Is Shanghai Fudan an HBM Concept Stock? Business Boundaries and Non-Volatile Memory Explained

Shanghai Fudan should not be simply classified as an HBM concept stock. Its storage-related business mainly focuses on non-volatile memory, including EEPROM, SPI NOR Flash, and SLC NAND Flash, rather than HBM, DRAM, or large-capacity 3D NAND manufacturing. When you analyze 01385.HK, a more accurate framework is “domestic IC design + niche memory + FPGA + security and identification chips + smart meter chips,” rather than placing it directly into the AI GPU high-bandwidth memory supply chain.

Key Takeaways

• Shanghai Fudan is not a typical HBM or DRAM memory manufacturer.
• Non-volatile memory is its main storage-related business.
• FPGA and related products are key high-margin indicators.
• Its AI relevance is more tied to edge computing and high-reliability chips.
• Investment analysis should focus on product mix, gross margin, and inventory risk.
• “Storage concept” should not be directly equated with “HBM concept.”

Can Shanghai Fudan Be Considered an HBM Concept Stock?

Shanghai Fudan should not be directly defined as an HBM concept stock. HBM is high-bandwidth stacked DRAM designed for AI GPUs, HPC, and high-end accelerators, usually involving DRAM dies, TSV, advanced packaging, and high-bandwidth interconnects with GPUs or ASICs. In its 2025 annual report, Shanghai Fudan’s disclosed major product lines are security and identification chips, non-volatile memory, smart meter chips, FPGA and related products, and integrated circuit testing services, not HBM.

The reason the market often searches Shanghai Fudan together with HBM mainly comes from three misunderstandings. First, many investors treat “memory chips” as one broad concept without distinguishing HBM, DRAM, NAND, NOR Flash, EEPROM, and SLC NAND. Second, Shanghai Fudan has non-volatile memory, FPGA, and high-reliability chip businesses, so it is often included in broader semiconductor, AI chip, domestic substitution, and memory-related themes. Third, the AI boom has strengthened market associations around “chips + memory + computing power,” but that association does not mean the company has direct HBM exposure.

HBM concept stocks require more specific business characteristics. For example, Micron HBM targets AI and supercomputing scenarios and emphasizes high-bandwidth memory, while Micron HBM3E further emphasizes high capacity, high bandwidth, and performance for AI innovation. Companies that directly benefit from HBM are usually clearly positioned in DRAM manufacturing, HBM stacking, advanced packaging, HBM testing, key materials, equipment, or GPU/AI accelerator supply chains. Calling a company an HBM play simply because it has “memory” products can easily mislead valuation judgment.

A more accurate label for Shanghai Fudan is a multi-product domestic IC design company, a niche non-volatile memory supplier, and an FPGA and high-reliability chip company. Its AI relevance can be discussed through edge computing, industrial control, signal processing, domestic substitution, and high-reliability applications, but it should not be forced into the HBM supply-demand cycle. If you are searching “Is Shanghai Fudan an HBM concept stock?”, the real question is not whether the company has the word “memory” in its business, but whether its memory type, customer scenarios, and profit model match the HBM business model.

Summary: Shanghai Fudan can be described as storage-related, domestic semiconductor-related, and AI edge-computing-related, but it should not be simply described as an HBM concept stock. HBM is high-bandwidth stacked DRAM for AI GPUs and high-performance computing, and its investment logic revolves around DRAM capacity, advanced packaging, AI accelerator demand, and high-bandwidth interconnects. Shanghai Fudan’s publicly disclosed storage business mainly focuses on non-volatile memory, including EEPROM, SPI NOR Flash, and SLC NAND Flash. Its real investment logic is closer to niche memory, FPGA, high-reliability chips, security and identification chips, and domestic IC design, rather than the global HBM supply chain.

How to Understand Non-Volatile Memory

Shanghai Fudan’s core storage business is non-volatile memory, meaning memory chips that can retain data after power is turned off. This is very different from HBM: HBM focuses on extremely high bandwidth, low latency, and AI computing acceleration, while non-volatile memory focuses on data retention, low power consumption, reliability, cost, and embedded applications. Shanghai Fudan’s NVM product line includes EEPROM, SPI NOR Flash, and NAND Flash, which is the key to understanding the company’s storage boundary.

EEPROM usually has smaller capacity and is suitable for storing parameters, identity information, calibration data, and small amounts of rewritable data. It is commonly used in automotive electronics, industrial control, smart meters, consumer electronics, and security identification. SPI NOR Flash is more suitable for code storage, boot programs, firmware, and embedded system applications, and is commonly used in IoT, communications equipment, consumer electronics, and industrial devices. SLC NAND Flash offers higher capacity and stronger reliability than multi-level NAND, making it suitable for certain embedded data storage, security, and industrial scenarios.

Shanghai Fudan’s 2025 annual report summary notes that EEPROM, NOR Flash, and NAND Flash are all non-volatile memory products, but each has differentiated cost advantages across different capacity ranges, forming relatively stable niche markets. This shows that the company’s storage business is not pursuing extreme bandwidth like HBM, but serving the need for reliable data retention in embedded systems.

Non-volatile memory is not completely unrelated to AI; its connection is just more indirect. AI does not only happen in data centers. It is also moving into edge devices, smart terminals, industrial control, automotive electronics, security systems, and smart meters. Edge AI devices need to store model parameters, configuration data, authentication information, operating status, and logs locally, which may increase demand for embedded memory chips. The annual report summary also notes that growing demand for artificial intelligence and edge computing may support memory chip demand, while automotive intelligence and edge AI are raising functional and performance requirements for memory products.

But non-volatile memory is not a substitute for HBM. NVM and HBM are completely different in technology route, speed profile, application scenario, customer structure, and supply chain position. NVM is more focused on embedded, industrial, automotive, security, communications, and consumer electronics applications. HBM is more focused on AI GPUs, HPC, data center accelerators, and high-end packaging systems. When investing, you should not treat every memory company as an HBM company just because “AI needs memory.” Nor should you benchmark Shanghai Fudan’s valuation directly against the HBM businesses of SK hynix, Micron, or Samsung simply because it has NVM products.

Summary: Non-volatile memory is the real boundary of Shanghai Fudan’s storage business. It is indeed related to AI edge computing, automotive electronics, industrial control, security, communications, and domestic substitution, but it is not HBM. You can understand NVM as the “long-term data retention layer” inside embedded systems, not the “high-speed memory bandwidth layer” inside AI GPUs. Shanghai Fudan’s NVM business should be analyzed through the market competition, product mix, customer certification, and gross margins of EEPROM, NOR Flash, and SLC NAND, rather than through HBM supply tightness, advanced packaging expansion, or AI GPU ramp-up logic.

Business Boundaries Between Shanghai Fudan and Mainstream Memory Leaders

Shanghai Fudan is not the same type of company as mainstream memory leaders such as SK hynix, Micron, and Samsung. The latter focus on mainstream high-volume memory such as DRAM, NAND, and HBM, where capital expenditure, advanced process technology, wafer manufacturing capability, and global pricing cycles are key. Shanghai Fudan follows the logic of an IC design company, with emphasis on product definition, R&D design, customer applications, and supply chain collaboration. Shanghai Fudan’s company profile shows that its products cover security and identification, smart meters, NVM, integrated circuit testing, and other areas.

Mainstream memory leaders are driven by capacity cycles. DRAM and NAND companies are usually analyzed through wafer capacity, capital expenditure, ASP, bit shipment, inventory days, AI server demand, and HBM supply-demand. HBM leaders also need to be analyzed through advanced packaging capacity, customer qualification, GPU platform attachment, yield, and unit pricing. Shanghai Fudan is different. It is not a large-scale wafer manufacturing platform, but a multi-product IC design company whose financial performance is more affected by product mix, downstream demand, R&D investment, inventory write-downs, and price competition.

Shanghai Fudan’s 2025 business structure also illustrates this difference. Its annual report shows 2025 revenue of about RMB 3.982 billion, up 10.92% year over year, while net profit attributable to shareholders was about RMB 232 million, down 59.42%. By product line, non-volatile memory revenue was about RMB 1.042 billion, down 8.30% year over year, while FPGA and related products generated about RMB 1.414 billion, up 25.19%. This shows that the company’s revenue growth was not driven by HBM, but by structural changes across multiple product lines.

Gross margin structure is even more important. In 2025, Shanghai Fudan’s FPGA and related products had a gross margin of about 74.82%, while non-volatile memory had a gross margin of about 66.08%. These high margins reflect product barriers, niche-market competition, and customer certification value, but they do not mean HBM exposure. Many investors see “storage + high gross margin” and immediately think of HBM. That is a common misunderstanding. Shanghai Fudan’s high-margin products are more related to FPGA, high-reliability applications, and niche IC design, not AI GPU high-bandwidth memory supply.

The term “storage concept” can also be misleading. Storage can refer to DRAM, NAND, HBM, EEPROM, NOR, SLC NAND, SSD, HDD, enterprise storage systems, object storage, and cloud data management. Each category has a completely different supply chain, customer base, cost structure, and valuation method. Shanghai Fudan is better placed within the framework of “niche memory + FPGA + domestic IC design,” rather than the framework of “global HBM supply chain.”

Summary: The business boundary between Shanghai Fudan and mainstream memory leaders is clear. It has a storage business, but it is not a DRAM/HBM giant. It has an AI-related narrative, but it is not part of the core AI GPU memory track. It has high-margin products, but that does not mean direct exposure to HBM supply tightness. When you analyze 01385.HK, you should break down NVM, FPGA, security and identification, smart meters, and testing services separately, rather than simply classifying the company as a memory chip leader. The real factors affecting valuation are product mix, gross margin, inventory, R&D investment, downstream applications, and domestic substitution progress.

How FPGA, Edge AI, and High-Reliability Chips Affect Valuation

Shanghai Fudan’s AI relevance comes more from FPGA, PSoC, high-reliability chips, edge computing, and domestic substitution than from HBM. FPGA can be used in signal processing, industrial control, communications, smart computing, and specific acceleration scenarios, but this logic is different from the data center track of AI GPU + HBM. When analyzing 01385.HK, you should treat FPGA and related products as one of the company’s core valuation elasticity variables, rather than focusing only on the “memory concept.”

FPGA is programmable, low-latency, parallel, and highly adaptable. It does not necessarily replace GPUs, but it can be valuable in industrial vision, communications, signal processing, edge computing, high-reliability systems, and certain AI inference scenarios. In its 2024 interim report, Shanghai Fudan stated that its FPGA product line includes tens-of-millions gate FPGA, hundreds-of-millions gate FPGA, billion-gate FPGA, and embedded programmable system-on-chip PSoC, and also mentioned FPAI, a reconfigurable chip for AI applications. This shows that the company’s AI relevance is closer to programmable logic and edge intelligence than to HBM.

In 2025, FPGA and related products became one of Shanghai Fudan’s most important high-margin product lines. The annual report shows that this product line generated about RMB 1.414 billion in revenue, up 25.19% year over year, with a gross margin of about 74.82%. This is more important than the question of whether the company is an HBM concept stock, because it directly affects earnings elasticity and valuation. If FPGA, PSoC, and high-reliability chips continue to expand among customers, Shanghai Fudan’s investment logic will lean more toward domestic high-end IC design rather than a general memory cycle.

You can break down Shanghai Fudan’s main business logic into five categories:

• FPGA logic: edge computing, low latency, signal processing, domestic substitution.
• NVM logic: embedded storage, automotive electronics, industrial, and security applications.
• Security identification logic: RFID, NFC, identity authentication, financial-grade security.
• Smart meter logic: grid upgrades, MCU, metering chips, demand cycle.
• Testing service logic: capacity utilization, depreciation, labor costs, margins.

The advantage of high-reliability chips is that customer certification cycles are long, switching costs can be high, and application requirements are strict. Once a product enters a specific customer or system, customer stickiness may be strong. But the risks are also real: long R&D cycles, unstable order timing, inventory pressure, and impairment may affect profit. In its 2025 English annual report, Shanghai Fudan noted that its products are used in financial services, social security, anti-counterfeiting and traceability, network communications, home appliances, automotive electronics, industrial control, signal processing, data centers, artificial intelligence, and other fields. A broad application base is a strength, but it also means demand divergence can affect different product lines differently.

Summary: Shanghai Fudan is not a core HBM company, but it is not completely unrelated to AI. Its AI relevance should be understood through FPGA, high-reliability chips, edge AI, industrial control, smart computing, and domestic substitution. FPGA and related product revenue and gross margin are more important variables to track than an HBM label. AI may increase demand for edge devices, industrial control, high-reliability chips, and embedded memory, but this elasticity is more indirect than the direct benefit HBM receives from AI GPU ramp-up. When judging Shanghai Fudan’s valuation, you should focus on whether FPGA and related products can continue to grow, rather than simply placing the company into the HBM track.

What Financial and Risk Signals Should Investors Track for Shanghai Fudan?

When you track Shanghai Fudan, the key question is not whether it is an HBM concept stock, but its product mix, gross margin, inventory, R&D spending, customer demand, and cash flow. In 2025, the company’s revenue grew, but net profit attributable to shareholders declined sharply, showing that revenue expansion does not necessarily translate into profit growth. Inventory write-downs, R&D expenses, product price competition, and diverging downstream demand can all directly affect the valuation of 01385.HK.

Product mix is more important than a single concept label. In 2025, non-volatile memory revenue declined mainly because of intensified industry competition, while security and identification chips, smart meter chips, and FPGA and related products recorded varying degrees of revenue growth. You need to analyze each business line separately: whether FPGA and related products continue contributing high margins, whether NVM returns to growth, whether security identification and smart meter chips benefit from downstream demand, and whether testing services can improve capacity utilization.

Inventory and impairment are important risks for semiconductor design companies. Many investors assume fabless companies do not face cycle pressure because they do not own fabs, but that is not true. Fabless companies still need to arrange wafer production, packaging and testing, inventory preparation, and customer delivery in advance. If downstream demand changes, product prices fall, or inventory cycles stretch, net realizable value can decline and affect profit. Shanghai Fudan’s profit pressure in 2025 was related to inventory write-down losses and impairment of intangible assets.

International investors also need to watch Hong Kong stock liquidity and valuation volatility. 01385.HK carries several labels at the same time: Hong Kong semiconductor stock, domestic substitution, storage, FPGA, and edge AI. These themes may increase short-term volatility, but long-term performance still returns to revenue, profit, cash flow, and R&D results. If you compare Shanghai Fudan with U.S.-listed Micron, Hong Kong-listed foundries, or A-share semiconductor design companies, you also need to consider A+H share structure, Hong Kong trading volume, RMB-denominated revenue, industry cycles, and geopolitical risks.

Trading costs should not be ignored either. If you follow Hong Kong semiconductor stocks such as Shanghai Fudan while also comparing U.S. names such as Micron, NetApp, Pure Storage, and NVIDIA, you should look beyond price movement and consider commissions, platform fees, external agency fees, and order-page fee displays. If the service is available in your region and you meet the relevant platform requirements, you can review Biya U.S. stock trading fees. Biya charges $0 commission for U.S. stock trading, while platform fees, external agency fees, and other costs are subject to the fee schedule and order page. Public market information and fee structures can help you make clearer decisions, but they do not constitute investment advice.

Summary: When investing in Shanghai Fudan, you should start from financial quality and product structure, rather than defining the company through the HBM concept. The key to understanding 01385.HK is to analyze NVM, FPGA, security and identification, smart meters, and testing services separately, and then assess how each product line contributes to revenue, gross margin, and cash flow. If high-margin FPGA and related products continue to grow, NVM competition eases, and inventory write-downs decline, the valuation foundation will be more stable. If price competition intensifies, inventory continues to weigh on profit, or R&D fails to convert into revenue, the stock may be vulnerable when concept-driven enthusiasm cools.

How Should Shanghai Fudan Be Positioned in a Semiconductor Portfolio?

Shanghai Fudan is better positioned as a Hong Kong-listed domestic IC design, niche memory, FPGA, and high-reliability chip name, rather than as a core HBM stock. If you already follow Micron, SK hynix, Samsung, ASMPT, Hua Hong Semiconductor, SMIC, NetApp, Pure Storage, and other companies, Shanghai Fudan can add exposure to “domestic IC design + embedded memory + FPGA,” but it cannot replace HBM or DRAM leaders.

The comparison set matters. When comparing Shanghai Fudan with Micron, SK hynix, and Samsung, the key is not whether one can replace the other, but how their industry positions differ. When comparing it with Hua Hong Semiconductor or SMIC, remember that Shanghai Fudan is an IC design company, not a foundry platform. When comparing it with FPGA, security chip, or MCU companies, you are closer to its real business boundary. When comparing it with NetApp or Pure Storage, you need to distinguish between a chip company and an enterprise storage systems company.

You can position Shanghai Fudan in the following way:

• If you focus on HBM, track SK hynix, Micron, Samsung, and the advanced packaging chain.
• If you focus on domestic IC design, Shanghai Fudan is more representative.
• If you focus on edge AI, watch FPGA, MCU, NVM, and security chips.
• If you focus on Hong Kong semiconductors, also compare valuation, liquidity, and cycle volatility.
• If you focus on long-term allocation, watch cash flow, R&D results, and product mix.

The most common mistake retail investors make is equating “storage” with HBM, “AI-related” with data center GPUs, and “high-margin products” with guaranteed high growth. A more stable approach is to read the annual report product categories first, then analyze revenue, gross margin, and risk disclosures. Shanghai Fudan’s opportunities lie in domestic IC design, niche memory, FPGA, and high-reliability applications. Its risks lie in product competition, R&D cycles, inventory impairment, and downstream demand divergence.

If you track both Hong Kong semiconductor stocks and U.S. AI storage names, you can use Biya to follow multi-asset market information and compare 01385.HK, Micron, NetApp, Pure Storage, and other companies through Hong Kong stock information and U.S. stock information. Service availability depends on your location, identity verification status, platform rules, and applicable laws and regulations. Before trading, you should check the order page, fee details, and risk disclosures. If you need to set up an account, you can review account registration based on your own circumstances, but any trading decision should still be based on your risk tolerance and independent judgment.

Summary: Shanghai Fudan’s proper position in a semiconductor portfolio is domestic IC design, niche NVM, FPGA, and high-reliability chips, not the HBM core track. For international investors, the real question is not whether Shanghai Fudan is an HBM concept stock, but whether its product structure can continue to deliver on domestic substitution, edge AI, industrial control, automotive electronics, and high-reliability chip demand. If placed in the right business framework, it can add a specific niche dimension to a Hong Kong semiconductor portfolio. If treated as a core HBM stock, investors may overestimate the direct pull from the AI GPU memory supply chain.

FAQ

Is Shanghai Fudan an HBM concept stock?

Shanghai Fudan should not be directly defined as an HBM concept stock. Its storage business mainly includes non-volatile memory such as EEPROM, NOR Flash, and SLC NAND Flash, rather than HBM, DRAM, or large-capacity 3D NAND manufacturing. A more accurate positioning is domestic IC design, niche memory, and FPGA-related business.

What products are included in Shanghai Fudan’s non-volatile memory business?

Shanghai Fudan’s non-volatile memory products mainly include EEPROM, SPI NOR Flash, and SLC NAND Flash. They are used in automotive electronics, consumer electronics, industrial, security, and communications scenarios, emphasizing data retention after power-off, reliability, and embedded applications. This is different from the AI GPU high-bandwidth memory route of HBM.

How is Shanghai Fudan different from Micron and SK hynix?

Micron and SK hynix are more focused on mainstream memory manufacturing such as DRAM, NAND, and HBM. Shanghai Fudan is more focused on IC design, niche non-volatile memory, FPGA, security and identification chips, and smart meter chips. Their industry positions, capital intensity, customer structures, and valuation logic are different.

Can Shanghai Fudan’s FPGA business benefit from AI demand?

Shanghai Fudan’s FPGA business may indirectly benefit from edge AI, industrial control, signal processing, and high-reliability applications, but it is not part of the AI GPU + HBM data center mainstream. Investors should focus on FPGA revenue, gross margin, customer certification, product iteration, and R&D conversion.

What indicators matter when investing in 01385.HK Shanghai Fudan?

Investors should focus on FPGA and related product revenue, non-volatile memory revenue, gross margin, inventory write-downs, R&D spending, cash flow, and downstream demand changes. Do not judge the stock only by HBM or AI labels. You should also consider Hong Kong stock liquidity and your own risk tolerance before trading.

Is Shanghai Fudan suitable for a memory stock portfolio?

Shanghai Fudan can be observed as a broad storage-related and domestic IC design stock, but it should not replace HBM or DRAM leaders. If included in a memory stock portfolio, its role should be clearly defined as niche NVM, FPGA, and embedded chip exposure, not the AI GPU high-bandwidth memory mainstream.