A significant surge is underway in the semiconductor sector, propelled by capital towards new highs, with stocks like Cambricon reaching 1900 yuan and Demingli hitting 500 yuan. On the first trading day after the holiday (May 6), the storage chip segment soared by 5%, and popular leaders such as Jiangbolong, Baiwei Storage, and Puran股份 collectively pushed their share prices into the next hundred-yuan tier, clearly illustrating a bullish posture reminiscent of the recent optical communication boom.
This represents a dramatic turnaround, transforming previously overlooked stocks into currently unattainable favorites. However, from a long-term perspective, the logic for storage and AI chips is not merely a game of current stock price levels but is driven by future industrial expectations and endogenous growth. With daily surges of 20% and doubling gains within months amplifying the profit effect, a new wave in the storage and AI chip market has indeed arrived.
The collective rise in AI chips and storage chips continues, with Cambricon and Demingli pushing their share prices to new highs. Recent sessions in U.S. and South Korean markets have been exceptionally active. As noted before the holiday, global industrial logic, sector cycle resonance, and the ripple effects from leading companies in the tech arena are strengthening. On the first post-holiday day, the semiconductor rally in the A-share market was ignited, led by storage chips, clearly demonstrating this interconnected logic.
Overnight in the U.S., semiconductor leader Intel surged nearly 13%, while storage chip giants Micron Technology and SanDisk both rose over 11%, also hitting new record highs. U.S. tech leaders, serving as global technology pioneers and major capital expenditure drivers, are deeply benefiting from the AI computing wave. Notably, renowned investor Dan Bin's Q1 U.S. stock portfolio included Micron Technology. Coincidentally, Micron's stock began a new uptrend starting in Q2, having risen nearly 90% since then.
The South Korean market delivered consecutive surprises. During the recent holiday period, news of the KOSPI index surging 5% (on May 4) made headlines, but it further astonished with another over 6% jump on May 6. Behind this, two semiconductor and storage chip companies, SK Hynix and Samsung Electronics, were core drivers of sentiment. On May 6 alone, their shares rose 10% and 14% respectively, both reaching new peaks.
The strong performance of South Korean semiconductor leaders is closely tied to their dominant market share in the global high-end storage chip sector. In the HBM (High Bandwidth Memory) field, a significant portion of the global market is firmly held by SK Hynix and Samsung. Furthermore, their latest financial reports showing high growth further corroborate the sector's robust health.
Against the backdrop of an upward industry cycle, A-share semiconductor, AI chip, and storage companies are benefiting from accelerated domestic substitution, supply chain synergy, and deep integration into global supply chains. The main semiconductor trend in A-shares on May 6 featured AI chips (represented by companies like Cambricon, whose stock reached a high of 1966 yuan, and Haiguang Information), storage chips (such as Demingli, Baiwei Storage, Jiangbolong), and the interconnected semiconductor industry chain (design, equipment, packaging), with the rally continuing to spread. The storage chip segment, in particular, has surged 30% since April 7, with the bull market pattern intensifying.
Bian Huizong, who has tracked and researched AI and storage chips for years, previously identified the inflection point opportunity in this trend early, pinpointing leaders and their launch points, such as Demingli in June last year, Tuojing Technology in July, Baiwei Storage in January this year, and Cambricon in April.
The long-cycle upturn in storage chip prosperity is being confirmed through the sequential realization of earnings. AI chips, storage chips, and semiconductors are closely symbiotic and interlinked. AI chip performance directly relies on semiconductor technology, while the internal structure and function of storage chips are based on semiconductor materials. AI chips and storage chips represent a synergy between computing power and storage capacity; AI chips provide powerful computational support, and storage chips offer data storage and retrieval capabilities.
While previous discussions focused on AI chips, the emphasis here is on storage. Storage chips, operating on binary principles, etch flowing digital data into silicon substrates through the interplay of electric currents and magnetic fields. Categorically, they are divided into two main types: volatile memory (RAM) and non-volatile memory (ROM).
Volatile memory, such as DRAM (Dynamic Random-Access Memory), stores data in capacitors requiring periodic refresh to maintain data integrity. HBM, mentioned earlier, is a mainstream type within this category, serving as the system's main memory in AI applications. Non-volatile memory stores data as electrical charges, retaining information after power loss. It supports faster sequential data access and is more suitable for applications requiring large capacity and high density. In AI, it acts as the proficient handler for storing model parameters, training datasets, and other data requiring long-term preservation.
A key question arises: How long can the storage chip upcycle last? Analysis from perspectives of supply-demand dynamics, pricing, and earnings realization provides insight. The surge in storage chip demand driven by AI computing needs solidifies the fundamental basis for the industry's prosperous cycle. A single AI server utilizes several times the storage capacity of a traditional server, creating sustained, highly elastic demand. However, the capital-intensive and long-cycle nature of storage chip manufacturing imposes rigid constraints on the supply side. Under conditions of supply-demand imbalance, product prices are driven upwards. Clearly, the price increase cycle for storage aligns with the new upward cycle in the stock prices of storage leaders.
Subsequently comes earnings verification. Following the conclusion of Q1 reporting season, the performance of storage chip concept stocks showed mixed but often impressive results, with many companies reporting doubled earnings as a baseline, and several achieving growth rates in the order of tenfold. Their earnings growth directly benefits from AI application demand and rising storage prices.
Propelled by the global explosion in AI computing demand, cyclical resonance, and market sentiment transmission, the likelihood of a sudden halt or sharp reversal for the storage and AI chip trend in the short term is extremely low. This industry cycle, driven by the underlying logic of AI, is not a short-term emotional fluctuation nor the typical rise and fall of traditional cyclical electronic components. Instead, it represents a resonance of new technologies, industrial core competencies, and capital. Within this continuously heating industry cycle, a new round of capital's monetary game has commenced.
Comments