Hefei is going to take a slice of Wuhan's market pie.

Changxin Technology, which immediately topped the A-share market capitalization ranking right after its listing, has released another major piece of news. 

On September 20, Changxin Technology announced at the World Manufacturing Convention held in Hefei, Anhui Province: Its fifth-generation DRAM (memory) technology platform has achieved mass production. Changxin stated that the number of dies produced per wafer on the G5 platform has increased by at least 50% compared with the previous generation platform, with improvements in storage density, power consumption, and cost control.

Changxin has shrunk the half-pitch of the active area of the memory array to 11.95 nanometers through quadruple patterning technology, which is close to the process of the world's top mass-produced memory platform. At the same time, Changxin also released two 24 Gigabyte (GB) LPDDR5X products built with the new technology platform. LPDDR5X is an energy-efficient DRAM mainly used in smartphones and other portable electronic devices.

Changxin said that compared with the previous generation of products, the data capacity of new products has increased by more than 50%, and they have entered formal mass production, fully entering the mainstream domestic flagship mobile phones in China. This means that the DRAM performance produced by Changxin Technology has caught up with the world-class level, and there is no generation gap anymore. Luo Xiaodong, Vice President of Changxin Technology and Head of the Marketing Center, said at the conference: Our process capability is now equivalent to the most advanced mass production node in the industry.

Reuters previously revealed that Changxin Technology is building new production lines in Beijing and Shanghai. 

In Beijing, Changxin Technology has long built a factory in Yizhuang, which is responsible for R&D and small-scale mass production. Now it plans to build another factory to expand the production line. In Shanghai, Changxin Technology has chosen to build a factory in the Lingang New Area of Shanghai. 

At present, Changxin Technology's two wafer production lines in Hefei and Beijing have a total capacity of more than 300,000 wafers per month. In addition, Changxin is building a total of 4 new wafer fabs in Shanghai and Hefei. It is expected that by the end of 2028, the total capacity will be increased to 464,000 wafers per month. 

Hefei, is going to share the market cake with Wuhan

The most critical point is that Reuters revealed: Changxin Technology is preparing to enter the NAND flash memory chip market dominated by South Korean companies such as Samsung Electronics and SK Hynix to expand its customer base. NAND flash memory is widely used in storage for mobile phones, computers and data centers.

Its NAND flash memory R&D production line is planned to be located in the new factory in Beijing. 

Changxin is the domestic leader in DRAM (memory), and the leader in NAND (flash memory) is Yangtze Memory Technologies Co., Ltd. (YMTC), which is preparing for its listing. In addition, Changxin Memory is the only domestic enterprise that has achieved large-scale mass production of memory chips. YMTC is the only memory IDM enterprise in mainland China that has the complete industrial chain capability of 3D NAND flash memory chips.

In terms of global market share: 

Memory: By the end of 2025, the three major original manufacturers Samsung, SK Hynix and Micron together accounted for more than 90% of the market share, and Changxin Memory only accounted for 7.67% of the global market. In the first quarter of 2026, Changxin's share rose to 8%, and in the second quarter it rose to 10%. 

Flash memory: The five major manufacturers in the NAND Flash market, Samsung, SK Hynix, Kioxia, Western Digital and Micron, together accounted for nearly 90% of the market share by the end of 2025, of which the top three accounted for about 65%. YMTC accounted for 11%. In the second quarter of this year, YMTC's share has risen to 14%, surpassing Micron and Kioxia, second only to Samsung and SK Hynix. 

Now that Changxin wants to enter the NAND market, it is obvious that this is not only to seize the market with South Korean Samsung, SK Hynix and Micron, but also inevitable to share the market cake with Wuhan. 

But it is not as simple as imagined: First, after all, the main products of Changxin and YMTC are in two different tracks, and there was no competitive relationship before. Second, in the field of flash memory, YMTC's technology is already very mature, ranking third in global market share, and its share is still rising.

Driven by the global AI wave, the chip industry has entered a super cycle in the past two years. As the two leading enterprises of domestic storage chips, Changxin Memory and YMTC have also ushered in a harvest period. 

Changxin's half-year revenue surged 8.7 times, and its net profit jumped directly from a loss of 2.3 billion yuan to 77.6 billion yuan. What is the concept of 77.6 billion yuan in net profit? It is more than three times that of Huawei. 

The half-year financial report released by Huawei shows that the attributable net profit is 23.428 billion yuan. For YMTC, its IPO documents show that in the first quarter, YMTC's revenue was 47.042 billion yuan, and its attributable net profit was 33.379 billion yuan, which is more profitable than Changxin. Now there are two real difficulties for Changxin to enter the NAND field.

The first one is the technical barrier. Although DRAM and NAND both belong to storage chips, their underlying technologies, product structures and manufacturing processes are different. Changxin has been betting on DRAM for the past ten years, advancing from the first-generation process to the fifth-generation, and has established a complete R&D, manufacturing and mass production system. Now it has become the number one DRAM manufacturer in China and the fourth largest in the world.

However, this set of capabilities cannot be simply copied to NAND. The core of NAND is 3D stacking, which needs to balance the number of storage layers, storage density, yield, performance and cost. YMTC has focused its main energy on 3D NAND since its establishment. After years of R&D and capacity expansion, it has formed a complete NAND industrial chain.

Changxin's transition from DRAM to NAND is equivalent to fighting a tough battle all over again.

The second, and more realistic one, is that the market share has already been occupied. After years of competition in the NAND market, giants such as Samsung, SK Hynix, Kioxia, Western Digital and Micron have long formed a stable supply system. The same is true in the domestic market, where YMTC has taken a step forward.

Moreover, YMTC is now growing at a high speed. In the first quarter of 2026, YMTC's global NAND market share reached 13%; in the second quarter, as the demand for enterprise-level SSDs from AI servers continued to explode, its shipment share further increased to 14%, taking the third place in the world. YMTC has already taken a large piece of the cake, and its own share is still expanding.

It took Changxin ten years to start from scratch in the memory track and reach a current market share of 10%. Now it wants to lay out the flash memory track, and its only advantage is that it has a large amount of capital in hand compared with the early days. But this is a completely different war. However, for Changxin, for Hefei, and for China, Changxin's layout of the flash memory track is of great significance.

For Changxin, once it makes a breakthrough in NAND, it will change from "walking on one leg with DRAM" to two-wheel drive of DRAM + NAND. After the two major storage tracks are connected, Changxin's product line, customer group and market space will be significantly expanded, and its dependence on a single product cycle can also be reduced.

More importantly, Changxin will truly have the confidence to launch comprehensive competition against global storage giants such as Samsung and SK Hynix. For Hefei, the significance lies in that it has another card in hand that can influence the national and even global industrial pattern.

If Changxin expands from DRAM to NAND, Hefei's storage industry will move from a single track to two tracks, and the industrial chain, talents, capital and upstream and downstream enterprises will further gather.

In the past, first-tier cities such as Beijing and Shanghai imported industries to Hefei. Now Hefei, relying on industry leaders like Changxin, is starting to attract R&D, capital and industrial resources to its surrounding areas in reverse.

For China, this battle is of even greater significance.

DRAM and NAND are the two core tracks of the storage industry. If Changxin can take over NAND, it means that China is expected to form stronger independent supply capabilities in both tracks at the same time.

Especially under the background of AI driving the surge in global storage demand and the intensification of international supply chain competition, having one more domestic storage giant that can enter the global mainstream market means that China has added another layer of industrial chain resilience in this key industry.

For Wuhan, on the surface it is competitive pressure, but from another perspective, it is the strongest driving force to promote YMTC's technology iteration. This means that YMTC will not only face the market competition pressure from South Korea and the United States in the future, but also face the challenge from Changxin Memory.

To cope with domestic and international challenges, YMTC needs to accelerate to catch up with the world's top technical level on the one hand, and build its own technical moat in China on the other hand.

At present, the total capacity of YMTC's first and second phases, which is 160,000 wafers per month, has reached full production. The third phase started construction in September 2025 and will be put into operation at the end of 2026, with a planned capacity of 100,000 wafers per month. In terms of technology, Samsung and SK Hynix began mass production of 300-layer NAND in 2025, and YMTC mass-produced 270-layer NAND in the same period. By 2027, YMTC will jump to 400 layers, narrowing the gap with Samsung and SK Hynix to about 1 year.

Behind Hefei's success

Hefei is the provincial capital with the fastest rising speed in the past two decades. In the next five years, Hefei will most likely surpass Zhengzhou and Changsha, and become the second largest city in central China.

Of course, Wuhan's position as the number one city in central China is very stable. After all, Hefei is making progress, and Wuhan is also making progress. Wuhan has a large base and a large increment, and Wuhan is still expanding its advantages.

However, Hefei Changxin's move to share the market cake with Wuhan YMTC, whether it succeeds or not, is enough to show that Hefei has become a pivotal heavyweight player in China's urban competition.

Even Hong Kong is starting to learn from Hefei. A latest opinion commentary article published by Singapore's Lianhe Zaobao, titled "Dai Qingcheng: Does Hong Kong need to learn from the 'Hefei Model'?" clearly mentions that the Hong Kong delegation of the National People's Congress also went to Anhui for inspection to learn about the construction of Hefei's science and innovation industry on the spot. After returning to Hong Kong, the representatives shared their inspection insights through the media one after another, praising Hefei's industrial development model, which has extremely strong reference significance for Hong Kong's current economic transformation.

Seeing this, one has to think about a question: What is behind Hefei's success? Personally, I think it is the comprehensive result of luck, effort, vision and concentration. Just like a person, behind the success, these four things are indispensable.

In addition to the above four points, in the tactical execution of industrial development, there are four key factors that determine success or failure:

First, adhere to long-termism, and do not easily change the selected track. After putting forward the strategy of "building the city through industry" in 2005, Hefei has hardly shaken its direction of developing advanced manufacturing. From the initial new display, to the later integrated circuits and new energy vehicles, to today's artificial intelligence, quantum technology, commercial aerospace and controllable nuclear fusion, Hefei has always focused on hard technology layout instead of chasing short-term trends.

It is particularly noteworthy that the new display, semiconductors, integrated circuits and new energy vehicles named in the 2013 government work report are still the pillar industries of Hefei more than ten years later. The industrial policy does not change every year, and the financial investment does not stop because of the industry trough. This strategic concentration is more important than any investment promotion slogan.

Second, build a complete industrial chain. In 2008, when almost everyone was questioning, Hefei took out funds equivalent to one-third of its fiscal revenue that year to support BOE to build its 6th-generation production line. Many people saw the 6 billion yuan investment, but what really matters is the story behind it.

As BOE grew into a global display leader, more than 70 upstream and downstream enterprises settled in Hefei one after another. One project eventually brought out a world-class display industrial cluster, which also made Hefei stand at the high ground of the global new display industry. Later, Changxin Memory drove the semiconductor industrial chain, and NIO drove the new energy vehicle industrial chain. The three industrial chains continued to intertwine and integrate, forming the widely known "Chip, Screen, Automobile, and Core Integration" ecosystem today.

Hefei's development path is not "introduce one enterprise", but "cultivate an ecosystem".

Third, be able to stand loneliness. BOE, NIO and Changxin all have one common feature — long investment cycle and extremely high risk. BOE suffered continuous losses at that time; NIO had difficulty in financing, and was even on the verge of life and death for a time; the storage chip industry where Changxin is located requires huge investment and has a long return cycle. Many local governments did not dare to invest, but Hefei did.

This courage is not gambling, but long-term investment based on professional judgment.

From investment promotion through funds, to industrial funds, and then to patient capital, Hefei has continuously improved its investment mechanism, adhered to the principle of "investing in early stages, investing in small enterprises, investing for the long term, and investing in hard technology", and is willing to accompany enterprises through the most difficult few years, instead of requiring results in one year and profits in two years. Facts have proved that few real large industries can grow in a short cycle.

Fourth, let professionals do professional work, and let the market play a decisive role. One of the most easily overlooked points of Hefei is that the government does not take full charge of the industry. Whether it is investing in BOE or laying out Changxin, it is not an arbitrary administrative decision, but relies on professional teams to conduct due diligence, expert evaluation, and then the industrial fund operates in a market-oriented manner.

More importantly, the government dares to enter, and also dares to withdraw. After BOE matured, state-owned capital gradually reduced its holdings to realize capital exit; after NIO completed the repurchase, the government's shareholding ratio decreased, and the enterprise was handed back to the market. The government is responsible for providing timely help, not for being the boss all the time; it takes early risks, but does not excessively intervene in business operations. This sense of boundary is precisely the most difficult thing for many places to learn.

Therefore, what Hefei cannot really replicate is not the successful investments one after another, but a complete set of industrial development logic.

It did not chase hot spots, but adhered to long-termism; it did not only recruit leading enterprises, but built an industrial ecosystem; it did not pursue short-term financial returns, but cultivated patient capital; it did not replace the market with administration, but let professionals do professional work.

BOE, Changxin and NIO are just the fruits harvested in different stages. What really supports the growth of these fruits is the development logic that has hardly changed for more than 20 years. This is the key for Hefei to become the provincial capital with the fastest rising speed in the past two decades and achieve leapfrog development.

This article is from the WeChat Official Account "Urban Finance", the author is Yu Fei, and it is published with authorization from 36Kr.