Semiconductor Industry Collaboration Unveils Open Ontology for AI-Powered EDA Workflows
New repository supports development of autonomous design systems through shared standards, validated use cases and reasoning frameworks.
The Silicon Integration Initiative (Si2) AI/ML Schema/Ontology Working Group announced the public release of its AI for EDA Ontology Repository, providing the semiconductor design community with a strong example of ontology and use cases to enable EDA agentic system experimentation for broader industry progress and collaboration.
As agentic AI systems continue to emerge, the semiconductor industry is reaching a critical inflexion point—evolving from point tools toward autonomous workflows. Agentic EDA enables multi-step reasoning systems across design flows and requires context awareness, causal understanding, and workflow orchestration.
An Ontology in EDA is a formal representation of:
- Concepts (e.g. nets, cells, timing paths)
- Relationships (e.g., “affects,” “depends on,” “constrains”)
- Rules and constraints
Within the EDA domain, Ontology provides the reasoning backbone for agentic systems by capturing:
- Workflow steps and dependencies
- Cause-and-effect relationships
- Design tradeoffs
- Domain vocabulary and semantics
The public repository is available under the Apache 2.0 open-source license and includes:
- The EDA ontology (TTL/OWL)
- Documented and validated use cases developed by Arizona State University and Drexel University
- Ontology validation collateral
- An MCP server supporting agent discovery and reasoning
- Full documentation package
The Si2 AI/ML in EDA Schema/Ontology Working Group is a collaborative effort involving Siemens EDA, Arizona State University, Drexel University, IBM, NC State University, NXP, Qualcomm, and Synopsys.
By making the ontology publicly available, the working group aims to foster industry-wide collaboration, accelerate innovation in agentic EDA systems, and establish a common semantic foundation for AI-enabled semiconductor design workflows.
