirts the company’s storied and complicated tenure in the China market.

Beijing’s blocking of the NXP bid was bound to happen, irrespective of the trade war. At its core, this isn’t about any single deal, but is a logical outcome of a brewing battle—between Qualcomm and China’s rising technological ambitions—over the future of international telecom standards and market leadership. Indeed, Qualcomm’s meteoric rise and gradual descent in China is emblematic of the country’s transformation from a market that passively accepted Western companies’ standards to a contender in the global technological race.

Qualcomm’s 2G Windfall in China

Zhu Rongji, China’s firebrand premier, wrote on the margins of a memo in March 1999, “Please have China Unicom consider adopting the CDMA standard and work with American companies.” This marked a major victory for Qualcomm’s seven-year push since 1992 to get a firm foothold in the China market, giving the company a significant edge in the global competition for second-generation (2G) cellular standards. To understand why that’s important, a brief detour into the development of 2G standards is warranted.

CDMA vs. GSM

The 1980s was a period in which global telecom standards were transitioning from analog to 2G. In the analog age, each user’s cell phone call was assigned a channel in which a single call could be transported. Since spectrum is a finite resource, the number of calls that can be made at the same time is limited. To put it differently, a highway is only so wide, which means only a certain number of cars can simultaneously fit across it before space runs out. Much like physical infrastructure, the constraints imposed by analog infrastructure meant that it could not accommodate huge volumes of calls and data.

The breakthrough in 2G technologies was that it allowed for multiple calls on the same channel, thereby circumventing the constraints of limited spectrum space. At the time, most of the world focused on a radio transmission technology called Time Division Multiple Access (TDMA). This technology improved spectrum usage efficiency by dividing the channel into multiple time slots and assigning them to different calls on the same radio channel. It was a solution to maximize the usage of channel space that was often wasted or under-utilized during a phone call. It is essentially the equivalent of allowing multiple cars to run on different schedules in the same lane on the highway, except it’s for multiple radio transmissions over a digital highway.

Qualcomm, founded in 1985 by UC San Diego professor Irwin Jacobs, pioneered another method that came to be known as Code Division Multiple Access (CDMA). This technology assigned each call with a code, and that call is then disassembled, transmitted, and then reassembled at the receiver’s end by using the code. Because of the code identifier, the call is no longer limited to stay in one channel but can hop on other channels as needed. Therefore, multiple users can speak at the same time. To use the highway analogy again, cars no longer need to stay in a single lane and can now also use other lanes whenever there is availability, allowing for more efficient use of all radio frequencies (see Figure 1).

Figure 1. How CDMA Works

figure2 1 - From Windfalls to Pitfalls: Qualcomm’s China Conundrum

Source: Qualcomm.

CDMA might be more technologically fitting for cellular communication than TDMA, but it was too late to the party. By the time Jacobs successfully prototyped the standard in 1989, the telecom industry had already sunk millions of dollars into TDMA infrastructure and incorporated it into a Global System for Mobile Communications (GSM). Moreover, since GSM was developed through a collaborative effort by multiple European telecom companies, there was more buy-in of the standard from industry stakeholders.

According to International Telecommunication Union’s report, by the end of 1998, roughly 130 million phones around the world were running on the GSM standard. Almost 90% of mobile phones in Europe, 35% in Asia, and 88% in Africa subscribed to the GSM standard. In contrast, Qualcomm’s 2G standard, which came to be called cdmaOne, had only 20 million subscribers with a minimal footprint on every continent in the world except for North America.

Just as the world was becoming more connected through mobile phones, Qualcomm appeared to be on the losing end of the standards competition. This is why seizing the China market was so integral to Qualcomm’s strategy. At the time, China was a market that had very low mobile penetration, and if hundreds of millions of Chinese started adopting devices with Qualcomm hardware, that would turn the tables on the 2G standards competition. In the standards race, the name of the game is market share.

Qualcomm CEO Jacobs reached the same conclusion when he visited China for the first time in 1992, the year Deng Xiaoping embarked on his “Southern Tour” to revive flagging economic reforms. Jacobs immediately set about finding ways to enter the China market. But Qualcomm’s initial overtures met resistance from a Chinese government that had already set its sights on the GSM standard.

The reason was simple. Beijing preferred GSM because it didn’t have to pay hefty royalties. Since GSM was a product of joint development among different countries, it had to be open source to enable collaboration. In contrast, the cdmaOne standard was developed virtually exclusively by Qualcomm, which meant it alone held numerous critical patents. Any manufacturer of CDMA-enabled mobile phones or network equipment had to pay for Qualcomm’s IP.

So in 1994, when the Chinese government decided the country needed to advance from analog to 2G standards like the rest of the world, it studied both GSM and cdmaOne as potential candidates and eventually decided to go with GSM for commercial applications.

Even so, that didn’t stop Qualcomm from testing the waters. The company recognized early on that it needed to enlist a key domestic constituency in China to support its efforts. It found an unconventional partner: the People’s Liberation Army (PLA). Although China had selected the GSM standard for commercial use, the PLA at the time was searching for a radio transmission technology that would be secure for military communication. Because Qualcomm’s CDMA technology was based on coded radio transmissions, it seemed like a good fit for what the PLA wanted.

While such a partnership would be unthinkable and forbidden today, the 1990s was a period in which the PLA had more latitude to engage in commercial activities that ranged from automobile manufacturing to pharmaceuticals and hotels.

In the Qualcomm case, the Ministry of Post and Communication (MoPC), the predecessor to the powerful Ministry of Industry and Information Technology (MIIT), ordered its local bureaus to set up a joint venture (JV) with local PLA divisions called “Great Wall.” The JV was granted a civil-military dual use license to experiment with cellular networks with the cdmaOne standard on the 800Mhz spectrum in four major cities: Beijing, Shanghai, Xi’an, and Guangzhou. By 1997, Qualcomm’s 2G networks for both commercial and military applications in these cities were up and running, with the potential to expand into other Chinese cities.

But the JV collapsed almost as soon as it was formed. Just a year after the 2G network went live, Chinese President Jiang Zemin issued an order that forbid the PLA from engaging in any commercial activities. The experimental CDMA network remained in place, however, though it never grew to cover more than half a million users, all of whom were later transferred to other networks.

The timing of this episode was peculiar. Some observers even suspected that the MoPC was anticipating this outcome all along and was setting up the JV to fail just to kill the CDMA standard in its infancy. In fact, while granting the CDMA dual-use license to Great Wall, the MoPC simultaneously accelerated the approval of a national GSM license on the 900MHz band to China Telecom, one of the national champions. MoPC’s true motivations will never be known, but one thing was clear enough to Qualcomm’s Jacobs, who said on the record that because the MoPC fully owned the 900 MHz band used for GSM, the ministry favored GSM.

These early setbacks didn’t dissuade Qualcomm from continuing its pursuit of the China market. The environment was different in 1999, when Beijing was wrapped up in intense negotiations to enter the World Trade Organization (WTO), which required winning Washington’s acceptance. Qualcomm, like many companies at the time, saw an opportunity to ramp up the pressure on China to open its market. From the Chinese vantage point, liberalizing the telecom sector could go a long way toward mollifying the United States and securing its support for WTO entry. And so, Premier Zhu inked his support for Qualcomm on March 2, 1999 as detailed above and gave his promise to the US delegation, led by Commerce Secretary Bill Daley, that was soon to arrive in Beijing.

Even with Zhu’s support for Qualcomm, negotiations were far from over. That’s because MoPC’s head at the time Wu Jichuan, a major proponent of China’s decision to adopt the GSM standard in 1994, was a hard-nosed negotiator. Wu wasn’t about to give in until Qualcomm met three demands: develop a new mobile phone model that can run on both CDMA and GSM networks; lower the royalty fees; and share the design of Qualcomm’s CDMA chipset. In other words, Wu wanted options to abandon CDMA at will, use the technology cheaply, and own the IP so that China can make its own chipsets.

Wu’s demands were of course a non-starter, and he probably knew it but wanted to play hardball anyway. These demands were viewed by the US government as China’s disingenuous attitude toward WTO entry. Daley raised the issue multiple times during broader talks with Chinese leaders. While underscoring the White House’s determination to push through key trade legislation that will support Beijing’s WTO entry, Daley also made it very clear that “among our commercial interests, this [Qualcomm] was a very important one.”

With US government support and intensive efforts, Qualcomm finally secured an in-person meeting with Premier Zhu. On October 6, 2000, Jacobs, Wu, Brent Scowcroft (former National Security Advisor who heads the consultancy The Scowcroft Group), and Yang Xiaozu (Chairman of China Unicom) all met in a conference room at Tsinghua University. Zhu was there to broker a compromise among the competing interests. He demanded all the parties present to write down a list of demands and disagreements and to “sort it out.”

At this critical juncture, China’s quest to enter the WTO was the priority, which meant Wu’s ambitious demands had to take a back seat for the time being. The impasse was broken: Qualcomm was allowed in the China market. In March 2002, a decade after Jacob’s first trip to China, China Unicom announced its commitment to deliver CDMA services to more than 350 cities.

When Success Comes Back to Bite

In the decade after Zhu lent his support to Qualcomm on the ledge of a memo, the American tech giant’s revenue stream from China grew from virtually nil to $2.4 billion, more than twice the amount from its home market and accounting for one-fifth of its global revenue (see Figure 2). Yet that very success led to two unintended consequences that would eventually turn the tables on Qualcomm: 1) It prompted China to nurse a grudge against the company, particularly toward its fee structure; 2) It clarified for Beijing that it needed to raise its game in global standards setting or else accept ponying up licensing fees in perpetuity.

Figure 2. Qualcomm’s Revenue Stream from China
Source: Qualcomm’s annual reports (1999 – 2008); author calculations.

But Qualcomm wasn’t done with profiting from 2G. It was hoping to extend its windfall in China to the 3G era, which officially commenced in January 2009 when Li Yizhong, the head of MIIT, announced China’s transition from 2G to 3G. In the months leading up to this announcement, MIIT had already been laying the groundwork by consolidating the state telecom industry from six major carriers to just three: China Mobile, China Unicom, and China Telecom. This “big three” competitive landscape, much like China’s state oil industry, is the one that endures today.

One of the rationales behind the industry restructuring was that each state giant would be awarded one of three competing 3G standards: WCDMA that evolved from GSM (license granted to China Unicom); CDMA2000 that evolved from Qualcomm’s own cdmaOne (license granted to China Telecom, which took over China Unicom’s CDMA network for $16 billion during the restructuring), and China’s homegrown Time Division Synchronous Code Division Multiple Access (TD-SCDMA) (license granted to China Mobile).

The “Double Dipping” Fee Structure

However you sliced it, this 3G standards landscape would significantly benefit Qualcomm. That’s because with anything that had the acronym “CDMA” in it, chances are Qualcomm owned some of the core IP since it was the original developer of the CDMA technology. According to Qualcomm’s 2009 financial report, both the WCDMA and CDMA2000 technologies were derived from CDMA and “are covered by our patents.” The company also claimed to hold critical patents for the TD-SCDMA standard.

With a firm grip on the core IP of the 3G era, Qualcomm made money by both licensing the IP and directly selling its own 3G-enabled chips to mobile phone vendors, who would still need to pay a royalty. Selling its own 3G chips may have been a larger contributor to the company’s revenue stream, but Qualcomm’s real profits were made from its licensing fees. By amassing hundreds of thousands of patents in cellular communication standards, Qualcomm’s fingerprints were virtually everywhere in the telecom industry. For years, the telecom industry had a running joke that while death and taxes are two certainties in life, paying royalties to Qualcomm was another certainty in the wireless industry.

Maintaining what amounted to a patent monopoly on 3G standards enabled Qualcomm to leverage a unique and highly lucrative licensing fee structure that is still largely in place today. It basically works like this: mobile phone manufacturers license Qualcomm’s technologies and pay the company royalties that are as much as 5% of the final sale price of the phone. This means the royalties increase with the phone price, even if Qualcomm’s technology inside the phone remains unchanged. In contrast, other telecom companies, such as Ericsson and Nokia, charge a flat fee for the specific technologies that licensees actually use.

To illustrate, if a basic mobile phone costs $400, then Qualcomm gets 5% of that in royalties, or $20 per phone. If the manufacturer decides to add a high-resolution camera, a bigger screen, or a sleeker case to soup up the phone, the price doubles to $800. Now the manufacturer has to pay $40 in royalties to Qualcomm even though the technologies licensed have not changed.

On top of paying royalties, as a 3G mobile phone manufacturer, you would either need to make your own 3G chips or buy from other chip makers. More likely than not, manufacturers end up buying chips from Qualcomm, so they have to pay the company again. This fee structure came to be known as “double dipping” and, needless to say, has irked many manufacturers.

Figure 3. Evolution of China’s Mobile Standards Adoption (1994 – 2014)
Source: Jefferies Equity Research.

This pricing strategy was also applied to the China market. But initially, it was foreign manufacturers that felt the brunt of Qualcomm’s fee structure. That’s because when the company entered China in 1999, the country was still a technological backwater incapable of producing quality mobile phones. Major carriers like China Unicom had to sign contracts with foreign manufacturers, such as Nokia and Ericsson, to import the phones. (Those manufacturers also relied on Qualcomm chips and IP, so were paying the company.)

It would take about another decade for Chinese manufacturers to acquire the capability to produce low end to “good enough” 3G mobile phones (see Figure 3). And that was when Chinese producers started to directly feel the pinch of Qualcomm’s double dipping strategy. Since Qualcomm held patents for all three 3G standards in China, manufacturers had little choice but to pay licensing fees. In addition, few Chinese manufacturers had the ability to make their own 3G chips, so they had to rely on foreign imports, including Qualcomm’s.

Figure 4. Mobile Phone’s Growth in China, 2000-2014 (in millions)
Source: Statista.

Demand for 3G mobile phones skyrocketed in China after 2009, and has grown 15 times in the 15 years since Qualcomm’s official entry into the China market. This led to another windfall for Qualcomm. By 2010, Qualcomm’s revenue from China reached $3 billion, surpassing that of South Korea. Just four years later, the company’s China market revenue for the first time exceeded combined revenue from the rest of the world, including the United States.

Yet as Qualcomm’s profit margins widened, Chinese mobile phone makers’ margins were being squeezed. Domestic original equipment manufacturers (OEMs) already had to keep prices low because of fierce competition that often resulted in price wars. Meanwhile, as Chinese manufacturers started to make more expensive phones with better displays and high-end cameras, they discovered that Qualcomm’s licensing fee kept on increasing, even though they were using the same IP.

Unsurprisingly, this did not sit well with Chinese OEMs nor with the Chinese government. What’s more, it wasn’t exclusive to China. The fee structure irritated many global mobile phone makers, especially as they were under the pressure of product cycles to constantly deliver new and more expensive features such as larger and better displays and fancier cameras. The price of their products went up, and like Chinese manufacturers, they also suddenly found themselves paying Qualcomm double or even triple the royalties for licensing essentially the same technologies.

In subsequent years, Qualcomm’s double dipping strategy would become a major source of conflict, not only in the China market but also in the broader telecom industry. Qualcomm has long argued since the 1990s that no matter what went in the phones, it was their technology that enabled them all. But this argument gained less traction in the 2000s. In the eyes of Qualcomm’s customers, the company’s technology was contributing less value to mobile phones yet the licensing fees kept on rising.

By 2015, Qualcomm was embroiled in controversies or being fined by regulators in Taiwan ($773 million), South Korea ($1.23 billion), and Europe ($853 million). Even Apple jumped on the bandwagon and fought multi-year legal battles with Qualcomm over this issue, arguing that it was engaged in “illegal practices.” By mid-2018, Apple announced that it would manufacture its own chips for the iPhone, completely moving away from Qualcomm’s chips.

It didn’t help matters that Qualcomm, at times, may have rubbed salt in the wound. In its 2014 annual financial report, the company noted, “particularly in China, certain licensees have disputed or underreported royalties owed to us under their license agreements with us, and certain companies have yet to enter into or delayed entering into license agreements with us for their use of our intellectual property, and such licensees and/or companies may continue to do so in the future.” While some Chinese manufacturers certainly found ways to circumvent royalty payments, Qualcomm still had all the chips in its corner.

Complaints in China grew louder and became harder to ignore for Chinese regulators. So they sprang into action. In November 2013, months before the issuance of 4G licenses, the National Development and Reform Commission (NDRC) initiated an investigation into whether Qualcomm’s licensing practices violated China’s Anti-Monopoly Law, which took effect in 2007.

As the investigation proceeded, Qualcomm was preparing for a fine of 1% to 10% of its previous year’s revenue and other remedies.

After the 14-month investigation concluded, Chinese regulators slapped a $975 million fine, equivalent to 3.7% of the company’s 2014 revenue, the largest fine ever in China for monopolistic practices. On top of the fine, Qualcomm agreed to lower its royalty rates on 3G devices to 5% and 3.5% for 4G devices, using a royalty base of 65% of the final sale price as opposed to 100%. So the company effectively lowered its royalty rates to 3.3% and 2.3%, respectively, on 3G and 4G devices, lower than in other foreign markets including India. In response, Qualcomm’s annual dividend saw a $0.60 cents per share reduction.

China’s first failed attempt on standards setting

What resulted was beyond Qualcomm’s expectations, but such an outcome should not have been a surprise. The writing was already on the wall four years before the investigation, when MIIT in 2009 unveiled its grand designs on promoting 3G standards.

The lack of domestically developed IP in mobile standards has clearly frustrated Chinese regulators to no end. They learned first-hand from Qualcomm how having a near monopoly on core technology patents is directly linked to market position and profits. From Beijing’s perspective, why should China passively accept standards when it had the market size to come up with competing standards to Qualcomm’s?

MIIT’s answer to that question was to order the China Academy of Telecommunications Technology (CATT) to collaborate with Germany’s Siemens to develop a new 3G standard that would come to be known as TD-SCDMA. In 2001, backed by all three Chinese carriers, TD-SCDMA was approved to join the global 3G standards governing body, the 3G Partnership Project (3GPP). However, it was China Mobile that was granted the TD-SCDMA license. Of the three state carriers, China Mobile was MIIT’s favorite and had dominated the 2G market (see Figure 5). But being the favorite also meant that China Mobile had the unenviable task of ensuring that the indigenous but commercially unproven 3G standard becomes a success.

Figure 5. China Mobile Lost Market Share from 2G to 3G
Source: Caixin.

Except the opposite happened. TD-SCDMA turned out to be far less developed than the prevailing 3G standards WCDMA or CDMA2000, both of which had proven to be commercially viable for years. No carrier outside of China ever used TD-SCDMA and even Chinese carriers, including China Mobile itself, sought to disassociate themselves with it.

But MIIT didn’t want to give up hope and ordered China Mobile to develop an entire 3G network based on TD-SCDMA. This was ostensibly a last-ditch effort to bolster the homegrown technology, but instead, China Mobile lost 10% of its market share over the four years it was being forced to support the weaker standard. “Other people had a head start and were running ahead of you on the main road. You can’t just give up, turn around and dig a separate lane,” as a Chinese telecom industry expert commented. “China’s TD-SCDMA led Chinese telecom companies to a detour from the mainstream.”

China’s effort to introduce a domestic 3G standard ended in failure, but its appetite for reducing dependence on foreign core technologies remained as strong as ever. The Chinese government had learned a hard lesson, but did not exactly hide its ambition to have another go at setting standards. It bided its time and largely went with the flow as the world moved to 4G standards.

But even then, two Chinese companies, Huawei and ZTE, had started to make some waves. According to Jefferies Equity Analysis, ZTE held 6% and Huawei 1% of all patents in 4G standards. That would change quickly as Huawei matured and trained its sights directly on Qualcomm. If Huawei’s effort to lead in global 5G standards succeeds, it will prove disruptive for Qualcomm’s business in China.

The 5G Race Is On

On July 26, 2018, China’s telecom giant Huawei presented a medal to Dr. Erdal Arikan, a Turkish expert in polar coding theory. The medal was designed and crafted by Monnaie de Paris with a Baccarat crystal. As extravagant as the medal was, its value was negligible compared to the royalties Huawei was about to collect by developing its own IP based on Arikan’s theory.

Huawei had been quietly pouring 15% of its annual revenue, or more than $61 billion, over the past decade to develop technologies that have the potential to become global 5G telecom standards. One such technology is based on Arikan’s polar coding theory. To understand why that’s important, a brief explanation of 5G standard development is needed.

Just like in the 2G and 3G eras, delegates from the world’s major telecom operators, networks, terminals and chipset vendors, and internet companies regularly met at 3GPP, the international governing body of telecom standards, to pitch technical solutions to various 5G challenges. One of the main problems that needed to be solved was reducing data transmission errors as the volume of data grew exponentially. More errors have crept into large volumes of data due to noise, interference, and fading.

A method called channel coding—which is basically repeating a piece of data to reduce errors—was developed to overcome the problem. To oversimplify, channel coding according to MIT basically works like this: if you were trying to transmit a message with only three bits, like 001, you could send it three times “001001001”. If an error crept in, and 001011001 was received instead, you could be reasonably sure that the correct string was 001.

Arikan’s polar coding theory is one such channel coding method that could be applied to improve data accuracy. So Huawei decided to back polar coding and invested billions into its commercialization. Within the course of eight years, this relatively new theory had become a viable solution in practice, surprising even Arikan.

The direct competitor to polar coding technology is, no surprise, Qualcomm’s low-density parity check (LDPC) technology. Compared to polar coding, LDPC has a much longer track record of commercial viability. The theory of LDPC was first introduced in 1963, 45 years earlier than polar coding. In subsequent decades, Qualcomm pioneered LDPC’s commercial application and developed critical patents. By the time polar coding was introduced in 2008, applications of LDPC had already been deployed in the real world.

The contest over whether LDPC or polar coding would become the global 5G standard for channel coding erupted on November 14, 2016 in Nevada, where 3GPP held meetings to vote on accepting a channel coding solution.

Debate was intense at the meeting, with companies picking sides. Western companies, led by Qualcomm, largely fell in line behind LDPC while numerous Asian manufacturers favored Huawei-backed polar coding. In an interview to the Wall Street Journal, an expert who was at the meeting recalled, “the Chinese decided this was important. This was one of the biggest political battles we’ve ever seen.”

Eventually, the two sides reached a compromise: both polar coding and LDPC would be adopted as part of the channel coding standard. This was a victory for Huawei as it gained a critical patent in the 5G global standard.

More such battles have been fought, and Chinese telecom companies have made considerable strides in establishing a foothold in 5G standards. According to technology research firm LexInnova, Huawei and ZTE today hold about 10% of critical 5G patents, compared to 15% for Qualcomm (see Figure 6).

Figure 6. Shares of Critical 5G Patents by Company
Source: LexInnova.

To some extent, the global standards race is a zero sum game in that only one technology will be ultimately suited to addressing one critical technical challenge. And the incentives are such that, like Qualcomm, each company is aiming for market dominance, not just market share. Therefore, the very nature of this competition means that Qualcomm increasingly finds its own dominance being chipped away by the emergence of formidable rivals—some of which are Chinese manufacturers who were once Qualcomm’s customers but are now using what they learned to compete with it.

As if fending off new competitors isn’t tough enough, Qualcomm also had to face pressure from the Chinese government to transfer its knowledge to Chinese companies. Although the government has long dangled the carrot of market access to get foreign companies to share certain technologies, the difference today is that the relative leverage has shifted.

Qualcomm still carries a lot of weight, but it is no longer the only player in town. Beijing has choices now, and if Qualcomm isn’t willing to play ball, the market share will go to a competing European firm or better yet, a rising Chinese company. This makes the trade-off challenging for Qualcomm: lose market share to Western tech giants today or lose market share to Chinese upstarts tomorrow.

Competition is also taking place in the area of hardware, namely advanced chipsets that are capable of supporting 5G data processing speeds. In fact, Moore’s Law’s famous prediction of computing speed doubling every two years was predicated on fitting ever more microscopic transistors on a chip. That’s because computing power is positively correlated with the number of transistors that can be piled onto a chip. The current generation of advanced mobile chipsets use 14-nanometer transistors.

But few Chinese companies have the ability to manufacture such chips. So eight months after NDRC slapped the fine on Qualcomm, the company agreed to form a JV with Huawei and China’s Semiconductor Manufacturing International Corp. (SMIC) to develop 14nm chips. This move was widely interpreted as a way to patch up relationships with the Chinese government, with little upside for Qualcomm otherwise.

These chips, however, quickly became obsolete. A true 5G network would enable users to download a full movie in 15 seconds, compared to 6 minutes in 4G. This means that the data processing capacity required for a 5G chip is much higher than that of 4G. The chips need to fit even more transistors, which means their size had to be reduced to at least 10nm.

Even global giants like Intel struggle with developing 10nm chips, let alone Chinese semiconductor fabricators. But Qualcomm in 2017 again decided to help SMIC’s subsidiary SJSemi to start the qualification of wafer bumping, a technique in chip manufacturing, to produce 10nm chips. This made SJSemi the first ever chip manufacturer in mainland China to enter the 10nm arena. Qualcomm at the time said that such collaboration “shows our commitment to support the upgrade of China’s local IC manufacturing industry and to better serve our Chinese customers.”

Currently, Samsung, Huawei, and Qualcomm are leading the pack in developing 5G chipsets. Huawei started its R&D efforts into 7nm processors in 2015 and has invested over $300 million in developing a prototype. On August 15, 2018, Samsung launched the first 10nm 5G chipset that’s fully compliant with 3GPP standards. Huawei immediately responded by announcing that it would launch its own 7nm 5G chipset Kirin 980 on August 31. Qualcomm, however, quietly launched its own 7nm Snapdragon chip ahead of Huawei on August 22.

New Battles on the Horizon

Qualcomm brought CDMA to China in the early days of Reform and Opening, even as the Chinese government had already decided to go in a different direction. But the American tech giant wasn’t defeated, using various leverage points like negotiations over China’s WTO entry to get into a market that was crucial to its long-term strategy.

Qualcomm’s persistence paid off handsomely: Beyond the billions of profits, without the China market, it would not have been able to dominate two generations of telecom standards. By having China adopt the 2G and 3G CDMA standards, Qualcomm’s market position in global telecom standards was cemented.

The American company’s success, however, left lasting impressions on the Chinese government and companies about the importance of leading in global telecom standards primarily through the development of indigenous IP. Qualcomm also didn’t help itself by alienating Chinese manufacturers and the telecom industry writ large with its lucrative fee structure that many viewed as unfair. In fact, China’s effort to set its own 3G standard with TD-SCDMA, albeit one that ended in failure, was a response to widespread domestic frustration over not having any influence in global standards.

Figure 7. Huawei vs. Qualcomm Patent Wall

patent wall 1 - From Windfalls to Pitfalls: Qualcomm’s China ConundrumQualcomm image source.

After a stellar run of 15 years in the China market, Qualcomm’s rise may be interrupted. As China’s telecom firms and mobile phone manufacturers have matured, and having absorbed the previous lessons of failure, they appear ready to challenge the industry leaders. For Chinese companies, Qualcomm’s experience taught them that if you win the patents race, you win the standards war. This is reflected in a Chinese company like Huawei, which has taken chapters from the Qualcomm playbook and has been obsessively filing patents (see Figure 7).

Qualcomm’s future prospect is arguably more uncertain than it has been in decades. It is stuck in a paradoxical position: the market that today contributes more than 60% of Qualcomm’s global revenue also happens to be the market that is most likely to challenge its dominant position. To make matters worse, this is coming at a crucial period of transition to the next-gen 5G standards in which no clear winner has been crowned.

This race is set to intensify, and so will the politics around it because technology is the main source of current US-China tensions. But ultimately, this is a competition between multinational companies—they are both proxies of respective national ambitions and potential collateral in the escalating conflict between their home countries. For Qualcomm, the battles it has fought and won so far in the China market appear to pale in comparison to the new battles on the horizon.