AI-Driven EDA Industry M&A Wave: Dual-Directional Boost for Quality and Efficiency

Securities Times reporter: Ruan Renshen

With the driving force of artificial intelligence (AI), the EDA (electronic design automation) industry is witnessing a merger and acquisition wave. Moving towards AI, embracing AI, has become a common choice for domestic and foreign EDA companies.

The largest M&A deal in the EDA industry

Recently, the State Market Supervision Administration conditionally approved Synopsys' acquisition of Ansys, a transaction valued at $350 billion. This has become the largest M&A deal in the history of the EDA industry.

This transaction is expected to bring new opportunities for Synopsys. According to data, Ansys specializes in engineering simulation software and covers not only IC design-related software but also simulates models for automobiles, aerospace, and other fields, with a market share of up to 42% in the simulation software industry. After the acquisition, Synopsys' potential market scale will increase by 1.5 times, meeting customers' needs for circuit and physical fusion under the driving force of AI.

"The design process of EDA includes design, simulation, verification, etc. Among them, the simulation step is highly dependent on algorithms, which have a natural connection with AI and can be easily empowered by AI to improve efficiency."

According to Chip and Semiconductor Vice President Wang Wei, as AI-driven hardware becomes more powerful, the scope of simulation-related fields will continue to expand.

After this transaction, Synopsys will also expand its traditional EDA industry "ability radius" by migrating simulation capabilities to high-end manufacturing industries such as automobiles and aerospace. In addition to Synopsys, other EDA giants are also showing similar expansion trends.

In March 2024, Cadence announced a $12.4 billion acquisition of BETA CAE Systems International AG, accelerating its intelligent system design strategy; in October 2024, Siemens EDA announced a $106 billion acquisition of Altair Engineering. According to reports, Altair's simulation technology covers fields such as mechanical, fluid, electromagnetic, and thermal management for automotive electronic systems.

With the driving force of AI, various terminal intelligent systems are becoming increasingly complex, and the demand for product design accuracy and high efficiency is no longer limited to the semiconductor industry. Statistics also show that EDA companies' M&A landscape has shifted from basic PCB design to overall electronic system solutions, covering non-semiconductor customers; at the same time, simulation and modeling capabilities have expanded significantly, highlighting the need for systematic solution capabilities.

From AI to AI

Several domestic EDA company insiders told reporters that under the driving force of AI, the importance of EDA system solution capabilities is increasing.

According to Wang Wei, EDA companies used to focus on design process optimization (DTCO), emphasizing software and manufacturing process integration. Now, with Moore's Law slowing down and advanced process development facing external barriers, the need for system technology joint optimization (STCO) has emerged, aiming to achieve "chip-to-package-to-system" co-optimization from a systems perspective.

This trend is driving EDA companies to continuously leverage AI to enhance their system capabilities, which will further be applied in critical industries such as high-end algorithm chips.

"The relationship between AI and EDA is a 'from-AI-to-AI' process." Wang Wei said. How to overcome the bottlenecks between various factors such as algorithm, storage, computing power, and energy consumption in the intelligent calculation system, ensuring high-efficiency output from AI data centers, poses a challenge for EDA companies.

According to He Jie, Vice President of Huawei's HiSilicon division, due to the increasing complexity of artificial intelligence and supercomputing chip architectures, as well as the rising costs of advanced process development, large-scale complex chip design validation processes are accelerating their entry into the early stages of chip development, aiming to shorten the R&D cycle, reduce risks, and improve chip quality. Additionally, AI computing typically involves collaborative work among multiple chips and nodes, with data transmission efficiency directly impacting overall performance; considering power consumption requirements is enormous, requiring continuous iteration on power management and heat dissipation.

Looking at the industry development trend, current EDA giants are following Moore's Law while researching more advanced processes, while also exploring faster interconnects for AI large-scale training and inference to provide stronger computing support. Among them, HBM (high-bandwidth memory) is a key technology for high-end algorithm chips, requiring the use of advanced 3D stacking packaging technology to integrate multiple memory chips vertically, thereby enhancing memory bandwidth and data transmission speed.

Industry insiders are calling for domestic industries to dare to try new schemes. Additionally, as typical application-oriented companies, EDA companies' healthy development depends on feedback and recommendations from downstream companies during the application process. The EDA industry faces a key challenge in data closure issues. Because EDA companies' data rely on terminal customer feedback, but design data is difficult to obtain, while AI tools lacking sufficient training materials will affect model reliability and cross-company and cross-domain usability.

AI-driven EDA

Despite facing multiple challenges and difficulties, the relationship between artificial intelligence and the EDA industry has already started to form a "dual-directional boost" for quality and efficiency. On one hand, EDA companies are upgrading their AI capabilities from chip design to system-level; on the other hand, AI is empowering EDA companies, significantly improving quality and efficiency.

"In the era of intelligence-driven innovation, artificial intelligence and the EDA industry will mutually catalyze and drive each other." Wu Xiaozhong said. International EDA companies are deploying AI-driven EDA fusion, focusing on three directions: AI-driven EDA tools, generative AI models for intelligent chip design, and innovative architecture combining AI and digital twin.

In April last year, Synopsys released "Synopsys.ai", considered the industry's first full-stack AI-driven EDA toolset, covering chip design from architecture to manufacturing; Cadence also proposed a generative AI solution for big data analysis, simplifying chip design workflows, and providing collaborative functions for simulation and prototyping design; Siemens EDA has multiple AI technologies applied to yield improvement, library building, digital verification, etc.

On the other hand, domestic EDA companies are also actively exploring and promoting AI applications in various scenarios.