Synopsys Sees AI Transforming Chip Design as Semiconductor-Specific Models Take Shape
Artificial intelligence is becoming one of the biggest forces reshaping the semiconductor design industry, and Synopsys believes the shift could dramatically accelerate electronic design automation, better known as EDA.
Shankar Krishnamoorthy, Synopsys’ chief product development officer, has highlighted how AI is beginning to change the way engineers approach chip development. According to him, the company is working with OpenAI on semiconductor-specific AI models designed to support the highly complex workflows involved in modern chip creation.
The goal is not simply to add generic AI tools into existing engineering software. Instead, Synopsys is focusing on models that understand the language, requirements, constraints, and technical depth of semiconductor design. This could make AI far more useful for engineers working on advanced processors, custom silicon, AI accelerators, and other next-generation chips.
EDA software plays a critical role in the semiconductor industry. It helps chipmakers design, test, verify, and optimize integrated circuits before they are manufactured. As chips become more advanced, the design process has grown increasingly complicated, requiring massive amounts of engineering time and computing power. AI has the potential to reduce that burden by automating repetitive tasks, improving design exploration, and helping engineers identify problems earlier in the development cycle.
Synopsys appears to see AI as a way to speed up the entire chip design process rather than replace human engineers. Semiconductor development still requires deep expertise, judgment, and creativity. However, AI could serve as a powerful assistant, helping teams move faster through tasks such as code generation, verification, documentation, layout optimization, and analysis of design trade-offs.
A major part of the strategy involves using different types of AI models for different kinds of work. Frontier models, which are large general-purpose AI systems, can be useful for broad reasoning, natural language interaction, and general productivity. At the same time, chip-specific models can be trained or optimized for semiconductor engineering tasks, making them better suited for technical workflows where accuracy and domain knowledge are essential.
This division of labor could become important as companies look for practical ways to integrate AI into professional engineering environments. General AI models may be helpful for communication and problem-solving, while specialized models can focus on the details of circuit design, verification, timing closure, power optimization, and manufacturing constraints.
For the semiconductor industry, the timing is significant. Demand for faster and more efficient chips continues to rise due to artificial intelligence, cloud computing, automotive technology, mobile devices, data centers, and high-performance computing. At the same time, designing advanced chips is becoming more expensive and time-consuming. If AI can help reduce development cycles, companies may be able to bring products to market faster while improving performance and efficiency.
Synopsys’ work with OpenAI points to a future where AI becomes deeply embedded in chip design platforms. Rather than acting as a standalone chatbot or external tool, AI could become part of the engineering environment itself, offering suggestions, running analysis, spotting issues, and helping teams make better decisions throughout the design process.
The impact could be especially meaningful for verification, one of the most demanding areas of chip development. Modern chips contain billions of transistors, and ensuring that every part functions correctly is a massive challenge. AI-assisted verification could help engineers detect bugs, generate test cases, and improve coverage more efficiently.
AI may also improve design space exploration, allowing engineers to compare more options in less time. Instead of manually testing a limited number of configurations, teams could use AI to explore many possible architectures, layouts, and performance targets. This may lead to better designs and more efficient use of resources.
While the industry is still in the early stages of adopting AI-driven EDA, Synopsys’ direction suggests that semiconductor design is entering a new era. The combination of advanced AI models and specialized engineering tools could reshape how chips are developed, tested, and optimized.
In the long run, AI may become a standard part of the semiconductor design workflow, much like traditional EDA tools are today. Companies that successfully combine domain expertise with powerful AI systems could gain a major advantage in speed, efficiency, and innovation.
Synopsys’ message is clear: AI is not just another software feature for chip design. It is becoming a major accelerator for the future of electronic design automation, with semiconductor-specific models playing a key role in helping engineers build the next generation of chips.
