In 2004, Zhu Yiming left his Silicon Valley job, carrying $100,000, to return to China and start a business. His dream was to "create proprietary memory chips." At that time, the global DRAM market was dominated by Samsung, SK Hynix, and Micron, which together controlled over 95% of the market share. Back then, almost no one believed he could succeed.

Undeterred by others' doubts, Zhu Yiming founded GigaDevice in Beijing in 2005, initially focusing on small-capacity NOR Flash memory. In 2016, GigaDevice was listed on the main board of the Shanghai Stock Exchange. That same year, he made a decision everyone thought was insane—entering the DRAM track by founding a new company. This company was the future ChangXin Memory Technologies (CXMT). A decade later, this company went public and shattered the A-share market's ceiling.

The "King of Stocks"

On July 27, 2026, CXMT officially debuted on the STAR Market of the Shanghai Stock Exchange, with the stock code 688825 and an IPO price of 8.66 yuan. In the first second of trading, the stock price jumped directly to 49.5 yuan, a gain of 471.59%, instantly pushing its total market capitalization past 3.31 trillion yuan. To put this in perspective, Industrial and Commercial Bank of China (ICBC) had a market cap of 2.75 trillion yuan at the time, with Kweichow Moutai and CATL trailing behind. CXMT opened as the highest-valued company on the A-share market. This wasn't a gradual climb; it was a full-throttle sprint straight to the top.

During intraday trading, CXMT's stock price once surged to 55.03 yuan, briefly pushing its total market cap to 3.68 trillion yuan, even surpassing Hong Kong's "king of stocks," Tencent. The full-day trading volume reached 141.1 billion yuan, setting a new single-day trading record for an individual A-share stock and making it the first A-share stock in history to exceed 100 billion yuan in single-day turnover. In terms of online subscriptions, 9.4288 million investors participated, setting a new historical record for the STAR Market.

A particular detail highlights the scale of the event. On the listing day, some investors reported trading lags, being unable to place or cancel orders. A securities firm branch manager bluntly stated, "There's never been a new stock with this much trading volume," as even the brokerage systems were overwhelmed. The stock closed at 49 yuan, up 465.82%, with a total market cap of 3.28 trillion yuan. From open to close, CXMT swept through A-share historical records: the largest IPO on the STAR Market, the first A-share tech stock to open with a market cap exceeding 3 trillion yuan, the first individual stock to surpass 100 billion yuan in single-day turnover, and more. For CXMT, this day wasn't just a "listing"; it was a "coronation," firmly planting itself in the position of "king of stocks."

The Winners

With the 3.28 trillion yuan market cap pie sliced up, who got the biggest pieces? The biggest winner was the Hefei state-owned assets system. A look at the equity structure reveals that the Hefei municipal state-owned assets system, through platforms like Qinghui Jidian, ChangXin Jicheng, Chan Tou Yi Hao, and Hefei Jian Chang, holds a combined stake of approximately 36.79%. Among these, Qinghui Jidian is the largest shareholder with a 21.67% stake, followed by ChangXin Jicheng with 11.71%, both ultimately controlled by the Hefei SASAC. Based on the closing market cap, the value of the Hefei state-owned assets' portion exceeds 1.2 trillion yuan. Market estimates suggest that over the past decade, the cumulative investment from Hefei state-owned assets was only between 24.8 billion and 30 billion yuan, yielding an investment return rate of nearly 50 times! This wasn't stock trading; it was a decade-long strategic game of chess.

The second-largest winner was the National Integrated Circuit Industry Investment Fund Phase II (Big Fund II), holding an 8.73% stake, corresponding to a market value of approximately 288.9 billion yuan. Anhui Provincial Investment Group holds a 7.91% stake, valued at about 262 billion yuan. Among industrial capital, Alibaba was the biggest winner. Through two entities, Alibaba holds a combined 4.97% stake, with a cumulative investment of about 7.6 billion yuan. At the closing market cap, this portion of equity is worth approximately 163 billion yuan, resulting in a paper profit exceeding 155 billion yuan. Then there's one person who made a quiet fortune: DeepSeek founder Liang Wenfeng. His controlled entities, High-Flyer Quant and Jiuzhang Asset Management, together secured about 20.2497 million shares, making a paper profit of over 700 million yuan at the open.

One group often overlooked is CXMT's employees. The company implemented two phases of employee stock ownership plans, cumulatively granted to 6,760 person-times. Based on the opening price, at least 237 employees have become paper millionaires. The stock plan from 2023 had a grant price of only 0.108 yuan per share, yielding a post-IPO return of over 400 times. For a single company's listing to mass-produce 237 millionaires is rare in A-share history.

The "Outsider"

But among all these winners, the most unique one is Zhu Yiming. According to the prospectus, Zhu Yiming indirectly holds approximately 1.5924 billion shares of CXMT through platforms like Qinghui Jidian, Hefei Jixin Sishiyi Hao, and GigaDevice. His pre-IPO stake was about 2.65%. Based on the 49 yuan closing price, the book value of this portion is approximately 78 billion yuan. Don't forget he also holds shares in GigaDevice. Zhu Yiming owns about 5.13% of GigaDevice. On that day, GigaDevice's market cap was about 304.6 billion yuan, making his stake worth 15.6 billion yuan. Combined, Zhu Yiming's paper net worth approached 94 billion yuan. Even conservatively, his personal net profit from this single CXMT venture exceeds 40 billion yuan. An engineer who returned from Silicon Valley with $100,000 and was once laughed at now has a net worth approaching 100 billion yuan.

But what's most surprising about Zhu Yiming isn't how much he earned, but how much he gave up. In 2018, he resigned as General Manager of GigaDevice to take on the full-time role of Chairman and CEO of CXMT. Before leaving, he made a solemn pledge: he would not take a single yuan in salary or bonus until CXMT turned a profit. He kept this promise for seven years. From 2018 to 2024, Zhu Yiming drew a "zero salary" for seven straight years. And that's not all. He also publicly promised to distribute his personal holdings of approximately 768 million CXMT shares to current employees within ten years, starting 36 months after the company's listing. Based on the current market cap, these shares are worth over 38 billion yuan. In other words, he directly transferred half his net worth to his employees. Furthermore, he pledged not to reduce his holdings for ten years after listing, meaning he would have to wait at least twenty years for a full exit. A man who worked for seven years without pay, gave away tens of billions in equity to employees, and promised not to sell his shares for a decade—in an era where everyone pursues "cashing out quickly," Zhu Yiming seems like an outsider.

The Big Gamble

Born in 1972 in Yancheng, Jiangsu Province, Zhu Yiming entered the Physics Department of Tsinghua University at age 17, completing a combined bachelor's and master's degree. After graduation, he went to the US for further studies, earning a Master's in Electronic Engineering from Stony Brook University. He then worked in Silicon Valley, becoming a project manager at a memory chip company. During his time in Silicon Valley, he witnessed a fact: memory chips are the most widely consumed and most standardized semiconductor product, the "food" for nearly all electronic devices. Yet, in this field, Chinese players were long absent.

In 2004, he decided to return to China. In 2005, GigaDevice was founded in Beijing, starting with NOR Flash. The path wasn't easy, but Zhu Yiming walked it steadily. In August 2016, GigaDevice was listed on the main board of the Shanghai Stock Exchange. By 2025, GigaDevice's revenue had reached 9.2 billion yuan. If the story ended here, Zhu Yiming would simply be a successful chip entrepreneur. In May 2016, Zhu Yiming reached an agreement with Hefei City to launch the "506 Project," with an initial investment of 18 billion yuan. In June of the same year, CXMT was founded, focusing on DRAM. What is DRAM? It's the RAM in your phone and the memory sticks in your computer. It seems unremarkable, but only three companies in the world can mass-produce it: Samsung, SK Hynix, and Micron. These three monopolize over 90% of the market share. What Zhu Yiming aimed to do was to carve out a position for a Chinese player in the crevice between these three giants. Industry insiders called him "crazy." The entire first phase of the National Big Fund was only 138.7 billion yuan, enough to build just two DRAM production lines. Insufficient funds, lack of technology, shortage of talent—none of the conditions were right. But Zhu Yiming still went ahead.

In 2018, he resigned as General Manager of GigaDevice to fully commit to CXMT. In September 2019, CXMT launched its self-designed 8Gb DDR4 product, achieving a "from zero to one" breakthrough for Mainland China's DRAM industry. In November of the same year, it received its first order. But a breakthrough didn't mean profitability. In the following years, CXMT continued to incur losses, losing 16.3 billion yuan in 2023 alone. External doubts grew louder, and the claim that "China can't make DRAM" became widespread.

The turning point came in 2023. That year, the global memory industry experienced its coldest winter in fifteen years. Samsung, SK Hynix, and Micron were forced to cut production. Yet CXMT made a stunning decision: to expand production counter-cyclically. Despite a 16.3 billion yuan loss, its annual R&D expenditure reached a high of 4.67 billion yuan, and its monthly production capacity increased from 90,000 wafers to 150,000 wafers. This was a massive gamble, betting on what the giants dared not bet on and doing what they dared not do. In 2025, the AI wave surged. The three giants shifted their production capacity towards higher-margin HBM, voluntarily giving up the traditional DRAM market. The capacity CXMT had built counter-cyclically perfectly plugged this structural gap. In 2025, CXMT achieved its first annual profit, with a net profit attributable to the parent company of 1.875 billion yuan. In the first quarter of 2026, single-quarter net profit surged to 24.762 billion yuan. The company expects a net profit of 50 to 57 billion yuan for the first half of 2026, averaging nearly 300 million yuan per day. From losing 16.3 billion yuan a year to earning 300 million yuan a day—only a little over two years separated these two points. Today, based on its Q1 2026 revenue, CXMT's global market share is about 8%, ranking fourth globally. There are only four DRAM manufacturers globally with complete IDM capabilities: Samsung, SK Hynix, Micron, and CXMT. In twenty years, Zhu Yiming has gone from being an engineer in Silicon Valley to becoming the only Chinese player at the global memory table.

The Trend

Zhu Yiming is incredibly capable. But CXMT couldn't have gotten this far on his strength alone. CXMT's story is essentially a microcosm of the era's theme of import substitution. For decades, China was in a state of "following from behind" in the chip field. As the most consumed type of chip, China imported nearly $60 billion worth of DRAM annually, with a self-sufficiency rate close to zero. Unless this situation changed, Chinese manufacturing would forever be constrained by others. In 2016, Hefei launched the "506 Project." The same year, the National Big Fund began deploying in memory chips. A capital formation covering multiple levels from local to national—including local state-owned capital, national industry funds, industrial capital, and market-oriented venture capital—began a decade-long journey of accompanying CXMT.

Over the decade, CXMT accumulated cumulative losses exceeding 30 billion yuan. Any purely market-oriented capital would have pulled out long ago. But the Hefei state-owned capital held on, and the National Big Fund held on. Not because they were smarter than the market, but because they were calculating not a 3-5 year payback period, but the nation's strategic account. This strategic determination was most evident in 2023. That year, the global memory industry was in a deep freeze. CXMT lost 16.3 billion yuan, but the Hefei state-owned capital and Big Fund didn't withdraw; instead, they continued to support capacity expansion. If they had backed down then, CXMT wouldn't have been able to catch the market opportunities brought by the 2025 AI wave. So, CXMT's success today stems from the confluence of four indispensable forces: Zhu Yiming's personal capability, a decade of Hefei state-owned capital's unwavering support, the strategic backing of the National Big Fund, and the historic opportunity of the AI era.

Today, China's self-sufficiency rate for memory chips has risen from less than 5% a few years ago to about 25% in 2025. From "almost zero" to "global fourth," CXMT took a decade. From "being strangled" to "making it ourselves," China's memory chip industry took a decade. Import substitution is not just a slogan; it's billions of yuan invested, countless nights of toil, and generations of technology caught up with.

Epilogue

On the day of CXMT's listing, many people only saw the stock price and market cap. But if you look deeper, a bigger change is underway. Hefei state-owned capital invested 24.8 billion yuan over a decade, yielding a paper return of over a trillion yuan. This wasn't generated by selling land; it came from investing in technology. For the past twenty years, local government finance mainly relied on land sales. A plot of land was auctioned, money came in, but the land was gone. This was a consumptive fiscal model, selling non-renewable resources. But now, things are different. Local governments are becoming "shareholders." They invest money in hard-tech enterprises. The companies grow, list, and their market caps soar. The value of the government's equity holdings rises accordingly. This is a creative fiscal model, investing in assets that continuously appreciate. Hefei wasn't the first to do this, but it has done it most successfully. From BOE to NIO to CXMT, Hefei has proven with one "gamble" after another that the second half of city competition is not about who sells more land, but who can "generate" more good companies.

This author believes that CXMT's listing signifies the end of the land finance era and the rise of technology finance and equity finance as new forms of wealth creation. It marks the end of an era driven by selling land, building buildings, and relying on real estate to stimulate the economy. It also marks the beginning of an era where driving development through investing in technology, fostering innovation, and leveraging equity appreciation is pulling back the curtain. CXMT, with its 3.28 trillion yuan market cap, is not just Zhu Yiming's victory, nor just the victory of one city, Hefei. It is a landmark footnote in China's great transformation from a "world factory" to a "technology powerhouse." Ten years ago, no one believed China could make its own DRAM. Ten years later, CXMT not only made it but also achieved fourth place globally with a 3 trillion yuan market cap. In the next ten years, there will be more CXMTs emerging.