The Origins of an Entrepreneurship Column and the Shifting Tides of the Semiconductor Industry
DaTong Chen
A few years back (2012), I was invited to write a yearlong entrepreneurship column for the Tsinghua alumni magazine Shuimu Tsinghua. Several installments recounted my experiences founding OmniVision Technologies in Silicon Valley and SpreadTrum Communications in Shanghai — a sort of summary of the first half of my life. But who could have guessed that in recent years, both companies would be acquired, delisted from the U.S. markets, and brought back to the motherland to embark on entirely new journeys! Looking back on those days, memories flood in and emotions run high. So I’ve put these stories together to share with everyone.
(Note: In 2013–14, Tsinghua Unigroup successively acquired SpreadTrum Communications and RDA Microelectronics, delisting them from NASDAQ and merging them into UNISOC. In 2016, a domestic consortium led by Hua Capital (now Primarius Capital) acquired OmniVision Technologies, delisting it from NASDAQ; in 2019, it was folded into the domestically listed Will Semiconductor, making it the top chip design stock in China.)
I. Starting Up in Silicon Valley: The “Accidental Entrepreneur”
In the early 1990s, mainland Chinese students had just arrived in Silicon Valley. The top priority was finding a job to put food on the table — who dared dream of starting a company? My entrepreneurship was honestly a misunderstanding. It’s more accurate to call it “being entrepreneured.”
It was early 1995. I’d been a senior engineer at National Semiconductor for two years, working on advanced analog semiconductor process development. I’d learned everything there was to learn, the job was comfortable, I was well-regarded, the project manager did whatever I suggested, and the occasional patent bonus was enough to cover household expenses… But when things get too quiet, the mind gets restless. I suddenly wanted to switch fields and learn integrated circuit (IC) chip design. Going from semiconductor processing to chip design is like going from being a tailor to being a fashion designer — your creative space expands enormously. IC design was the glamour profession in Silicon Valley’s semiconductor world, but the difficulty goes without saying. Still, yours truly never shied away from a challenge. Besides, a mainland Chinese friend of mine, a chemical engineering PhD pushing fifty, had just successfully made the switch to IC design engineer. Compared to him, at least I had a proper semiconductor background.
My plan at the time was to spend six months reading six or seven chip design textbooks, memorize some jargon, polish up my résumé, dress it up a bit, and then go job hunting and hope for the best. I was only halfway through when Professor Wu Qiming (Tsinghua Radio Engineering Department, Class of '56, my master’s thesis advisor; his sister Wu Qidi, Tsinghua Radio Engineering Department, Class of '65, later became president of Tongji University and Vice Minister of Education) came to find me. He asked if I knew any bipolar IC experts who could help his small company (Opus) solve a technical problem. I’d just happened to read a book — Bipolar IC Design — so I put on a brave face, passed myself off as an expert, and said: “Let me give it a shot.”
At the time, Opus had landed a project to reverse-engineer a TV remote control chip. They’d dissected the original chip’s circuitry but couldn’t figure out how it worked. So every day after work, I’d go moonlight at Opus, then rush home late at night to frantically look things up. Two or three weeks later, the problem was solved. Felt pretty good about myself.
A month later, Opus CEO Mr. Hong came to me again and said: “We’re starting a new company. Want to join?” I asked: “What’s this new company going to make?” He answered: “CMOS image sensors (CIS).” Never heard of them. I asked again: “Would I be doing process or design?” He said: “Design.” I agreed on the spot — well, that saves me the trouble of job hunting! Two days later, Old Hong came back and said: “Since you’re joining the company, why don’t you be a co-founder?” I asked: “What’s a co-founder?”
He explained for ages, and I still didn’t really get it. Just like that, I became a “co-founder” of OmniVision Technologies — I’d been “entrepreneured.” Only later did I realize that being a co-founder is no joke. From that point on, the enormous pressure of the company’s success or failure followed me like a shadow, impossible to shake off. On the clock or off, every waking moment was spent worrying about the company. It truly was “choosing a different way of life” — stepping onto a road of no return!
After I’d “joined the gang,” another co-founder, Raymond Wu, who’d come to the U.S. from Taiwan, quietly revealed to me a jaw-dropping secret: the company’s seed money was raised through what he called a “half-bluff, half-hustle”! At the time, there was only one startup in the entire world developing CIS — a British company called VVL. Raymond, resourceful as ever, somehow got his hands on their engineering samples, wiped off the company markings, and passed them off as our OV “R&D samples.” Of course, these “samples” couldn’t be shown around Silicon Valley, so Raymond took them back to Taiwan and demoed them everywhere. Eventually, through a college classmate, he met Stanley Chui, the second son of the HCG ceramics family (yes, the bathroom fixtures people). Under Raymond’s persuasion, Stanley wanted to play around with high-tech and invested $2 million to establish OmniVision. (The above is solely Raymond’s account and has not been independently verified.) When I learned the truth, I broke out in a cold sweat — what kind of pirate ship had I boarded?! But getting on a ship is easier than getting off, so I had to grit my teeth and press on. If the company failed, this would become a “scandal.” But if it succeeded, it would probably become an “anecdote”!
OmniVision Technologies (OVT) was founded with four co-founders:
Shiao-Ying Hong was CEO, Raymond Wu handled marketing, T.C. Tshu was in charge of digital circuits, and yours truly was responsible for analog circuit design — which just happened to be the technical core of the CMOS image sensor (CIS).
With such a heavy burden on my shoulders, this fake expert’s heart was anything but confident. So I quickly called in reinforcements, inviting Zhang Zhongxuan, my classmate from Tsinghua’s Microelectronics Institute master’s/PhD program, to also be a co-founder. Zhang was an analog circuit wizard, already a senior design manager at a well-known Silicon Valley IC design company. But to my dismay, just a few weeks after joining, he was lured back by his former company with over a million dollars in stock options and a design director title. Alas, our little temple simply couldn’t accommodate such a great Buddha.
Left with no choice, I urgently recruited several junior alumni from Tsinghua’s Microelectronics Institute to form OVT’s core technical team, including: He Xinping (top of the Class of '80), Liu Jun (Class of '80), Yang Hongli and Dong Qi (top two of the Class of '82)… Brilliant talent, no question, but not one of them had actually designed an IC product before. Since I couldn’t rely on anyone else, I had to rely on myself. So this fake expert learned while teaching — whatever textbooks I finished, everyone else read in turn. When we didn’t understand something, we discussed it together. And somehow, within six months, we produced our first sample.
In 1995, only two or three startups worldwide were developing CIS. But in February 1996, the International Solid-State Circuits Conference (ISSCC) organized a CIS tutorial that was packed to the rafters. Within the next six months, twenty to thirty companies stormed into the field, including behemoths like Intel, HP, Sony, National Semiconductor, and others — each with teams of hundreds and budgets in the hundreds of millions. Overnight, the landscape shifted dramatically, surrounded by giant predators, danger lurking at every turn. By comparison, our little OVT outfit — “a dozen people with seven or eight rifles,” a ragtag operation — was laughably outmatched.
But since we were already on the boat with no way back, we had no choice but to grit our teeth and keep going. We Tsinghua engineering men had two traits: First, we weren’t afraid of hard work. Everyone worked 12+ hours a day (the company provided dinner) and six and a half days a week, so our development pace was blazing fast. Second, we didn’t believe in the impossible. We were all young and fearless, guerrilla fighters with unconventional tactics, bold in innovation. This allowed our products to vastly outperform those of the big European and American companies in performance, cost, and power consumption.
Although we took some detours on the technical development path, OVT successfully developed the world’s first single-chip color CMOS image sensor (CIS) in 1997. Compared to the traditional four-chip CCD image sensor, CIS offered improvements of tens to even hundreds of times in cost, size, and power consumption. Within just a few years, it triggered an industry revolution — a generational shift in technology — and made the dream of camera phones a reality.
From personal experience, I can tell you: the vast majority of big European, American, and Japanese companies are dinosaurs — paper tigers. I also came to understand the competitive law of survival: “the fast fish eats the slow fish.” That’s why in Silicon Valley, over 95% of all innovation is driven by small startups. And the big companies? They survive on brand recognition, distribution channels, and economies of scale, while continuously acquiring small companies to obtain new technologies.
By the late 1990s , CIS was being widely adopted in webcams, and OmniVision captured over half the global market share. In July 2000, it successfully went public on the U.S. NASDAQ. After that, yours truly got restless again, itching to go back to China and start something new. So I left OVT — but that’s another story.
Over the following decade-plus, He Xinping and Yang Hongli successively took over as COO of OVT, leading the company through continuous growth and keeping it at the forefront of the CIS industry. In recent years, OVT even supplied camera chips for Apple’s iPhone 4 and the new iPad.
Even more interesting — when you count them all up — a total of six CIS chip design companies were founded by graduates of Tsinghua’s Radio Engineering Department (including GalaxyCore in Shanghai, founded by Zhao Lixin, Class of '85). Together, they captured half the global market.
II. Returning to China: The Birth of SpreadTrum
In July 2000, after OmniVision went public on NASDAQ, my confidence was sky-high and I was itching to make a move: Since we Chinese students could build something in Silicon Valley, it was time to go back to China and give it a shot!
At the time, China’s semiconductor industry was in dire straits, struggling to get off the ground. After the 1990s, as international giants flooded in, China’s hundreds of state-owned semiconductor enterprises were nearly wiped out in the brutal market competition (save for a few defense-related companies). The nationally sponsored Hua Hong NEC (State Project 909, construction started in 1995) was hemorrhaging money year after year with a bleak outlook. The handful of IC design companies supporting it could only produce low-end products like transit cards and phone cards. In the surging global tide of digitization, China faced a critical vacuum in the semiconductor industry — the very backbone and core of the information society! The facts proved that the highly market-driven semiconductor industry was virtually impossible to sustain under the rigid state-planned economic system. The only hope was to encourage private innovative enterprises (and new market-oriented state companies) under market mechanisms to rebuild China’s semiconductor industry — a painful path of rebirth from the ashes!
In the year 2000, amid this crisis, the Ministry of Information Industry issued Document No. 18, explicitly encouraging the development of the software and integrated circuit industries for the first time. Simultaneously, China’s first internationally standard semiconductor wafer foundry, SMIC, broke ground in Shanghai’s Zhangjiang Hi-Tech Park, bringing a glimmer of hope to China’s semiconductor industry.
In the fall of 2000, I set foot on my homeland for the first time in 11 years, beginning an overseas scholar’s journey home and into entrepreneurship. The first question in starting a company: what product to make? For this, I traveled the length and breadth of China — north to south, east to west — conducting market research. Finally, through the introduction of my Tsinghua classmate Ji Weiwei (Radio Engineering Department, Class of '77), I had lunch with Vice Minister Qu Weizhi of the Ministry of Information Industry. During the meal, Minister Qu gave a detailed account of the embarrassing state of China’s mobile phone chips: China, home to the world’s largest number of mobile phone users, had to import every single core handset chip from Europe and America. To break this monopoly, the Ministry of Information Industry in 1997 organized a concerted effort among domestic companies. After several years and billions of yuan spent, all they had to show for it was a pile of project acceptance reports — zero progress on commercialization. She said: “There’s nothing we can do about 2G anymore. If 3G turns out the same way, we really can’t answer to the nation!” I knew absolutely nothing about 3G, but I’d just heard Qiao Peng (Tsinghua Radio Engineering Department, Class of '81 master’s program; co-founder of UTStarcom Wireless and Amimon) mention it. So I picked up the phone and called Silicon Valley, and the very next day, Qiao Peng rushed to Beijing in a whirlwind. Just like that, our product direction was set: develop 3G mobile phone core chipsets (essentially the CPU of a phone). In the mobile communications tidal wave, the market scale would be many times that of PCs.
We returned to Silicon Valley with high hopes and immediately started recruiting. The first person I sought out was Wu Ping — Tsinghua Radio Engineering Department, Microelectronics major, Class of '79. At the time, he was R&D Director at Silicon Valley’s MobileLink, in charge of developing 2G handset core chips. The pleasant surprise was that Wu Ping not only wanted to start a business too, but already had a preliminary team that had been working at it for a while, having weathered several storms and currently wavering in uncertainty. Our arrival was like a shot of adrenaline — two groups merged into one, creating a super-luxurious team that looked beautiful on paper. And so, off we went, full steam ahead.
The next step: finding money. Everything in this world falls into two categories: spending money and making money. Everything that involves spending money is a painful pleasure, but when it comes to making money (fundraising, soliciting donations, product sales…) — that’s when things get truly, brutally difficult. That’s the real test of one’s mettle! Our super-luxurious team of Silicon Valley overseas scholars was no exception.
It started beautifully. In mid-December, our very first meeting with Acorn Campus (an incubator founded by successful Chinese entrepreneurs in Silicon Valley) resulted in a commitment of several million dollars. They even proactively proposed bringing in Lip-Bu Tan — sorry, I mean Chen Wufu (legendary Chinese entrepreneur who had already successfully founded five high-tech companies) as interim CEO. The next day, we happily flew back to Beijing to attend a special symposium on China’s 3G industry development organized by the Ministry of Information Industry. But ten-odd days later, when we returned to Silicon Valley basking in glory, the winds suddenly shifted. Out of nowhere, another team (mostly composed of Taiwanese students) showed up to compete for the same project. Acorn suggested the two teams merge and receive joint funding. Both sides met several times, but our philosophies were just too different to reconcile. And so, that opportunity was lost.
Soon after, the Enron scandal broke, the U.S. telecom bubble burst, and Silicon Valley’s economy plunged into deep recession, with icy winds howling. Our fundraising hit a wall, and the team started to crack. By mid-March 2001, we’d even agreed that if there was no breakthrough within a month, we’d have to abandon the project and find another path.
But crisis breeds opportunity. Not long after, Wu Ping’s trip to Taiwan brought good news: MediaTek Chairman Tsai Ming-kai was willing to invest, finally solving our funding problem. (Interestingly, a few years later, MediaTek and SpreadTrum would jointly create China’s “shanzhai” phone market and become each other’s biggest competitors in that space.) And so, SpreadTrum Communications was officially founded in April 2001. After the team went through several rounds of comings and goings, the final founders were: Wu Ping, Ji Jin (Tsinghua Radio Engineering Department, Class of '77), Fan Renyong (Nanjing University, Class of '78), Zhang Xiang (Zhejiang University, resigned in 2004), and myself.
III. SpreadTrum’s Rise: From 2.5G to 3G
When SpreadTrum Communications was founded in early April 2001, the five founders divided responsibilities as follows: Wu Ping as CEO, myself as CTO, Zhang Xiang as China General Manager (resigned in 2004), and Fan Renyong and Ji Jin as Vice Presidents. At the time, SpreadTrum was up against European and American super-heavyweights: Texas Instruments (TI), Motorola, Siemens, Philips, and more. Mobile phone core chipsets are among the most complex integrated circuits: they demand tens of millions of gates of ultra-high integration while requiring ultra-low power consumption for long standby times. Even harder, to maintain uninterrupted, stable calls, the software algorithms must handle real-time handoffs between cells under endlessly varying, complex conditions — the staggering development difficulty and massive testing workload speak for themselves! At international companies, developing a new generation of handset chips typically required 1,000–2,000 hardware and software engineers working together for 5–7 years, spending $500 million to $1 billion before the product was ready for market. SpreadTrum’s first round of funding was just $6 million. The challenge was severe.
Back then, very few people in China had chip design experience, and high-end talent was virtually nonexistent. So we assembled a chip design team of twenty-some people in Silicon Valley (which was later moved back to China), while simultaneously hiring fifty to sixty engineers in Shanghai to form the software team. Tsinghua Radio Engineering alumni Lu Bin, Xie Fei, Kang Yi, Zhao Tong, and others returned to China one after another, mentoring and training hands-on. Within a few years, domestic staff made leaps and bounds, taking on virtually all software development work.
Silicon Valley is the world’s innovation center — not just for its globally leading technology, but more importantly for the innovation ecosystem it has built over decades, including the investment environment, entrepreneurial spirit, and corporate management systems. When we brought Silicon Valley experience back to China and combined it with local advantages, SpreadTrum created an industry miracle: from founding, 6 months to complete the 2.5G handset chip design; within 10 months, chip verification was essentially done; within 12 months, initial software integration was complete and we made our first call; after another year of field testing and various certifications, chips went into mass production at the 24-month mark!
In 2002, when SpreadTrum needed its second round of funding, it was right in the aftermath of the dot-com bust and 9/11 — Silicon Valley was in a deep slump, and fundraising was harder than ever. The company was watching its cash burn down to fumes. At that critical moment, on Wu Ping’s initiative, executives and U.S. employees took steep pay cuts, helping the company survive the crisis. And the fact that we could raise money at all was decisively thanks to SpreadTrum’s rapid development and bold innovation: we creatively integrated three chips (digital, analog, and power management) into one, and that single chip was just one-third the size of a competitor’s digital chip alone!
By 2003, we had chip products, but who would buy them? Nokia, Ericsson, Motorola — those international brands? Don’t even think about it! Domestic customers were bolder, willing to try new things, willing to be the “guinea pigs” — but they lacked R&D capability. So SpreadTrum went all-in, doing everything: from chip to software, to PCB and casing design, to certification testing. The customer just needed to tweak the phone’s appearance and change the boot screen, and the product was done — a true “turnkey solution” in every sense! And just like that, overnight, countless branded phone manufacturers sprang up in Shenzhen, giving rise to the now-famous “shanzhai” (knockoff) phone phenomenon! Come to think of it, together with Taiwan’s MediaTek, we could probably be called the co-founders of the “shanzhai model.”
The transformation from a pure tech company to a market-driven company requires gut-wrenching, bone-deep change. The “shanzhai” phone craze started with MP3 music functionality. We figured, since it’s a music phone, the sound quality has to be top-notch — must be dual-channel at 128 kbps. At the time, SpreadTrum’s chip used software MP3, only supporting single-channel at 64 kbps. So we went full speed ahead designing a new chip. But before we were even halfway done, the market was suddenly flooded with countless MP3 phones — all running MediaTek chips, all single-channel at 32 kbps! We could’ve kicked ourselves. Lesson painfully learned. So when MP4 video phones became the trend, we got smart: we hacked together a “quasi-MP4” (actually dynamic JPEG technology) and caught the market window.
Market demand is the driving force behind innovation. From 2004 to 2009 (before the smartphone era), 80–90% of multimedia phone innovations originated from the Shenzhen (Huaqiangbei) “shanzhai” phone market. From MP3/MP4 phones, big-screen/loud-speaker phones, phones with built-in radio/TV, to ridiculous things like cigarette-box phones and Ferrari phones — they were wildly varied and endlessly creative. The most iconic was probably the dual-SIM dual-standby feature, saving countless migrant workers exorbitant roaming charges. This need was brought to our attention by Xia Xin CEO Li Xiaozhong during a customer visit. Our R&D team studied it deeply and creatively proposed a pure software solution, launching a product within six months. It took the market by storm, became a must-have feature, and forced even international brands to follow suit.
And so, with high cost-performance ratios, turnkey solutions, flexible local support, and the “fast fish eats slow fish” approach, MediaTek and SpreadTrum competed fiercely in the domestic market, growing at breakneck speed, and within just three or four years pushed TI, ADI, Freescale (Motorola), and other giant companies out of the Chinese market. The industry had a vivid metaphor for this: Two martial arts masters fighting inside a tent, battling until the sky went dark, neither able to defeat the other. When they stepped outside, they found bodies strewn everywhere — all bystanders struck down by the internal energy radiating from the two masters’ fight.
SpreadTrum’s annual revenue, from 2003 to 2007, grew 2–3x every year, reaching nearly 1 billion RMB, and the company successfully went public on the U.S. NASDAQ. In the years that followed, it went through life-and-death struggles and emerged reborn through blood and fire. By 2013, with 1,400 employees and revenue exceeding 7 billion RMB, it had become China’s largest independent semiconductor design company.
Around 2000, over a dozen startups worldwide were developing 3G handset chips, but only SpreadTrum survived. Why? It wasn’t that we were especially clever — we just did two things differently. First, Wu Ping suggested we “hang a sheep’s head but sell dog meat”: when raising money, we said we’d build 3G; once we had the money, we built 2.5G for the existing market first. At the time, the industry predicted the 3G market would take off in 2002–2003, but in reality, 3G didn’t start gaining traction until after 2007. Most companies didn’t survive to see that day! Second, we were fortunate to return to China, where we discovered the “shanzhai” market and took the “surround the cities from the countryside” Jinggang Mountain path. To have been born at the right time, to have witnessed and participated in China’s great rise — what incredible fortune!
IV. TD-SCDMA: China’s First International Standard in a Century
At the turn of the century, the world was undergoing dramatic transformation. And the single most significant change was the emergence of a new market rivaling Europe and North America in scale — led by China, with India, Russia, Brazil, and Southeast Asia following close behind. This would fundamentally reshape the global industrial landscape in the decades to come!
Those who control the market set the rules of the game — it’s only natural. So for the past century, virtually all industrial standards were dominated by European and American corporations. Standards were originally meant to unify specifications and ensure product compatibility between manufacturers, promoting industry development. But after 1990, this original purpose was thoroughly twisted. Controlling industrial standards became one of the primary weapons for Western mega-corporations to monopolize technology and markets and crush competitors. The battle over standards, therefore, concerns China’s overall ability to innovate independently and upgrade its industries — it affects the long-term survival of Chinese industry! But rights cannot be given as charity; they must be fought for. The arduous growth of TD-SCDMA — China’s first international standard in a century, a third-generation mobile communication (3G) standard — is a microcosm of this struggle.
In 2000, China’s Telecommunications Research Institute submitted a 3G standard proposal, TD-SCDMA, to the International Telecommunication Union just before the deadline. Unexpectedly, it sailed through and was quickly approved, becoming one of three international 3G standards (the other two being CDMA2000, the North American standard, and WCDMA, the European standard, currently used by China Telecom and China Unicom respectively). But then, something strange happened: for years, not a single major Western company developed TD handset chips. Only then did the realization dawn — letting your standard pass was giving the Chinese government face (after all, they’d already been raking in fortunes in the Chinese market). Without handset chips, the TD standard was just a worthless piece of paper!
The government panicked and went begging far and wide, finally managing to set up two Sino-foreign joint ventures in 2002: Shanghai Commit (which went bankrupt in 2008) and Beijing T3G (later acquired by Philips Semiconductor, now virtually invisible in the TD market). Companies like Philips and Nokia were just tossing in a few million dollars to go through the motions, keeping the government happy while taking their sweet time: chip design alone would take three years, with the first test tape-out not happening until 2005. At that rate — to borrow a Chinese saying — the flowers would have wilted long ago. TD would be dead for sure!
At this critical juncture, someone finally couldn’t stand watching anymore. Grabbing a spear and leaping onto a horse, he let out a great shout: “Here I am!” The assembled bigwigs’ hearts skipped a beat. They looked up — and couldn’t help but laugh. A nobody, charging in solo: SpreadTrum Communications was going to take on the Western giants single-handedly! In early 2003, SpreadTrum had only about two hundred people. Our 2G chip development was nearly done but not yet in mass production, and the company was in its most difficult phase. Wu Ping and I thought long and hard through the night, debating back and forth. In the end, we remembered why we’d come back to China in the first place. If we couldn’t stand up and fight at this critical moment, we’d live with a guilty conscience for the rest of our lives! And so, we gritted our teeth and made the tough decision, convincing the board to halt the half-finished WCDMA project and go all-in on TD-SCDMA chips.
To understand the TD standard, SpreadTrum decided to collaborate closely with Datang Mobile. This required an in-depth technical discussion between both teams. Unfortunately, it was May 2003 — the height of the SARS epidemic. Datang’s team was in Beijing, SpreadTrum was in Shanghai, staring at each other across the distance, unable to meet, desperate with anxiety. By June, we couldn’t wait any longer. Both teams agreed to meet at the Sun Island Resort in Shanghai for about ten days of intensive discussions. (At the time, the entire resort was deserted — it was as if we’d booked the whole place.) After the meeting, SpreadTrum proposed a bold timeline: chip design complete in six months, first phone call within a year!
That was like poking a hornet’s nest. The industry was abuzz with skepticism; nobody believed it. The Ministry of Information Industry immediately dispatched ten of China’s top experts, with just one question: “Why does everyone else need three years for design while you claim six months?” We talked until our mouths went dry, citing our rich experience, advanced CAD tools, Silicon Valley work-till-you-drop spirit, and track record of success… The experts remained half-convinced at best. Soon, word spread through the industry: yet another “returning overseas scholar” scam company. We swallowed our frustration, buried our heads in work, completed the design in February 2004 and sent it for tape-out. Samples came back at the end of April. By the end of May, we’d made our first physical-layer phone call. Only then did the industry start to believe in SpreadTrum. Zhang Qi, then Director of the Products Division at the Ministry of Information Industry (and a Tsinghua Radio Engineering Department alumna), immediately summoned us for a briefing upon hearing the news. Breaking with standard protocol, she arranged project funding for SpreadTrum, relieving the company’s most urgent financial pressure.
But this was only the first step in TD’s commercialization. Afterward, the Western heavyweights began maneuvering aggressively, spreading narratives like “TD cannot build a standalone network; it can only supplement other 3G networks.” Under various pressures, over the ensuing years, TD went through indoor testing, outdoor testing, small-scale networking, large-scale testing — with countless obstacles and dangers at every step, truly nerve-wracking!
The most perilous moment came in the spring of 2006. To push back against the ITU, the Ministry of Information Industry required TD to upgrade from “inter-frequency networking” to “same-frequency networking.” This dramatically increased the difficulty of interference mitigation between adjacent cells. The number of neighboring cells to be monitored in real-time jumped from 4 to 7 or more, increasing the chip’s decoding computational demand by 2.5–3x. To make matters worse, we had to pass the November field test in Qingdao, with absolutely no time to develop a new chip. The only option was to optimize software algorithms and squeeze out computing time by brute force. It was nearly mission impossible! I had no choice but to take charge personally, brainstorming with the algorithm team — every wild idea, every off-the-wall thought, thrown against the wall — until finally, by the end of April, we identified a technical path with a sliver of hope. Then it was overtime, around the clock, racing against time. We barely made it to the initial field test in Qingdao in September. But when the results came in, they were predictably dismal: the call success rate during road testing was only 28%, against a standard of 95% or higher. Two months left. I led a strike team of dozens, stationed in Qingdao. Every night — all-night road testing. Mornings — rest. Noon — meeting to review, analyze problems, propose solutions. Afternoon — coordinate with headquarters, modify software at every layer, deliver new versions. Evening — back to road testing. We pushed ourselves to the absolute limit for two solid months. Finally, in the ultimate test, we passed — average call success rate of 97–98%. Another miracle!
And so, SpreadTrum worked hand in hand with chip vendors like Datang, T3G, and Chongqing University of Posts and Telecommunications (CQUPT) Xinke, base station manufacturers like Huawei, ZTE, and Xincom, and handset makers like Hisense, Lenovo, and Bird Mobile. Together, encouraging one another, enduring untold hardships, we pushed forward step by step. Finally, in early 2009, the TD operating license was awarded to China Mobile, and the commercialization process began.
Over the three years since, the TD industry has continued to develop and improve. Sales of TD handsets are projected at 80 million units per year, accounting for over 40% of the domestic 3G market. China’s first international standard in a century has finally been successfully commercialized!
TD’s development has also produced two additional effects: First, the entire TD supply chain (base stations, chips, and handsets) is completely dominated by domestic companies — Huawei, ZTE, SpreadTrum, Lianxin, Hisense, Lenovo — enormously boosting the upgrade of the mobile communications industry. Second, TD’s successful large-scale commercialization has greatly strengthened China’s voice in setting 4G standards: China’s 4G standard, TD-LTE, has been accepted as one of two international 4G standards, poised to capture 30–40% of the global market!
Looking back on the ten-year arduous journey of TD, I deeply feel that everything can only rely on ourselves. I believe that as long as we stay true to our convictions, build our confidence, and keep our heads down and work hard, the tide of our independent innovation and industrial rise is unstoppable!
