AI-driven demand indicates a prolonged semiconductor upcycle extending into 2026 and beyond.

AI-driven demand indicates a prolonged semiconductor upcycle extending into 2026 and beyond.

      Credit: UBS

      During the UBS Greater China Conference 2026 in Shanghai on Wednesday, Jimmy Yu, a Semiconductor Analyst for China Technology at UBS Securities, presented his insights on the cycles of the semiconductor industry in China and globally, the growth dynamics driven by AI, and investment themes. He indicated that with the rapid advancement of AI, the global semiconductor industry is entering a new upturn, expected to demonstrate stronger growth momentum and durability compared to previous cycles.

      From a global standpoint, Yu anticipates the semiconductor market will near $700 billion in 2025 and exceed $1 trillion in 2026, equating to year-on-year growth exceeding 40%. By 2027, the market is forecasted to reach approximately $1.18 trillion, with growth remaining in the double digits. While such growth is historically not unusual, Yu emphasized that the crucial distinction this time is the structural demand transformation triggered by AI.

      The memory segment is projected to be a key contributor to this cycle. Yu mentioned that the memory market could expand by nearly 50% by around 2027, primarily due to rising prices enhancing overall revenue. Even without considering memory, the demand for logic chips, foundry services, and related sectors remains strong, underpinning ongoing double-digit growth from 2025 to 2027.

      On the demand front, AI applications are rapidly proliferating beyond data centers and cloud computing into areas like financial services, enterprise applications, and security-related sectors. Although some applications are still in their infancy, their potential market size is becoming increasingly apparent, Yu noted. Consequently, investment in AI infrastructure is likely to emerge as one of the most reliable growth drivers for the semiconductor industry in the coming years.

      Regarding industry cyclicality, Yu pointed out that semiconductors remain a fundamentally cyclical sector. Based on global revenue growth projections, the industry could hit a near-term peak around the third quarter of this year. However, indicators such as foundry capacity utilization, capital expenditures, and corporate profitability suggest that the cycle peak may be postponed until the third quarter of next year. He added that equity markets usually anticipate cyclical changes one to two quarters in advance, leading investors to position themselves ahead of underlying fundamentals.

      Focusing on China, Yu stated that the development of AI infrastructure is significantly enhancing domestic demand for computing chips, advanced process nodes, advanced packaging, and semiconductor equipment. Though China still trails behind its international counterparts in advanced manufacturing capabilities, there is considerable scope for improvement in localization and self-sufficiency.

      In recent years, revenue growth among Chinese semiconductor equipment companies has surpassed the industry average, with several subsegments showing signs of sustained progress.

      AI is transforming the demand landscape of the semiconductor industry. Demand for high-performance, low-power chips is accelerating across various applications, from cloud computing and edge technologies to autonomous vehicles, robotics, and intelligent devices. Supported by both global and China-specific factors, the current semiconductor upcycle is anticipated to continue into 2026 and possibly beyond.

      Jessie Wu is a tech reporter based in Shanghai, covering consumer electronics, semiconductors, and the gaming industry for TechNode. You can connect with her via e-mail: jessie.wu@technode.com. More by Jessie Wu.

AI-driven demand indicates a prolonged semiconductor upcycle extending into 2026 and beyond. AI-driven demand indicates a prolonged semiconductor upcycle extending into 2026 and beyond. AI-driven demand indicates a prolonged semiconductor upcycle extending into 2026 and beyond. AI-driven demand indicates a prolonged semiconductor upcycle extending into 2026 and beyond.

Other articles

From the laboratory to the living room: how CSOT transforms research into dependable displays. From the laboratory to the living room: how CSOT transforms research into dependable displays. With the rising prevalence of consumer electronics and smart devices, display technology is taking on a more significant role in electronic devices. Chinese Within Makera's investment in worldwide creators as desktop CNC technology gains traction. Within Makera's investment in worldwide creators as desktop CNC technology gains traction. In recent years, as digital manufacturing technology has become increasingly available, desktop-level tools for creators and small enterprises have begun to How Flippa is Breaking Down Language Barriers in Global Transactions How Flippa is Breaking Down Language Barriers in Global Transactions Flippa is eliminating language obstacles in international mergers and acquisitions with its AI-driven multi-language Deal Room, assisting buyers and sellers in finalizing cross-border digital business transactions. How Flippa Is Eliminating Language Barriers in International Deal-Making How Flippa Is Eliminating Language Barriers in International Deal-Making Flippa is breaking down language obstacles in global M&A through its AI-driven multi-language Deal Room, assisting buyers and sellers in finalizing cross-border digital business transactions. Inside Makera's investment in global makers as desktop CNC technology reaches its potential. Inside Makera's investment in global makers as desktop CNC technology reaches its potential. In recent years, with the increasing accessibility of digital manufacturing technology, desktop-level tools for creators and small enterprises have begun to Investigating physical AI: Linkerbot is focused on equipping robots with human-like hands to operate in the real world. Investigating physical AI: Linkerbot is focused on equipping robots with human-like hands to operate in the real world. If large language models act as the brains of machines, then agile robotic hands are evolving into their physical forms. As AI transitions from the digital realm into

AI-driven demand indicates a prolonged semiconductor upcycle extending into 2026 and beyond.

During the UBS Greater China Conference 2026 in Shanghai on Wednesday, Jimmy Yu, a Semiconductor Analyst for China at UBS Securities, presented his