
The relationship between Shanghai Fudan 01385.HK and memory chips mainly lies in non-volatile memory, not HBM, DRAM, or consumer SSD NAND. You can understand it as a niche memory name within Hong Kong-listed semiconductor localization themes. Its memory products cover EEPROM, NOR Flash, and SLC NAND, mainly serving automotive electronics, industrial control, communications, smart meters, security systems, IoT devices, and other embedded applications. If you are tracking AI-related memory themes, you first need to distinguish mainstream high-capacity memory from embedded high-reliability memory, so you do not treat all “memory chip concepts” as the same stock logic.

Shanghai Fudan 01385.HK can be considered a “memory chip-related stock,” but it should not be simply described as a “pure memory chip stock.” A more accurate description is that it is an integrated chip design company covering security and identification chips, FPGA, non-volatile memory, smart meter chips, and testing services. Among these segments, non-volatile memory is the core business line connecting the company to the memory chip theme. In the 2025 annual report summary of Shanghai Fudan Microelectronics Group Co., Ltd., the company disclosed that its non-volatile memory products mainly include EEPROM, NOR Flash, and SLC NAND Flash.
This means Shanghai Fudan’s memory logic is not the HBM used in AI servers, nor is it comparable to mainstream DRAM/NAND giants such as Samsung, SK Hynix, or Micron. Its memory business is more focused on embedded, industrial-grade, automotive-grade, and domestic substitution-related storage chips. Its products are usually not the SSDs that consumers directly see when buying computer storage, but small- or mid-capacity chips embedded inside smart meters, communications equipment, security devices, modules, industrial control systems, and automotive electronic components.
| Business Line | Representative Products | Connection to Memory Logic | Key Factors to Watch |
|---|---|---|---|
| Non-volatile memory | EEPROM, NOR Flash, SLC NAND | Yes, the most direct link | Pricing cycle, capacity expansion, customer certification |
| FPGA and other products | FPGA, PSoC, edge AI-related products | Indirect | Edge AI, domestic substitution |
| Security and identification chips | RFID, NFC, security MCU | Indirect | Payment, identity recognition, IoT |
| Smart meter chips | Smart meter MCU, general MCU | Indirect | Grid upgrades, overseas smart meters |
| Testing services | Wafer testing, finished product testing | Indirect | Semiconductor testing demand |
From a stock classification perspective, the keywords for 01385.HK are not simply “memory price increase,” but “Hong Kong semiconductor + domestic chips + niche memory + FPGA + embedded applications.” If the market is speculating on HBM, AI server memory, and GPU supply chains, Shanghai Fudan may not be the most direct beneficiary. If the market is focused on domestic non-volatile memory, automotive-grade chips, IoT modules, industrial equipment, and security chips, its relevance becomes much clearer.
You should also note that Shanghai Fudan has both A-share code 688385 and Hong Kong stock code 01385.HK. They refer to the same company, but market liquidity, valuation level, investor structure, trading currency, and trading rules are not exactly the same. Hong Kong-listed 01385.HK is more suitable for observing the company’s relative position within the semiconductor sector from an international investor’s perspective, rather than only looking at short-term sentiment in A-share chip themes.
Summary: Shanghai Fudan 01385.HK is not a pure HBM or DRAM leader, but it can be included in a “memory chip-related stock” watchlist. Its core connection is non-volatile memory, including EEPROM, NOR Flash, and SLC NAND. When evaluating the stock, you need to treat it as a diversified chip design company, rather than directly applying every memory chip market cycle to it.

EEPROM, NOR Flash, and SLC NAND all belong to non-volatile memory, meaning they can retain data after power is turned off. However, they solve different problems. EEPROM is mainly used for small-capacity parameter and identity data storage. NOR Flash is more suitable for code, firmware, and boot programs. SLC NAND is used for higher-capacity data or code storage where reliability remains important. Shanghai Fudan’s non-volatile memory NVM product line lists three major categories: EEPROM, SPI NOR Flash, and NAND Flash, with capacities ranging from Kbit to Gbit levels.
EEPROM is like a “small notebook” inside a device. It does not store large files, but configuration parameters, identity information, calibration data, and status records. For example, metering parameters in smart meters, identity data in modules, configuration values in industrial equipment, and other small but important data may need to be preserved after power loss and support repeated read/write operations. Shanghai Fudan’s annual report states that its EEPROM series supports I²C, I3C, SPI, Micro Wire, and other interfaces, covering applications such as mobile phone modules, smart meters, communications, home appliances, automotive electronics, medical instruments, and industrial control meters.
NOR Flash is more like the “boot drive” inside a device. It is suitable for storing firmware, bootloaders, control programs, and directly readable code. Routers, set-top boxes, security cameras, IoT modules, display panels, automotive electronic control units, and similar devices may all need NOR Flash to store boot code. Shanghai Fudan’s SPI NOR Flash product page shows that the series supports SPI, Dual SPI, Quad SPI, QPI, and other interfaces, while emphasizing high reliability, high security, and compact packaging.
SLC NAND is in a different position. It is usually better suited than NOR Flash for higher-capacity requirements and has a more favorable cost per unit of capacity, but system design usually needs to consider ECC, bad block management, and interface adaptation. Shanghai Fudan’s SLC NAND Flash product page states that its SPI NAND covers 0.5Gbit to 8Gbit, includes 1.8V and 3.3V series, and provides features such as embedded ECC, write protection, and UID.
| Type | Typical Capacity | Better Suited For | Common Scenarios | Investment Implication |
|---|---|---|---|---|
| EEPROM | Kbit to Mbit | Parameters, identity data, calibration data | Meters, modules, industrial control, automotive electronics | Stable, fragmented, small-capacity demand |
| NOR Flash | Mbit to Gbit | Firmware, boot programs, code | IoT, security, communications, displays, automotive | Watch reliability, interface, and customer certification |
| SLC NAND | Gbit level | Larger-capacity code or data | Communications, security, wearables, automotive, medical | Watch capacity upgrades and cost advantage |
From an engineering perspective, EEPROM focuses on “frequent read/write operations and reliable small-data retention,” NOR Flash focuses on “fast reading and code execution,” while SLC NAND focuses on “higher capacity and a balance between cost and reliability.” Micron NOR Flash also emphasizes that NOR Flash is commonly used in scenarios requiring fast code execution, high reliability, and secure data storage. Infineon NOR flash memory lists automotive, industrial, and embedded systems as core application areas.
Summary: EEPROM, NOR Flash, and SLC NAND are all memory chips, but they are not the same business. EEPROM is about small-capacity, high-reliability storage. NOR Flash is about code and firmware storage. SLC NAND is about higher capacity and cost structure. The value of Shanghai Fudan’s memory business comes from the long-term embedding of these products in different terminal devices, not from a single consumer storage market.

Shanghai Fudan is more of a niche non-volatile memory company than an HBM mainline stock. The distinction is simple: HBM mainly serves GPUs, AI accelerators, and high-performance computing, with bandwidth, stacked packaging, DRAM capacity, and advanced packaging as core indicators. Shanghai Fudan’s EEPROM, NOR Flash, and SLC NAND are more focused on embedded devices, industrial control, automotive electronics, smart meters, communications modules, and security devices. Both fall under the broad memory category, but downstream customers, product form, pricing cycles, and competitive landscapes are very different.
HBM is a key memory component in the AI compute chain and is usually tied to NVIDIA GPUs, AI servers, cloud capital expenditure, and advanced packaging capacity. Ordinary NAND and DRAM are more affected by server, PC, smartphone, data center, and consumer electronics cycles. By contrast, niche non-volatile memory may not have very large capacity per chip, but it places greater emphasis on reliability, consistency, long product life cycles, interface compatibility, stable supply, and customer qualification.
Shanghai Fudan’s annual report describes the industry clearly: non-volatile memory is widely used in automotive electronics, consumer electronics, computers, network communications, industrial electronics, and security monitoring. EEPROM, NOR Flash, and NAND Flash each have differentiated cost advantages across different capacity ranges, forming stable application areas and niche markets. This shows that the company’s memory business is not merely a bet on one broad commodity pricing cycle, but rather product coverage across multiple embedded scenarios.
| Comparison | HBM / DRAM / High-Capacity NAND | EEPROM / NOR Flash / SLC NAND |
|---|---|---|
| Main Demand | AI servers, GPUs, data centers, PCs, smartphones | Industrial control, automotive electronics, communications, security, IoT |
| Core Metrics | Bandwidth, capacity, advanced packaging, process technology | Reliability, interface, endurance, supply stability |
| Pricing Logic | More tied to commodity memory cycles | More dependent on niche markets and customer mix |
| Competitors | Samsung, SK Hynix, Micron, etc. | Fudan Micro, GigaDevice, Winbond, Infineon, etc. |
| Investment Pitfall | Assuming AI memory price increases directly apply | Need to look at specific products and financial contribution |
This difference also explains why “memory industry price increases” cannot be mechanically applied to 01385.HK. For example, Omdia noted that NAND revenue and pricing rose significantly in the first quarter of 2026, influenced by AI, data centers, and supply constraints. But such a high-capacity NAND cycle does not necessarily mean Shanghai Fudan’s SLC NAND revenue will move in the same proportion. For Shanghai Fudan, you need to watch its own customer orders, inventory impairment, product certification, mature-node supply, and fragmented downstream demand.
One important feature of niche memory is that customer introduction cycles may be longer, but once a product enters industrial, automotive, communications, smart meter, or similar terminal supply chains, replacement may not happen frequently. For investors, this means short-term explosiveness may be lower than that of HBM leaders, but the stock may still offer medium- to long-term observation value in domestic substitution, automotive-grade chips, industrial digitalization, and IoT expansion.
Summary: Shanghai Fudan and HBM both sit within the broad “memory-related” framework, but they occupy very different positions. HBM is high-bandwidth memory for AI computing, while Shanghai Fudan’s memory business focuses on EEPROM, NOR Flash, SLC NAND, and other non-volatile memory products. When analyzing 01385.HK, do not look only at AI memory hype; you also need to examine the company’s own product structure and customer demand.
Demand elasticity for Shanghai Fudan’s memory chips mainly comes from automotive electronics, industrial control, smart meters, network communications, security monitoring, IoT modules, medical instruments, and wearable devices. These scenarios do not share the goal of maximizing storage capacity. Instead, they require chips to retain data after power loss, read firmware reliably, support long-term supply, and operate in complex environments. If downstream devices continue becoming smarter, more connected, and more localized in their supply chains, demand for EEPROM, NOR Flash, and SLC NAND may expand across many fragmented applications.
Start with EEPROM. It usually appears in places where devices need to save “small but important” data, such as meter parameters, device calibration values, identity authentication information, module configuration, and status records. Smart meters, industrial control meters, automotive electronics, home appliances, and medical instruments all require high reliability. The value of each chip may not be large, but use cases are broad, customer bases are diverse, and long-tail demand can form.
Next is NOR Flash. It is often used for code storage, firmware upgrades, and device booting. IoT modules, routers, security cameras, display screens, set-top boxes, Ukeys, WiFi/Bluetooth modules, and similar devices often require stable Flash storage for boot programs. As device functions become more complex, firmware capacity may also rise. Companies such as Winbond SLC NAND have also discussed the possibility of higher-capacity code storage shifting from NOR to high-reliability NAND in certain automotive scenarios.
SLC NAND is more suitable for embedded devices that need higher capacity while still emphasizing reliability. Communications equipment, security monitoring systems, automotive devices, medical instruments, and wearable products may need to store system data without requiring the very large capacity of consumer SSDs. Shanghai Fudan’s SPI NAND products, with compact packaging, low pin count, ECC, and different voltage series, correspond well to these embedded design needs.
Six variables may affect demand for Shanghai Fudan’s memory products:
Downstream structure also determines that Shanghai Fudan’s memory business should not be judged only by consumer electronics cycles. Smartphone, PC, and SSD inventory cycles affect industry sentiment, but the company’s real elasticity may come more from industrial-grade, high-reliability, and niche customer certification. For ordinary investors, there is no need to track every terminal data point. It is more important to monitor whether the company discloses new products, new certifications, new customers, capacity upgrades, and revenue growth in non-volatile memory.
Summary: Demand for Shanghai Fudan’s memory chips is not driven by a single explosive application, but is distributed across automotive electronics, industrial control, smart meters, communications, security, IoT, and medical devices. As terminal equipment becomes smarter, more connected, and more localized, non-volatile memory has stable use cases. However, this demand needs to be verified through financial revenue, customer adoption, and inventory changes, rather than judged solely by sector themes.
To assess 01385.HK’s memory upside, you should not only look at the “memory chip concept.” You need to evaluate revenue growth, the share of non-volatile memory, gross margin changes, R&D expenses, inventory impairment, and cash flow. Shanghai Fudan’s 2026 first-quarter report showed revenue of about RMB 1.032 billion, up 16.23% year over year. Non-volatile memory revenue was about RMB 293 million, up 23.39% year over year. This indicates a stage of recovery in the memory product line, but earnings elasticity still depends on R&D spending, inventory, and product pricing.
In its 2026 FIRST QUARTERLY REPORT, the company disclosed R&D investment of about RMB 316 million in the first quarter, accounting for 30.65% of revenue. This is important for a semiconductor design company. High R&D spending supports product iteration and process platform expansion, but it can also suppress short-term profit release. Since Shanghai Fudan has multiple product lines beyond memory, including FPGA, security and identification chips, and smart meter chips, R&D expenses cannot be simply attributed to one business.
Inventory is another key signal. The report mentioned that price increases in non-volatile memory led to the reversal of part of the inventory impairment provision, reducing asset impairment losses. This shows that memory pricing cycles can affect profit performance. If prices rise and inventory structure improves, earnings may benefit. If demand weakens, inventory accumulates, or product prices fall, impairment pressure may return.
| Financial Metric | Meaning for Memory Business | Positive Signal | Risk Signal |
|---|---|---|---|
| Non-volatile memory revenue | Measures product-line momentum | Growth faster than company average | Growth slowdown or falling share |
| Gross margin | Reflects pricing and product mix | Higher share of high-reliability products | Price declines, cost pressure |
| Inventory and impairment | Reflects cycle pressure | Impairment reversal, inventory decline | Inventory buildup, more impairment |
| R&D spending | Affects long-term competitiveness | Continuous new product launches | High expenses without revenue conversion |
| Operating cash flow | Measures collection and operating quality | Improving cash flow | High receivables and inventory burden |
| Customer certification | Determines long-term order stability | More industrial and automotive customers | Certification progress below expectations |
You also need to check whether companywide profit moves in line with revenue. In 2025, the company’s revenue increased, but net profit attributable to shareholders declined. This shows that semiconductor companies cannot be judged only by headline revenue growth. Product pricing, expenses, impairment, R&D pace, share-based payments, government subsidies, and product mix can all affect profit. For 01385.HK, growth in non-volatile memory revenue is a positive signal, but it is not a complete reason to buy the stock.
At the trading level, fees and exchange rates also affect actual returns. Hong Kong and U.S. semiconductor stocks are often placed in the same watchlist. If you compare Shanghai Fudan, A-share chip ETFs, U.S. memory stocks, and AI semiconductor companies at the same time, you are not only comparing stock price movements, but also trading costs, FX costs, portfolio concentration, and liquidity. Biya can be used for multi-asset market and trading management. For U.S. stock trading, U.S. stock trading fees should be checked against the fee schedule and order display. Biya charges USD 0 commission for U.S. stock trading, while platform fees, external institutional fees, and other fees should be verified according to actual rules. Availability of related services depends on the user’s location, identity verification results, platform rules, and applicable laws and regulations.
Summary: The memory upside of 01385.HK should be broken down from financial reports, not guessed from sector themes. Key indicators include non-volatile memory revenue, pricing changes, inventory impairment, R&D spending, and cash flow. Short-term price increases may improve earnings, but high R&D, inventory cycles, and a multi-business structure can also affect final results. Before trading, investors should also check fees, exchange rates, and market liquidity instead of focusing only on the theme.
Ordinary investors can place Shanghai Fudan in a “Hong Kong semiconductor localization + niche memory + FPGA” watchlist, rather than treating it as a peer of Micron, SK Hynix, or Samsung Electronics. A more practical approach is to use a four-step framework: product line, downstream demand, financial delivery, and valuation. First understand what chips the company sells, then identify where demand comes from, then check whether financial results confirm the story, and finally evaluate Hong Kong valuation, liquidity, and AH-share differences.
The first step is labeling the company correctly. Shanghai Fudan is not a pure HBM stock, not a pure NAND stock, and not a pure AI stock. Its strength lies in a multi-product portfolio: non-volatile memory provides the memory concept entry point, FPGA connects to edge AI, while security and identification chips and smart meter MCUs add other downstream exposures. If the label is wrong, the valuation comparison that follows will also be wrong.
The second step is to watch product-line upgrades. Is EEPROM entering more high-reliability scenarios? Are NOR Flash capacity, interfaces, and packaging continuing to expand? Is SLC NAND gaining traction in communications, security, automotive electronics, and other scenarios? These questions are more important than simply asking whether “memory prices are rising.”
The third step is financial verification. Is non-volatile memory revenue continuing to improve? Is gross margin recovering? Is inventory falling? Is R&D spending producing new products? Is operating cash flow healthy? If there is only industry heat but no company-level data follow-through, stock price movements may be driven mostly by sentiment.
Before buying or tracking the stock, you can check the following questions:
If you follow cross-market semiconductor investments, you can use U.S. stock information search to compare U.S. memory, AI chip, and semiconductor equipment companies, then look at 01385.HK to understand where Hong Kong semiconductor names stand. If your portfolio involves U.S. dollars, Hong Kong dollars, and other currencies, real-time exchange rates should also be included in cost calculations. Public market information, trading rules, and fee structures are only analytical inputs and do not constitute investment advice.
Summary: Shanghai Fudan is better viewed as a composite chip stock, not a simple pure-play memory name. For beginners, the four-step framework is more useful: product line, downstream demand, financial delivery, and valuation plus trading costs. Once you distinguish niche memory from the HBM mainline, you can better understand the real position of 01385.HK in the Hong Kong semiconductor universe.
If you track Shanghai Fudan 01385.HK, A-share chip ETFs, U.S. memory stocks, AI semiconductor equipment, and digital assets at the same time, what you need to manage is not only “which stock is hotter,” but also trading hours, pricing currencies, fee structures, FX changes, and portfolio concentration across different markets. Biya is a global multi-asset trading wallet that supports U.S. stocks, Hong Kong stocks, and cryptocurrency trading, as well as payments in more than 40 local currencies. Before using any related service, you should confirm your location, identity verification status, platform rules, and applicable laws and regulations. Semiconductor stocks can be volatile, especially when memory cycles and AI themes move together. Before trading, you should fully understand order types, fee structures, liquidity risks, and your own risk tolerance.
Shanghai Fudan 01385.HK is related to memory chips, but it is not an HBM mainline company. Its memory business mainly includes EEPROM, NOR Flash, and SLC NAND, which are more focused on non-volatile memory and embedded applications. HBM is more directly driven by AI servers, GPUs, and advanced packaging, so the two should not be treated as the same theme.
Shanghai Fudan EEPROM products are mainly used for small-capacity, high-reliability data retention. Typical use cases include smart meter parameters, module identity information, automotive electronics configuration, industrial control records, and medical device calibration data. The focus is not large capacity, but power-off retention, stable read/write performance, and long-term reliability.
Shanghai Fudan NOR Flash can be used for code, firmware, and boot program storage in MCU or embedded devices. IoT modules, industrial control equipment, communications devices, and automotive electronic control units often need stable booting and firmware upgrades, so NOR Flash has strong complementary demand with MCUs, SoCs, and communications modules.
Shanghai Fudan SLC NAND is more focused on embedded and high-reliability applications. It is not the same as the high-capacity TLC or QLC NAND commonly used in consumer SSDs. SLC NAND stores 1 bit of data per cell and usually emphasizes endurance, reliability, and industrial adaptability, but its capacity and cost structure differ from consumer SSD NAND.
When evaluating Hong Kong-listed 01385.HK, key indicators include non-volatile memory revenue, gross margin, inventory impairment, R&D spending, operating cash flow, and growth differences across product lines. Looking only at the “memory chip concept” is not enough. Pricing cycles, customer certification, product structure, and Hong Kong stock liquidity can all affect investment judgment.
Shanghai Fudan A-share 688385 and Hong Kong stock 01385.HK refer to the same company, but they trade in different markets. They may differ in trading currency, trading hours, liquidity, investor structure, valuation level, and transaction costs. When comparing AH shares, investors should refer to real-time prices, exchange rates, and account rules.
*This article is provided for general information purposes and does not constitute legal, tax or other professional advice from BiyaPay or its subsidiaries and its affiliates, and it is not intended as a substitute for obtaining advice from a financial advisor or any other professional.
We make no representations, warranties or warranties, express or implied, as to the accuracy, completeness or timeliness of the contents of this publication.



