Synopsys Wins the AI Silicon Brain Drain

The reported move of Amit Gupta, Jeff Dyck, and Ryan Silk from Siemens EDA to Synopsys matters because this is more than a routine transfer of experienced executives and engineers. Together, the three represent a concentrated body of expertise in applying artificial intelligence to semiconductor design—one of the most strategically important areas in electronic design automation.

Gupta was a central figure in Siemens EDA’s AI strategy and custom-IC business. He founded Solido Design Automation, whose machine-learning technology addressed process variation and yield verification in analog and custom chips. Siemens acquired Solido in 2017, and Gupta subsequently led both its custom-IC activities and the development of the Fuse EDA AI system. Dyck, another former Solido leader, helped create and industrialize its machine-learning technology. Silk worked on the engineering implementation of Fuse, including AI-assisted debugging and workflow automation. Their value therefore lies not only in individual ability, but in years of shared experience turning AI research into dependable commercial EDA products.

That distinction is important. Chip design is an unusually difficult environment for AI. An agent can occasionally make a mistake without catastrophic consequences; a design tool cannot casually hallucinate an electrical property or incorrectly declare a billion-transistor chip ready for fabrication. An advanced mask set can cost tens of millions of dollars, while a respin can delay a product by months. Successful EDA AI consequently requires semiconductor knowledge, high-quality engineering data, rigorous validation and a deep understanding of how chip designers actually work. Teams possessing that combination are scarce.

For Synopsys, the hires could accelerate the transition from AI-enhanced tools to genuinely agentic engineering workflows. Synopsys has been developing agents that configure tools, analyze results and iterate through design or verification tasks. It has also introduced autonomous workflows with Microsoft and AMD, including a debug-closure workflow for which it reported substantial cycle-time reductions. Gupta, Dyck and Silk bring relevant experience in building precisely the connective layer between AI models, established EDA engines and engineers.

The move also complements Synopsys’ broader strategy. Following its acquisition of Ansys, Synopsys is attempting to connect semiconductor design with multiphysics analysis at the system level. Its expanded relationship with Nvidia adds accelerated computing and AI infrastructure. The former Siemens team could help Synopsys combine these assets into practical products rather than a collection of impressive—but separate—technologies. Gupta’s product and commercial experience may be particularly valuable in determining which autonomous capabilities customers will trust and purchase.

For Siemens, the departures would be a meaningful competitive setback, although not necessarily a fatal one. Fuse has been presented as the front end for AI capabilities across Siemens’ EDA portfolio and as a foundation for self-verifying agents. Losing several people closely associated with that platform risks delaying its roadmap, disrupting customer relationships and weakening continuity between its custom-IC and central-AI organizations. The timing is especially sensitive because Synopsys, Cadence and Siemens are all racing to define the operating layer through which engineers will use future design tools.

The broader implication is that EDA competition is shifting from ownership of individual point tools toward control of AI-driven workflows. Customers rarely replace proven signoff products quickly, but an effective agent can influence which tools get invoked, how often they are used and where design knowledge accumulates. In that sense, the AI orchestration layer could become a new strategic control point.

Bottom line: I spent 10+ years with Amit, Jeff, and Ryan at Solido so I know this by first hand experience. The three hires will not decide the EDA market. Siemens retains valuable products, engineers and customer relationships, while AI adoption remains constrained by accuracy, security and data-access concerns. Nevertheless, the movement of an intact, commercially experienced AI team can compress years of organizational learning. If Synopsys gives the group sufficient authority and integrates it effectively, the result could alter not simply which company has better AI features, but which company defines how the next generation of chips is designed, absolutely.

Also Read:

Synopsys Cloud at DAC 2026

NVIDIA Brings Agentic AI Toolkit to EDA

Synopsys and Intel Foundry Enable System-Level Design on Intel 14A