The recent launch of OpenAI's next-generation flagship model, GPT-6 Astra, has been touted as the company's most intelligent and alignment-focused creation to date. Notably, this model played an integral role in designing OpenAI's proprietary inference chip, Jalapeño, establishing a powerful "AI designs chips, chips run AI" feedback loop. This raises a key question: what does this deep involvement of AI in semiconductor design and its ability to slash R&D timelines signal for the broader industry?
The transformative impact of AI on chip design can be examined through three primary dimensions that are reshaping the industry's landscape.
1. The Full-Scale AI Transformation of EDA Tools Boosts Industry Efficiency
At the 2026 IDAS Design Automation Industry Summit in early September, the launch of China's first Agentic EDA intelligent-agent strategy system and the debut of the country's inaugural 3D chip physical design tool marked a significant milestone. These innovations integrate AI capabilities deeply into the entire chip design workflow. EDA, often regarded as the "machine tool" of chip design, is the critical software backbone supporting the modern semiconductor industry. The core value of AI-enhanced EDA lies in establishing a complete operational loop that encompasses verification goals, expert engineering methodologies, real-time tool feedback, and traceable evidence chains, enabling autonomous iterative optimization. This model has the potential to dramatically enhance complex chip verification efficiency, building a robust technical foundation for the domestic chip sector to overcome intricate development bottlenecks and achieve significant performance leaps.
2. Diverse Downstream Demand Expands the Chip Design Landscape
The explosive growth of AI large models and AI servers is currently driving a rapid surge in demand for high-end chips. Gartner projects that revenue from the AI data center ecosystem will rise from 36.5% of global semiconductor revenue in 2026 to more than 53% by 2030, with value creation shifting from "standalone accelerator chips" to a "full-stack infrastructure" approach. Beyond AI, demand for chip design is flourishing across multiple sectors. Analysts predict that the fastest-growing areas for advanced process node chips will include cloud-based AI training and inference, high-performance computing, data centers, and high-speed networking, as well as AI-enabled smartphones, AI PCs, smart glasses, autonomous driving, robotics, and industrial intelligence. According to the World Semiconductor Trade Statistics (WSTS), global semiconductor sales are expected to reach $1.655 trillion in 2026, a year-over-year increase of 108%.
3. AI and Chip Development Enter a Positive Feedback Cycle of Continuous Evolution
AI is progressively evolving into a "virtual engineer" throughout the chip design process, a trend that can be summarized by two mutually reinforcing paths: "Design for AI" and "AI for Design." The former leverages EDA, IP, and system analysis capabilities to help customers construct more powerful AI infrastructure. The latter deeply embeds AI into the design workflow, using intelligent agents to redefine R&D efficiency and accelerate product innovation. In this dynamic, AI generates more powerful computing capacity, which in turn revolutionizes engineering processes, creating a self-accelerating virtuous cycle.
So, how can investors capture the opportunities emerging in chip design?
In the first half of this year, rapid advancements in AI, computing power, and electric vehicles have continued to expand demand within the semiconductor industry, lifting overall sector momentum. Wind data reveals that constituent stocks of the Shanghai Stock Exchange Sci-Tech Innovation Chip Design Thematic Index (950162.CSI) reported a 59% year-on-year increase in operating revenue and a 282% surge in net profit attributable to parent companies during this period. Analysts note that the domestic chip industry in H1 2026 is characterized by "high overall growth, structural divergence, and a pronounced contribution from manufacturing and memory." Unlike previous recoveries driven by consumer electronics, this upward cycle is primarily fueled by AI servers, data centers, and high-performance computing demand, leading to synchronized improvements across memory, logic, packaging, testing, and equipment materials. Looking ahead, the industry's growth drivers in 2027 are likely to shift from "price increases and supply-demand mismatches" to "real demand, domestic substitution, and technological upgrades." The twin forces of the AI dividend and import substitution are expected to jointly support the sector's medium-term upward trajectory.
From the perspective of leading international institutions, the resilience of AI capital expenditure remains the central theme across the supply chain. Custom AI chips and Agentic AI are opening up new incremental opportunities. As AI chip customization deepens, design complexity rises, exacerbating the shortage of skilled engineers. EDA companies are well-positioned to leverage Agentic AI to automate parts of the design, verification, and optimization workflow, converting efficiency gains into commercial value. By 2030, Agentic AI could generate an additional market opportunity of approximately $3.7 billion annually for the EDA industry.
In summary, AI's empowerment of the semiconductor sector underscores a long-term trend of efficiency improvements, demand expansion, and technological iteration within chip design. The Sci-Tech Innovation chip design track demonstrates clear long-term growth potential. The Sci-Tech Innovation Chip Design ETF managed by Yinhua Fund (589350), which tracks the Shanghai Stock Exchange Sci-Tech Innovation Chip Design Thematic Index (950162.CSI), selects listed company securities involved in chip design on the STAR Market as its index samples. This vehicle may assist investors in capturing the benefits of the industry's sustained growth.
Risk Disclosure: Investment involves risk. Investors should carefully read the fund contract, fund prospectus, and fund product information summary before making any investment decisions. Past performance does not indicate future results.