AI computing power and storage demand explode, chip capacity expansion kicks off! Semiconductor equipment welcomes a new bull market.

Recent research reports from Wall Street financial giant Citigroup stated that in the global wave of AI computing power infrastructure construction and the macro background of the "storage chip super cycle", semiconductor equipment manufacturers - especially ASML Holding NV ADR (ASML.US), Lam Research Corporation (LRCX.US), and Applied Materials (AMAT.US) - these three semiconductor equipment giants will be the biggest beneficiaries of the rapid expansion of AI chips (AI GPU/AI ASIC) and DRAM/NAND storage chip production capacity. Citigroup predicted in this report that the global semiconductor equipment sector will usher in a "Phase 2 bull market cycle," meaning that after the super bull market of 2024-25, it is likely to see a new round of bullish trajectory. Led by Citigroup's well-known semiconductor industry analyst Atif Malik, the Citigroup analyst team predicts that as demand for AI chips and storage chips continues to surge, Taiwan Semiconductor Manufacturing Co., Ltd. Sponsored ADR, Samsung Electronics, and Intel Corporation - the three largest chip manufacturers in the world - as well as storage chip manufacturer SK Hynix, are expected to significantly increase their capital spending (capex) guidance for 2026 and beyond in their upcoming financial reports. This forecast suggests that the semiconductor equipment expenditure of wafer fab equipment (WFE) in 2026 may be more likely to approach its "most optimistic outlook." In terms of stock price performance, the US semiconductor equipment sector has been very strong since the beginning of the year. ASML Holding NV ADR's ADR price hit a record high at the beginning of 2026, with a single-day increase of over 8% on January 2nd, and a total increase of 16% since the beginning of the year, reaching a market value close to $500 billion; Lam Research's stock price has also hit historical highs since the second half of 2025, with a 20% increase since the beginning of 2026; Applied Materials' stock price has also hit new highs since the beginning of 2026, with a 15% increase since the beginning of the year. The Citigroup analyst team emphasized that Micron Technology, Inc., a major competitor to SK Hynix and Samsung Electronics in the storage chip market, announced in its December 2025 earnings call that its capital spending for fiscal year 2026 (ending in August 2026) had been raised from $18 billion to $20 billion, a significant increase of 45% year-on-year. Micron also indicated that capital spending for fiscal year 2027 would continue to increase. As a direct competitor to Samsung Electronics and SK Hynix in the storage chip market, Micron's expansion measures may prompt these two South Korean storage chip manufacturers to take corresponding capital spending expansion actions to maintain their market position. It is worth noting that market expectations for Taiwan Semiconductor Manufacturing Co., Ltd. Sponsored ADR's strong chip production capacity expansion are not only focused on NVIDIA Corporation, AMD, and Broadcom Inc., the three leading AI chip companies, bringing in massive orders for data center AI chips, as well as the huge consumer electronics chip orders that Apple Inc. brings in every year. In the enterprise-level high-performance SSD sector (belonging to NAND terminal applications), the high-performance NVMe (especially PCIe Gen5/Gen6) SSD controller chips rely heavily on Taiwan Semiconductor Manufacturing Co., Ltd. Sponsored ADR's high-end process capacity - meaning that Taiwan Semiconductor Manufacturing Co., Ltd. Sponsored ADR's current capacity is far from enough to meet the "endless orders" brought by AI computing power and storage, and a significant expansion of capacity is imminent. After Alphabet Inc. Class C launched the Gemini3 AI application ecosystem in late November, this cutting-edge AI application software quickly became popular worldwide, driving an instant surge in AI computing power demand for Alphabet Inc. Class C. The Gemini3 series products brought in an incredibly large amount of AI token processing power, causing Alphabet Inc. Class C to significantly reduce the free access limits for Gemini 3 Pro and Nano Banana Pro, and temporarily restrict subscriptions for Pro users. This, coupled with recent trade export data from South Korea showing strong demand for SK Hynix and Samsung Electronics' HBM storage systems and enterprise-grade SSDs, further validates Wall Street's assertion that the "AI boom is still in the early stages of infrastructure construction where computing power supply cannot meet demand." As global tech giants such as Microsoft Corporation, Alphabet Inc. Class C, and Meta lead the way in the construction of ultra-large-scale AI data centers, driving the accelerated expansion of advanced process AI chip production and CoWoS/3D advanced packaging capacity, as well as DRAM/NAND storage chip capacity expansion, the long-term bullish logic of the semiconductor equipment sector is becoming increasingly robust. According to Wall Street giants Morgan Stanley, Citigroup, Loop Capital, and Wedbush, the global investment wave in artificial intelligence infrastructure centered on AI computing power hardware is far from over; it is just the beginning. Under the unprecedented "AI inference computing power demand storm," driven by continuous demand until 2030, the total scale of this round of AI infrastructure investment wave is expected to reach $3 trillion to $4 trillion. In 2026, a wave of capacity expansion for chip manufacturers The research report released by Citigroup shows that the main investment strategy for chip stocks in 2026 is not simply "broadly bullish on semiconductors," but rather focuses specifically on the semiconductor equipment industry (i.e., related to WFE). Citigroup's semiconductor investment strategy is centered on the value chain of "chip manufacturing giants increasing capex to expand the WFE market, leading to orders/revenue/profit expansion for semiconductor equipment leaders," betting on the continued growth of the semiconductor equipment sector in 2026. Citigroup's latest estimates show a global WFE market size model of approximately $115 billion in 2026 (meaning a year-on-year increase of +10%, far above the average growth rate of the past decade), with the three major chip manufacturers (Taiwan Semiconductor Manufacturing Co., Ltd. Sponsored ADR, Samsung, and Intel) accounting for approximately 59% of this total. The research also emphasizes the growth trend observed, indicating that the overall WFE in 2026 will be "closer to" their bullish scenario of $126 billion. In particular, for the world's leading chip manufacturer Taiwan Semiconductor Manufacturing Co., Ltd. Sponsored ADR, Citigroup predicts a capex guidance range of $46-48 billion for 2026 and expects a significant upward revision for the full year (communication with institutional investors suggests it could reach $50 billion). Taiwan Semiconductor Manufacturing Co., Ltd. Sponsored ADR had already raised its capital expenditure range for 2025 from the previous range of $38-42 billion to $40-42 billion in their earnings call in October of last year. Citigroup believes that the chain of "computing power - storage - advanced chip manufacturing" driven by the AI infrastructure craze determines the strong stickiness of semiconductor equipment capex compared to any previous cycle: the massive demand for computing power for AI training/inference not only drives the demand for advanced logic chips but also significantly boosts the demand for high-end storage chips (especially HBM/enterprise SSD related). As the complexity of chip manufacturing processes increases, the equipment required for each wafer's "front-end advanced processes/step count" to increase, resulting in continuous demand and improved visibility of orders for equipment. Citigroup's WFE model breakdown assumptions explicitly indicate a +30% growth in NAND, +12% in DRAM, and +10% in Foundry/Logic (based on their model perspective), suggesting that 2026 will see a more balanced expansion in advanced chip manufacturing processes, rather than solely relying on the AI computing sector. This trend typically benefits semiconductor equipment leaders with broader coverage and a complete product portfolio (such as focusing on deposition/etching/cleaning/measurement/advanced packaging and other multiple processes). The latest industry outlook data released by the World Semiconductor Trade Statistics Organization (WSTS) indicates that the global semiconductor demand expansion momentum is expected to continue to be strong in 2026, with MCU and analog chips seeing strong recovery following the ongoing soft demand since the end of 2022. WSTS predicts that following a strong rebound in 2024, the global semiconductor market will grow by 22.5% in 2025, with a total value of $772.2 billion, higher than the outlook given by WSTS in the spring; in 2026, the total value of the semiconductor market is expected to significantly expand to $975.5 billion on top of the strong growth in 2025, approaching the $1 trillion market scale target for 2030 forecasted by SEMI, indicating a potential 26% year-on-year increase. According to WSTS, this consecutive two-year strong growth trend will mainly benefit from the strong momentum in the logic chip sector dominated by AI GPUs, as well as the continuous strong momentum in the storage sector dominated by HBM storage systems, DDR5 RDIMM, and enterprise-grade data center SSDs, with both sectors expected to achieve incredibly strong double-digit growth. This is due to the continued strong expansion demand in areas such as AI inference systems and cloud computing infrastructure. Lithography, etching, thin-film deposition, and advanced packaging equipment entering a period of surging demand The Citigroup analyst team mentioned that the "Phase 2 cycle" signifies a shift in valuation anchors from "valuation rebound" to "continuously rising profits": when the overall WFE market shifts from the benchmark scenario towards a bullish scenario, the profit elasticity of semiconductor equipment leaders may be even greater than their revenue elasticity (due to economies of scale, capacity utilization increases, and a higher proportion of high-end processes). Therefore, Citigroup chooses ASML Holding NV ADR, Lam Research Corporation, and Applied Materials' semiconductor equipment combination to express the perspective of a positive growth trajectory. ASML Holding NV ADR, Lam Research Corporation, and Applied Materials, these three semiconductor equipment giants, cover lithography, etching, thin-film deposition, and advanced packaging equipment, all of which are key semiconductor equipment sectors benefiting the most from the AI infrastructure craze and storage super cycle. The Netherlands-based lithography giant ASML Holding NV ADR's EUV lithography machines can be considered essential semiconductor equipment for the world's largest chip manufacturers like Taiwan Semiconductor Manufacturing Co., Ltd. Sponsored ADR and Samsung Electronics to create the essential core DRAM for cutting-edge AI applications such as ChatGPT, Claude, etc., under the unprecedented global AI boom since 2023. These machines are also crucial for the storage giants such as SK Hynix and Micron Technology, Inc. to develop HBM storage systems, data center enterprise-grade SSDs/DDR, and other core storage devices required during the ongoing "storage super cycle" until possibly reaching 2027. AI GPU/AI ASIC accelerators' performance leap heavily depends on advanced logic nodes (from 3nm to 2nm, or even more advanced nodes like 1.8nm, 1.6nm), for which the key layers have to be achieved using EUV or even High-NA EUV to achieve smaller line widths and higher yields. ASML Holding NV ADR's EUV/High-NA EUV equipment specifically targets the mass production demand for 3nm and sub-2nm logic and globally leading-performance DRAMs, making them bottleneck capital equipment for expanding advanced processes. Meanwhile, AI training/inference is also fueling growth on the "storage side": the HBM storage systems paired with AI GPU/AI ASICs require not only further miniaturization of DRAM manufacturing nodes but also demand for etching, thin-film deposition, CMP processes, and the crucial stacked packaging and interconnect links (TSV/hybrid bonding for advanced packaging), significantly increasing manufacturing steps and equipment density. According to Applied Materials' latest technical interpretation, the HBM manufacturing process requires approximately 19 additional material engineering steps compared to traditional DRAM, with approximately 75% of these steps covered by their most advanced semiconductor equipment. They have also launched a bonding system for advanced packaging/storage chip stacking, making HBM and advanced packaging equipment long-term strong growth vectors for the company. GAA (gate-all-around)/backside power delivery (BPD), and other new chip manufacturing node equipment, will be the core DRIVE driving the company's next strong growth. Compared to Applied Materials, Lam Research's advantage lies in the high aspect ratio (HAR) etching/deposition and related process capabilities necessary for advanced HBM storage, as well as expertise in 3D NAND/advanced DRAM structure and interconnect that highly depend on Lam's exclusive HAR processes.