Enhanced demand within the AI semiconductor sector is not only driving increased requirements for wafer fabrication, advanced packaging, and equipment but is also concurrently elevating the value of upstream toolchains, including Electronic Design Automation (EDA), Intellectual Property (IP), verification, and design services. Domestic EDA providers are expected to progress from the adoption of individual point tools towards the replacement of full-flow solutions. Companies possessing core algorithmic expertise, established customer validation, and ecosystem synergy capabilities are poised for sustained benefits. The analysis suggests domestic EDA stands to gain from both the semiconductor upcycle and the AI dividend, with relevant companies including Empyrean Technology Co.,Ltd. (SHE: 301269, Buy), Primarius Technologies Co.,Ltd. (SHA: 688206, Not Rated), and Semitronix Corporation (SHE: 301095, Not Rated). The core viewpoints are as follows:
Rising Complexity in AI Chip Design Benefits EDA Infrastructure
As the demand for large model training and inference continues to grow, the accelerated upgrade of components such as GPUs, AI ASICs, HBM, Chiplet, advanced packaging, and high-speed interconnects is driving a shift in chip design from single SoCs towards multi-dielet, heterogeneous integration, and system-level co-design. Compared to traditional chips, AI chips impose higher demands on performance, power, area, signal integrity, thermal management, and verification coverage, significantly increasing design iteration cycles and verification complexity. EDA tools are integral throughout the entire process, encompassing logic synthesis, simulation and verification, place and route, timing analysis, physical verification, power analysis, yield optimization, and advanced packaging design. IP cores provide reusable functional modules like CPUs, GPUs, NPUs, interfaces, and memory controllers. Together, they form the fundamental infrastructure for chip design. The analysis posits that the semiconductor upcycle, fueled by AI, elevates the value of the entire upstream toolchain.
Verification Emerges as Key Growth Driver for EDA, AI Amplifies Tool Usage
In advanced chip design, verification often consumes a substantial portion of R&D time due to the high certainty required before tape-out. Functional verification, formal verification, hardware emulation, timing sign-off, and physical verification necessitate repeated confirmation across numerous scenarios, constraints, and process corners. As complexity increases with advanced nodes, Chiplets, and AI chips, verification intensity and simulation demands continue to rise, reinforcing strong customer lock-in for core tools like VCS, Xcelium, JasperGold, Palladium, PrimeTime, and Calibre. Unlike general software, chip design tools require deep integration with foundry process technologies, IP models, customer methodologies, testbenches, and sign-off flows. Switching tools involves not just software migration but also impacts verification environments, script ecosystems, engineer expertise, and foundry certifications. The analysis suggests EDA's business model exhibits a strong "reverse compounding" effect: the longer a customer uses a platform and the deeper their process integration, the higher the switching cost. The integration of AI into EDA is unlikely to simply replace traditional tools. Instead, it may lead to higher tool invocation frequency and greater computational power consumption through automated test case generation, script writing, PPA (Power, Performance, Area) exploration, bug localization, and multi-round optimization, thereby strengthening the overall value proposition of EDA platforms.
Domestic EDA Poised for Breakthroughs in Mature Nodes, Manufacturing, and New Paradigms
The EDA landscape has long been dominated by overseas players like Synopsys, Cadence, and Siemens EDA, with their barriers entrenched in core algorithms, process adaptation, foundry certification, customer flow adoption, and engineer ecosystems. While domestic vendors face challenges in fully replacing these leaders with comprehensive full-flow tools for advanced nodes in the short term, they have established clearer paths for breakthroughs in areas such as mature nodes, analog full-flows, manufacturing-oriented EDA, device modeling, yield analysis, digital verification, and advanced packaging. Empyrean Technology Co.,Ltd. is making continuous progress in analog full-flows and certain digital backend tools. Primarius Technologies Co.,Ltd. focuses on device modeling and SPICE simulation. Semitronix Corporation approaches the market from test chips, yield analysis, and manufacturing data management. Concurrently, new design paradigms like AI + open-source EDA, the RISC-V ecosystem, Chiplet, 3DIC, and the Tau (τ) law provide domestic EDA vendors with entry windows distinct from traditional 2D full-flow tools. The analysis concludes that, driven by the confluence of rising AI chip complexity, supply chain autonomy, and post-Moore's Law design paradigm shifts, domestic EDA is expected to gradually evolve from the adoption of point tools towards the replacement of flow-level solutions. Companies with accumulated core algorithms, a foundation in customer validation, and ecosystem synergy capabilities are positioned for sustained benefits.
Key Investment Risks
The report highlights potential risks including intensifying industry competition and slower-than-expected advancement of AI technologies.
Comments