Coreay Technology, a manufacturer of packaging substrates for AI chips, is being eagerly sought for investment by OPPO, BYD, Shenghong and other investors.
Domestic Xinai Technology has obtained financing from multiple industrial capitals to focus on the R&D and production of high-end chip packaging substrates.
After an AI chip is manufactured, it needs to be mounted on a "base" several times larger than the chip to be powered on, dissipate heat, and communicate with the circuit board. This base is called a packaging substrate, which serves as the bridge between the chip and the external world. The number of manufacturers capable of producing high-end substrates across the world can be counted on one hand, and they are all located in Japan, South Korea and China's Taiwan region.
On October 19, a Nanjing-based company named Xinai Technology announced the completion of its nearly 1.5 billion yuan Series A2 financing. The list of investors is quite notable: OPPO that manufactures mobile phones, BYD that produces automobiles, Rockchip that develops end-side chips, and Shenghong Technology that makes circuit boards all participated in the financing, followed by state-owned capitals at both provincial and municipal levels in Jiangsu Province and a number of existing shareholders making additional investments.
These industrial parties are scrambling to take shares in a company that was founded just five years ago, as they are not buying the current profits, but the high-end substrate production capacity in the future.
01
The Base of Chips
Over 80% of the Market Is Controlled by Japan, South Korea and Taiwan Region of China
First, let's figure out which link the substrate industry is stuck at.
After the chip is manufactured, it cannot be directly soldered to the circuit board due to excessively dense pins, excessive heat generation and unsmooth signal transmission. A layer of "substrate" must be placed in the middle: lead out the circuits of the chip, rearrange them, and then connect them to the circuit board. The more advanced and larger the chip is, the more layers the substrate has, the finer the circuits are, and the higher the technical difficulty will be. AI chips usually have a size of over a thousand square millimeters, which is several times larger than mobile phone chips, so they put much stricter requirements on the substrate.
The global market of this type of board is highly monopolized. According to industry research data, the world's top 10 IC substrate manufacturers all come from Japan, South Korea and China's Taiwan region, which together account for about 85% of the global output value in 2025; the top-ranked Unimicron alone accounts for about 20%. The overall share of Chinese mainland manufacturers is very small.
If we dig deeper into the upstream, the supply constraint is even more severe. More than 95% of the global share of ABF build-up film, the core material for manufacturing high-end substrates, is held by a Japanese company — Ajinomoto. This enterprise, which is well-known for its monosodium glutamate, has controlled most of the upstream throat of AI chip packaging. According to industry data, ABF materials only account for 5% to 7% of the substrate cost, but once there is a material shortage, the entire high-end substrate production line will have to shut down.
The supply side is also not sufficient. At the SEMICON industry conference in September this year, high-end substrates were named the second supply chain bottleneck after memory: the supply gap of ABF substrates exceeds 30%, and the orders of leading manufacturers have been scheduled to the second quarter of 2027.
NVIDIA's GPUs need substrates, and domestic AI chips also need substrates. Whoever owns the production capacity controls the neck of the entire computing power chain. This is exactly the background why Xinai Technology's this round of financing is scrambled for by industrial parties.
02
The 8th Early Employee of MediaTek
Comes Back to Make Up for Lessons
Zhang Chuihong, the founder of Xinai Technology, has a resume that can be regarded as the "ceiling" level in the semiconductor circle.
He is a founding employee of MediaTek, one of the first eight employees of the company, and once served as Senior Vice President. He has been in charge of core businesses including supply chain management, wafer process development, CPU and EDA; he and his team are the main inventors of Coreless ETS technology, the mainstream substrate technology for mobile phones. He holds an MBA degree from Tsinghua University and a Master's degree in Electrical and Computer Engineering from New York University.
In May 2021, Zhang Chuihong founded Xinai Technology in Pukou, Nanjing, and built the entire advanced substrate manufacturing system from scratch. The plant construction speed was very fast. In July 2023, the company began to produce samples, and within 7 months after the samples were delivered, its products passed the certification of leading domestic packaging and testing manufacturers such as JCET and Hua Tian Technology.
A person who has worked in the semiconductor substrate industry in Taiwan for most of his life, why did he come to Nanjing to start all over again? Because he is developing FCBGA packaging substrates, which are applied to the packaging of AI and server core chips such as CPU, GPU, FPGA and network ASIC, and this high-end category is exactly what the Chinese mainland lacks the most and relies most on imports.
The capital market has given sufficient corresponding support to Xinai Technology which has such "ambition".
Over the past four years, Xinai Technology has accumulated financing of more than 5 billion yuan. In September this year, Nanjing and Xinai Technology signed a new 5 billion yuan project — the advanced packaging substrate project for AI and chiplet architecture, which is planned to build a production line with an annual output of more than 700,000 high-end substrates, and will achieve an annual output value of 6 billion yuan after reaching full production.
Interestingly, just a few days before the announcement of this round of financing, Zhang Chuihong publicly stated in an industry dialogue: be alert to the "glass substrate craze". Glass substrate is regarded as the next-generation alternative direction of ABF substrate, and it is also one of the R&D directions that Xinai will invest in with the funds raised in this round — the money will be invested in that direction, but the founder first poured a basin of cold water on the overheated market sentiment.
03
Mobile Phone Manufacturers, Automobile Manufacturers and Chip Manufacturers
Are Scrambling to Become Shareholders
The most noteworthy part of this round of financing is not the 1.5 billion yuan amount, but the investors.
There are two new industrial investors: Shenghong Technology and Rockchip. According to the report by Xinhua Daily Press Group, Shenghong Technology provides order and process coordination from the demand side of PCB and substrates, while Rockchip carries out technical linkage from the chip definition end. The two companies respectively occupy the downstream application and upstream design links of the packaging substrate industrial chain. Earlier, OPPO and BYD had already appeared in Xinai's shareholder list.
Mobile phone manufacturers, automobile manufacturers, chip manufacturers and circuit board manufacturers, the downstream players from all four links have joined in. This is not a typical financial investment logic. The most direct demand for industrial capitals to take shares in a substrate company is to "lock production capacity": when AI chips are to be mass produced, high-end substrates are indispensable consumables. Instead of waiting for market allocation, it is better to tie the production capacity to their own partners first.
Xinai Technology itself also confirmed this path. The company said that its relevant products have passed the verification of leading domestic AI chip enterprises. This round of capacity expansion is based on the existing customer verification, and will synchronously improve the FCBGA production capacity, yield and delivery capacity along with the mass production of customers' products.
In other words, what Xinai is betting on is not "whether I can make the product", but "whether I can supply enough when domestic AI chips are mass produced". The investors are betting on the same thing.
From "passing certification" to "large-scale supply", there are three barriers in the middle: yield, production capacity and stable delivery. This is also the reason why the state-owned capitals at both provincial and municipal levels in Jiangsu are willing to enter the market as patient capital, because this track features long cycle and heavy assets, and requires investors who can afford to wait.
However, the final performance of Xinai Technology will still be judged by two figures.
The first is the commissioning schedule of the new Nanjing project at the end of 2026, whether the production line with an annual output of 700,000 high-end substrates can be put into operation as scheduled; the second is the actual loading rate of domestic substrates in domestic AI chips, and how long the distance from verification to mass production is. 85% of the global substrate production capacity is controlled by others. Xinai has torn half of this layer of paper, but the second half of the journey is more difficult than the first half.