Socionext Inc., a provider of custom system-on-chip (SoC) design and development services, is incorporating agentic AI into its SoC design processes and extending AI-based design methods to the physical design of large-scale SoCs using advanced process technologies. For its physical design initiatives, Socionext is collaborating with Cadence.
Applying agentic AI to SoC design and development
Socionext is integrating agentic AI into front-end design to support design and verification tasks, engineering analysis and decision-making. The company is also using AI agents within its design and development processes to apply accumulated design knowledge across engineering activities.
These efforts are intended to support engineering productivity and enable engineers to focus on design decisions and other development work.
Applying AI-based methods to SoC physical design
Socionext is extending its AI-based design work to physical design. Physical design for large-scale SoCs built with advanced process technologies involves multiple design constraints and requires coordination among electronic design automation (EDA) tools and design flows.
In collaboration with Cadence, Socionext is applying AI-based design methods to the physical design of a large-scale 3nm SoC. The company is using the Cadence InnoStack AI Super Agent and plans to evaluate its use in additional SoC development projects.
Integrating engineering knowledge into physical design workflows
Socionext is using retrieval-augmented generation (RAG), Model Context Protocol (MCP) support, customized AI skills and AI agents to incorporate internal design knowledge into physical design workflows. Current and planned applications include floorplanning, custom timing-analysis checks and checklist automation.
These applications are intended to make internal engineering knowledge more accessible and reusable across projects and teams.
Extending AI-based design across SoC development
In addition to automating selected tasks, Socionext is working to retain and reuse engineering knowledge across projects, teams and future development activities. The company intends to use this approach to support design-process changes while maintaining established quality and reliability requirements.
Socionext plans to connect its front-end design AI initiatives with its physical design capabilities to support a more integrated SoC development flow. The company will continue its collaboration with Cadence and evaluate the resulting capabilities in additional SoC development projects. It is also developing custom AI agents and engineering assistants that use internal expertise to support productivity and knowledge reuse.
For more information, visit cadence.com.
The post Socionext expands AI-based design across SoC development appeared first on Engineering.com.