Industrial: Ten major industry trends worth attention in 2026

date
11:38 18/02/2026
avatar
GMT Eight
As we enter 2026, we closely follow the iteration of industry trends and policy guidance. Based on the initial assessment, we update and supplement, and launch the top ten industry trends for 2026.
In early 2025, we focused on the upgrading of traditional industries and the cultivation of new quality productivity, proposing ten major industry trends worth watching. Looking back, most of these directions have become the core investment themes throughout 2025. As we step into 2026, we closely follow the iteration of industry trends and policy guidance, updating and supplementing our earlier judgments, and releasing the ten major industry trends for 2026 for the reference of all investors: 1. AI Applications: The global AI competition is in full swing, with model iteration driving deepening application scenarios. In early 2025, after major global giants significantly increased capital expenditure, the market's focus on AI shifted to whether the huge capital expenditure by tech giants could truly be transformed into commercialized applications. After the concentrated pricing of the "AI bubble" in late 2025, with the intensive iteration and release of large models domestically and internationally, breakthroughs were achieved in various application scenarios. In 2026, the global AI competition is still in full swing, and the iteration of model capabilities is expected to continue to drive the deepening of AI application scenarios, with the landing of popular applications being highly anticipated. Overseas, the competition landscape for AI applications has entered a new stage of multipolarity, with multiple disruptive models being launched, creating demonstration effects domestically. The overseas AI application competition landscape is transitioning from dominance by OpenAI to multipolarity. ChatGPT continues to dominate the generalized traffic core entry, occupying a large market share in areas such as chat, translation, and image generation, continuing to lead globally. In September 2025, after the launch of the video generation model Sora by OpenAI, it quickly reached the top of the US App Store's free list in its first week, further consolidating its ecosystem advantage. However, with more major players intensifying their efforts, the AI application competition landscape has clearly shifted towards multipolarity: Google is rapidly catching up with the GeminiPro series, XAI (Grok) is rising with real-time data and differentiated experiences, Meta, Google, Anthropic, etc., integrating AI with social, search, and productivity tools to build closed-loop ecosystems. For example, in February 2026, Anthropic PBC launched a legal AI plugin, sparking widespread discussions in the market about the substitution effects of AI in sales, marketing, financial software, and other areas. Rising to balance the closed-loop ecosystems of giants are rapidly rising open-source forces. In January 2026, OpenClaw became popular with an open-source model, with its token calls on the OpenRouter platform ranking first as of February 15, 2026. At the same time, AIAgent is moving towards autonomous intelligence, with OpenAI releasing its first AI intelligent body product, Operator, in January 2025, and Anthropic launching ClaudeOpus4.6 and AgentTeams experimental features in February 2026, expected to accelerate the commercialization of vertical high-value AI Agent scenarios. In 2026, amidst the competition among overseas AI giants, more disruptive large models are expected to be launched, continuing to create demonstration effects for the development of AI applications domestically. Domestically, AI applications are reaching a tipping point, attracting more tech giants to enter the market, and model iteration and scenario penetration are expected to resonate. 2025 was a critical year for the iteration of AI models and the landing of applications in China. DeepSeek, as the dark horse of the AI field in 2025, completed multiple rounds of model iteration. After the open source of DeepSeek-R1 and V3.2 series, leading models like Ali Tongyi Thousand Questions and MiniMax accelerated their open source efforts, leading the way in AI application landing. Among them, Ali's Qwen3 series, with its 256K long context, million-token extension capabilities and cost advantages, has become the global benchmark for open source. Additionally, in 2025, the focus of AI competition in China shifted from pure model parameter competition to application landing capabilities. Domestic AI application giants have successively secured their positions through core business operations, with ByteDance upgrading to a cross-ecosystem intelligent entity and integrating with Douyin and Today's Headlines, Ali's Tongyi Thousand Questions integrated into the e-commerce and Dingding ecosystem, and Tencent's Yuanbao deeply embedded in the WeChat and QQ social chains. Domestic enterprises are accelerating the layout of AI applications for consumers and industries. Ali has initially built an entire chain of "AI ordering, AI ticket booking, AI customer service," Tencent's Yuanbao Spring Festival red envelope activity has driven user growth, ByteDance has "Douyin" as the core consumer entry point, and in February 2026 released the new generation video generation model Seedance 2.0, achieving a breakthrough in AI video from generation to commercial use. Looking ahead to 2026, domestic AI applications are expected to transition from model implementation to scenario monetization, from conceptual narratives to performance realization, and model iteration and scenario penetration are expected to resonate. In terms of configuration, we recommend focusing on: 1) Continuous iteration of multimodal models, expected to continue to empower the fields of film and television, games, and advertising, among others. 2) As competition for universal access reaches saturation, vertical high-value AI Agent scenarios are expected to accelerate commercialization, with industries characterized by clear rules and high labor costs benefiting first, including medical, financial, legal, education, tax, and government sectors. 2. AI Computing Power: Overseas capital expenditure maintains high growth, accelerating domestic substitution Overseas, the 2026 capital expenditure of tech giants is expected to see high growth, with new technologies and demands accelerating the industry chain. In 2025, the capital expenditure of major cloud service providers overseas significantly increased, with the combined capital expenditure of the top four cloud service providers in North America totaling $359.2 billion. In 2026, the combined capital expenditure guidance for Amazon, Google, Meta, and Microsoft was approximately $598.7 billion, with a projected year-on-year increase of 67%, reflecting the continued demand for AI computing power under the logic of an arms race. Facing the explosive growth in AI computing power demand worldwide and the need for high-speed data transmission, new technologies and demands in the industry chain are accelerating. On the one hand, with data centers scaling up to ultra-large clusters, 1.6T and even 3.2T high-speed optical modules are being rapidly commercialized, and cutting-edge technologies such as silicon photonics, CPO, etc., are continually breaking through. On the other hand, the demand for massive computing power is driving data centers into a phase of rapid expansion, leading to a widening gap in key infrastructure supply such as optical fibers and liquid cooling, as well as high demand across various segments of the industry chain. Domestically, leading tech companies are increasing their capital expenditure and accelerating the iteration of large domestic models, driving continuous improvements in the performance of domestic chips and building a self-controllable domestic substitution ecosystem. As the core engine of computing power, chips in China have traditionally relied on foreign chips like NVIDIA's A800 and H800 for training. In 2025, the US expanded export controls on high-end AI chips, accelerating the need for domestic substitution. Leading companies such as Huawei's Ascend, Cambrian, and Hygon Information Technology continue to increase capital expenditure and research and development investment, improving their market share and performance. At the same time, the deep integration of self-developed chips with large domestic models is driving optimization of computing power architecture. For example, Huawei's Ascend chip and the DeepSeek series of models have completed deep integration, Cambrian's collaboration with SenseTime's SEKO series of multi-modal models, tailored for commercialization. As AI penetrates into end-to-end scenarios, domestic telecom operators and cloud service companies are expected to further increase their capital expenditure, deepening the substitution of the chip industry chain, with domestic computing power companies and upstream semiconductor equipment materials offering ample growth opportunities. In terms of configuration, we recommend focusing on: 1) The ongoing surge in global demand for computing power presents a significant opportunity for Chinese companies in the optical module industry, with domestic manufacturers occupying a major share of the global market and vast room for growth. Focus on leading domestic optical module suppliers; as optical communication advances to higher data rates, high-speed optical modules, optical fibers, CPO, liquid cooling, and copper connections are expected to continue to benefit. 2) The market share of self-developed chips and domestic devices in the inference market is expected to accelerate, so continue to monitor leading domestic AI chip companies. 3. Storage: AI-driven super cycle commences, supply-demand situation continues The demand for AI training and inference is surging, driving the storage industry into a new super cycle. Historically, the storage industry has exhibited strong cyclical characteristics due to supply-demand mismatch, driven by new computing platforms. In 2025, the surge in AI computing power demand emerged as the main driver of increased storage demands, fundamentally transforming the supply-demand dynamics in the storage sector. With traditional server's demand for DRAM and NAND being 8-10 times that of conventional servers, AI servers in 2025 accounted for 53% of the global monthly memory production capacity, leading to severe shortages of high-end storage components such as DDR5 and HBM. By the fourth quarter of 2025, global storage prices had risen by over 40%, and the supply-demand imbalance may continue to push storage prices higher in 2026. On the supply side, production capacities remain constrained, and the supply-demand situation continues, affecting upstream equipment and testing stages. In light of high demand, manufacturers are actively reducing traditional production capacities. International companies like Samsung, SK Hynix, Micron, etc., are phasing out DDR4 products to focus on DDR5 and HBM, leading to a significant shortage of consumer-grade storage components. While domestic storage manufacturers are ramping up production, the long lead time for fab construction and equipment delivery makes it difficult to alleviate these short-term shortages. For example, the delivery time for consumer-grade memory equipment is approximately 4-9 months, with new production capacities expected to gradually come online in the second half of 2026. However, the enduring demand driven by AI-powered data centers could lead to sustained shortages. Currently, China's Changxin Storage and Yangtze Storage have initiated the largest expansion plans in their history, with new facilities planned for operation in Shanghai and Wuhan by 2027. Overall, the storage industry in 2026 may struggle to achieve comprehensive and effective supply growth. This trend is expected to propagate upstream to semiconductor equipment and testing stages, benefiting domestic storage chip designers, module manufacturers, semiconductor equipment companies, and testing enterprises. In terms of configuration, we recommend focusing on: 1) In light of continued tight supply-demand dynamics, the buoyant AI market, and the protracted expansion timelines, leading domestic storage companies are expected to demonstrate sustained prospects; 2) The super cycle in the storage industry will drive a wave of expansion, with room for growth in upstream equipment and testing sectors. 4. Commercial Aerospace: A battleground for US-China rivalry, critical core technologies achieved domestically Commercial aerospace, as a battleground for US-China rivalry, has become strategically important, with a growing focus, high-level emphasis, and continuous policy support. Satellite orbits and frequency spectrum resources are non-renewable and scarce strategic assets with significant dual-use applications for military and civilian purposes. Since 2025, commercial aerospace has become a key arena for technological competition between the US and China. Leveraging their technical advantages, US companies like SpaceX and Blue Origin have built complete ecosystems covering rocket launches, satellite internet services, and deep space exploration. In December 2025, seeking to ensure US space superiority, the US government issued the "Ensuring US Space Superiority" executive order. In China, commercial aerospace has been prominently featured in the government work reports for two consecutive years, with the 2025 report explicitly stating the need to accelerate the development of the commercial aerospace industry. The 3rd plenary session of the 20th Central Committee of the Communist Party of China's report articulates the need to "promote the high-quality and safe development of commercial aerospace." In November 2025, a special policy for commercial aerospace was announced, titled "Action Plan to Promote the High-Quality and Safe Development of Commercial Aerospace (2025-2027)," which includes 22 key initiatives, the establishment of a national commercial Addsino Co.,Ltd. fund with an initial scale of 20 billion RMB, and the promotion of research facility access, government procurement support, and local subsidies. Supporting policies in the financial markets are also reinforcing this trend, with expedited IPO approvals for commercial aerospace enterprises and advantages businesses poised to accelerate their entry into the capital markets. China has introduced an expedited approval process for commercial aerospace IPOs, with over 10 companies initiating the process by February 15, 2026. Companies like Blue Arrow Aerospace, Star Engines, and China Aerospace Science and Technology Corporation have made significant progress towards actual IPOs. Additionally, in December 2025, the Shanghai Stock Exchange issued "Guidelines for Application of Issuance and Listing Review Rules No. 9," providing detailed regulations on the application of the Science and Technology Innovation Board's fifth set of listing standards for commercial rocket companies. This will better support high-quality commercial rocket companies that have not yet achieved a certain level of revenue, accelerating the innovation and development of commercial aerospace, and actively serving China's aerospace power strategy. Domestically, the pace of manufacturing by major aerospace players is gradually taking shape, with deep participation in the supply chain. Chinese companies have captured significant market share in terms of commercial aerospace shipments over the years, with ShenZhen New Industries Biomedical Engineering leading the market. In 2025, UBTECH ROBOTICS launched the full-size industrial humanoid robot Walker S2, securing multiple billion-dollar deals throughout the year, totaling over 1.3 billion RMB in orders. Shenzhen New Industries Biomedical Engineering initiated the first batch of commercial deliveries for the Thousand Questions series, surpassing a 200 million RMB purchase agreement with Ningbo Joyson Electronic Corp. DOBOT has entered the intelligent robot race with its Dobot Atom model, offering a "nimble operation + straight knee walking" feature, which attracted significant global orders. Chinese suppliers are deeply involved in the commercial aerospace supply chain, catering to both domestic and overseas markets. In terms of configuration, we recommend focusing on: 1) Chinese domestic aerospace manufacturers leveraging a low-cost, comprehensive domestic parts supply chain to accelerate large-scale production, enhancing customer engagement with scenario-based solutions; 2) 2026 marks the debut of commercial aerospace production, making core, technology-agnostic upstream components the primary beneficiaries. As demand for commercial aerospace goods grows abroad, areas like ball screws, reducers, electro-motors, sensors, and others are anticipated to realize significant gains. This translation has drawn insights from multiple complex and technical areas, each with specific industry terminology and concepts. For the most accurate understanding, it's recommended to refer to the original text in Chinese for a detailed grasp of the content.