Siemens today released Solido Characterizer, an AI-powered library characterization tool that accelerates the creation of SPICE-based Liberty files for semiconductor design, a critical step in chip development that has become a bottleneck as process nodes advance. By integrating predictive AI and a purpose-built simulator, the software reduces Liberty file generation from weeks to days, addressing growing complexity in process corners, tighter electrical margins, and emerging data formats such as LVF.
Technology Overview
Solido Characterizer delivers a 7x throughput improvement through two complementary innovations. Its enhanced AI engine achieves a 5x speedup for multi-PVT (process, voltage, temperature) creation and advanced LVF (library variation format) techniques. The industry’s first AI-accelerated characterization simulator, Solido LibSPICE, adds an additional 2x performance boost. Together, these capabilities ensure libraries are production-ready across all process nodes without compromising data quality or schedule predictability.
Manufacturing Implications
The tool enables teams to scale characterization across multiple IP blocks and design groups while maintaining accuracy and timeline consistency. Integration with Solido Analytics provides real-time QA insights, progress monitoring, and automated rerun capabilities, streamlining workflows. Solido Characterizer also works with Solido Generator, which uses baseline Liberty files to train AI models for generating additional library views without SPICE simulation, completing an end-to-end library creation flow. For advanced workflows, it can combine with Solido Fuse, built on the Fuse EDA AI system, to enable generative and agentic AI capabilities.
Market Context
Foundries and in-house design teams face mounting pressure from new process technologies that introduce tighter electrical margins and larger corner counts. Siemens reports that early adopters have validated the tool’s impact: GlobalFoundries achieved a 20–30 percent speedup in internal flows while maintaining production accuracy correlated to SPICE models. Anatrix used the tool to characterize radiation-hardened digital gate libraries, with silicon testing confirming post-layout parasitic simulations accurately predicted IP behavior, supporting first-pass success.
As semiconductor design complexity continues to escalate, AI-driven characterization represents a fundamental shift in library development economics. Solido Characterizer positions engineering teams to meet rising demands for speed and precision across all process nodes, enabling faster time-to-market for increasingly complex designs. The integration of generative and agentic AI workflows further suggests that characterization will become a more automated, scalable function within the broader EDA ecosystem.
