China's Largest Unlisted Independent Memory Manufacturer: Hongxinyu's Path to Self-Developed Controllers

Title context: China's Largest Unlisted Independent Memory Manufacturer: Hongxinyu's Path to Self-Developed Controllers Text: As China's largest unlisted independent memory manufacturer, what is truly worth watching about Hongxinyu may not just be its scale, but also the chip capabilities hidden behind its products. , publicly filed application documents show that Hongxinyu achieved full-year operating revenue of RMB 11.238 billion and net profit of RMB 1.364 billion in 2025; in January-April 2026, performance accelerated, with operating revenue of RMB 8.012 billion and net profit of RMB 3.841 billion, with profit for the single period already exceeding the full-year level of 2025. When the market turns its attention to this memory company, a more noteworthy technological main line gradually emerges: Hongxinyu did not begin to venture into controllers only after its business grew large; rather, from its early days, it embarked on a path of coordinated development of controller chips and memory modules. From consumer-grade to enterprise-grade: Hongxinyu's next step does not start from zero If "controller chips + memory modules" is the technology route Hongxinyu laid out since its founding, then the achievement of its first strategic development cycle was to build scale, customers, and technological accumulation in the consumer-grade memory market and rank among the industry's leading companies. Centering on the needs of consumer electronics customers, the company gradually formed three mutually supporting capabilities: understanding customers' application needs, translating those needs into specific product features, and driving new products through validation and commercialization. These capabilities are also an important foundation for Hongxinyu's expansion into more application scenarios. The consumer-grade business validated the path of technology and product coordination, while the enterprise-grade market provides new room for this set of capabilities. Accumulated experience in controller design, firmware development, memory media analysis, and testing and validation is expected to help Hongxinyu extend the product development and customer service experience formed in the consumer-grade market to enterprise-grade memory. The application documents disclose that the company has begun mass production of memory products used in enterprise-grade scenarios, which means the related expansion has entered the product implementation stage. Looking again at Hongxinyu's controller path from this perspective, its significance is not only to support existing revenue, but also to provide a technological foundation for the company's expansion from consumer-grade to enterprise-grade: entering new scenarios with already accumulated capabilities, and then using the demands of new scenarios to drive technological upgrading. Layout from the start: beyond modules, there is also a chip To understand this route, one must first understand the role of controller chips. For NAND Flash-based memory products, flash particles provide the storage medium, while the controller chip and firmware handle data management tasks. How data is read and written, how errors are corrected, and how wear is controlled are all related to this software-hardware system. The same flash particles, paired with different controllers, firmware, and product designs, may ultimately deliver different performance, power consumption, and reliability. Therefore, the controller is not an ordinary accessory on a module, but an important link affecting product competitiveness. Hongxinyu's development timeline provides concrete evidence for this technology route. The company was founded in December 2018, and in 2019 it launched product lines including embedded memory, solid-state drives, and mobile storage; in the same year, the company's official website recorded the tape-out of the HG2319 USB 3.2 flash drive controller chip. The industrialization of memory products and the layout of controller chips were already unfolding simultaneously in the company's early stage. Since then, controller products have gradually extended along different application directions. According to disclosures on the company's official website, in 2020, HG2319 achieved mass production, and the HG2309 USB 2.0 flash drive controller was taped out; in 2021, the HG2283 PCIe Gen34 SSD controller and the HG2232 eMMC 5.1 controller were taped out; in 2022, the HG2283 and HG2229 eMMC 5.1 controllers achieved mass production. In 2023, the HG2259 SATA SSD controller was taped out; in 2024, the company disclosed that its PCIe Gen44 SSD controller was taped out in the fourth quarter, and the HG2325 UFS 2.2 controller also completed tape-out; in 2025, the HG2233 eMMC 5.1 controller was taped out. Extending from USB to SSD, eMMC, and UFS, Hongxinyu's controller layout gradually covers more memory applications. The significance of these milestones is not simply the addition of several chip models, but rather that controller development has continued to advance. Of course, tape-out, sampling, and mass production represent different stages and cannot be conflated; the value of a technology layout must ultimately be demonstrated through product validation and commercial delivery. A reference to a mainland version of Phison: what is similar is the development path If one needs to find a reference to help investors understand Hongxinyu, Phison Electronics, a company listed in the Taiwan capital market, is a name that cannot be avoided. Phison Electronics (Phison, ticker 8299) has long been engaged in the NAND Flash controller chip and memory solution business, with capabilities covering chip design, firmware development, system integration, and memory products. The industry lesson it provides is that memory companies can, by mastering controllers and firmware, extend competition from single hardware supply to the design and optimization of complete solutions. From the business path of "controller chips + memory modules," one side of Hongxinyu can be understood as a "mainland version of Phison." The analogy here is to the model of technology and product coordination, and does not mean that the two companies are already completely identical in technology generation, product coverage, or market position. Between the two, there is not only similarity in model. The application documents disclose that Phison participated in Hongxinyu's initial investment through an entity under it and further increased its investment in 2020; Hongxinyu also cooperated with Phison in areas such as technical services and intellectual property licensing in its early days. At the same time, the documents disclose that the company subsequently established independent R&D capabilities and continued to accumulate its own technology and intellectual property. Therefore, a more accurate statement is: on the basis of industrial cooperation, Hongxinyu laid out controller and module coordination from its early days and continued to strengthen independent R&D capabilities. This is closer to its actual development process than the narrative that "all technologies were independently developed from scratch." From chips to products: R&D investment is forming more complete capabilities In 2026, Hongxinyu's controller route made new public progress. At COMPUTEX 2026, the company released its self-developed HG2325 UFS 2.2 controller and supporting embedded memory solutions. According to the official introduction, the chip adopts a 22nm process, is equipped with a self-developed 4KB LDPC hardware error-correction engine, and supports mechanisms such as wear leveling, bad block management, and power-loss data protection. What is noteworthy about this release is that the controller chip and supporting memory solution were unveiled together. For customers, what is needed is not just chip parameters, but products that can run stably in actual terminals; for Hongxinyu, the potential value of controller-module coordination is precisely to create more direct feedback between chip development and terminal demand. The R&D pipeline also continues to extend forward. According to publicly filed application documents in July 2026, in addition to controllers already in mass production, the company has eight other independently designed and developed controllers at different stages: UFS 2.2 and eMMC 5.1 are in the sampling stage; USB 3.2, UFS 4.1, and PCIe 4.0 SSD are in the tape-out stage; PCIe 5.0 SSD and low-power UFS 2.2 are in the back-end design stage; UFS 5.0 is in the front-end design stage. Supporting this route is continuous investment. From 2023 to 2025, Hongxinyu's R&D expenses were approximately RMB 283 million, RMB 392 million, and RMB 493 million, respectively, totaling approximately RMB 1.168 billion over the three years. Among these, one reason for the increase in R&D expenses in 2025 was an increase in technical service fees mainly related to controller chip R&D. These are the company's overall R&D expenses and cannot all be equated with controller chip investment, but they can reflect the intensity of its continued building of technological capabilities. The company has also been recognized as the Guangdong Provincial Engineering Technology Research Center for High-Performance Storage Control Chips and has built a technology system around chip development, firmware, testing, and media analysis. It should be distinguished that its controllers mainly serve its own memory products; not all products use self-developed controllers, nor does this mean the company manufactures its own memory wafers. The label of "China's largest unlisted independent memory manufacturer" has brought Hongxinyu into the view of more investors, while the development trajectory of its controller chips has made the company's technological profile clearer. What is worth watching is not only "how many memory products it has sold," but also whether it can continuously translate chips, firmware, and application experience into products customers are willing to choose. From laying out "controller chips + modules" in its early days, to building scale and a customer base in the consumer-grade market, and then extending into enterprise-grade scenarios, Hongxinyu's growth logic is becoming clear: it must not only do well with existing products, but also ensure that its long-accumulated technology and commercialization capabilities are validated in new markets.