Siemens Advances Self-verifying Agentic AI Workflows for Semiconductor and PCB Design

Estimated reading time: 4 minutes

Siemens announced an expansion of its strategic partnership with NVIDIA to deliver self-verifying agentic AI workflows to EDA, helping semiconductor and printed circuit board (PCB) engineering teams move from autonomous task orchestration toward more trusted, continuously validated engineering outcomes. The new capabilities build on Siemens’ recently introduced Fuse™ EDA AI Agent system and add new NVIDIA AI technology to help long-running, domain scoped, AI agents reason, act and continuously validate decisions against deterministic, physics-based EDA engines.

The updates to Siemens’ Fuse EDA AI Agent are designed to help the semiconductor and PCB industry move beyond simple automation and deliver trusted and verifiable outcomes that improve result quality, time-to-results, tool-calling reliability and token efficiency across long-running engineering workloads. The Fuse EDA AI Agent system is now integrated into Siemens’ recently launched Intelligence Center X, supporting agent creation and orchestration in Siemens’ enterprise industrial AI environment. Intelligence Center X delivers AI-driven, coordinated processes across design, manufacturing and supply chain – enhancing the comprehensive Digital Twin to support smarter execution and more trusted outcomes.

"Our expanded collaboration with NVIDIA enhances our domain-specific industrial AI, physics-based EDA engines and accelerated computing to create trusted, self-verifying AI workflows," said Amit Gupta, senior vice president and chief AI strategy officer, Siemens EDA, Siemens Digital Industries Software. “By enabling autonomous and long-running EDA AI agents to continuously validate their decisions against proven engineering tools accurately and efficiently, we help customers accelerate development, improve design quality and increase confidence in trusted engineering outcomes.”

“Semiconductor and PCB design are among the most complex engineering challenges in the world, and AI agents need trusted tools to reason, act and verify their work,” said Timothy Costa, vice president and general manager of computational engineering at NVIDIA. “Siemens is combining its EDA software with NVIDIA accelerated computing, AI software and Nemotron open models to create self-verifying agentic workflows that improve design quality, speed up results and give engineering teams greater confidence in every step of the process.”

Enabling trusted self-verifying AI agents for semiconductor and PCB design

The Fuse EDA AI Agent solution now combines Siemens engineering intelligence and trusted verification capabilities with NVIDIA AI infrastructure technologies to execute complex semiconductor design tasks with enhanced speed and precision:

Optimized EDA agents built with NVIDIA NeMo Gym open library for agentic environments improve result quality, speed and token efficiency for semiconductor and PCB design. This empowers customers with EDA design flows that get smarter over time with self-verifying agents which learn from every project to automatically improve execution strategies, optimize workflows and make better use of accumulated context.

The NVIDIA OpenShell secure runtime enables enterprise scale design teams to run autonomous agents across their entire EDA environment with enterprise-grade security, access controls and audit trails within governed runtime environments.

Advanced reasoning with the latest NVIDIA Nemotron models and Switchyard accelerates design with AI agents that understand complex engineering trade-offs at the speed of AI reasoning and supports complex engineering workflows with improved performance and token efficiency.

NVIDIA accelerated computing and CUDA-X libraries helps design teams achieve signoff-quality results in hours instead of days, powering both AI reasoning and EDA engines for dramatic speedups without sacrificing accuracy.

Together, these capabilities enable secure and token-efficient multi-agent coordination, real-time feedback and improved convergence across semiconductor and PCB design and verification flows, helping customers achieve trusted engineering outcomes.

Comprehensive end-to-end design automation

Building on these capabilities, Siemens’ Fuse EDA AI Agent orchestrates workflows across the semiconductor and PCB lifecycle. Domain-specific agents collaborate across synthesis, verification, implementation and validation processes to optimize outcomes holistically.

The new offering spans Siemens’ EDA technologies across the full semiconductor and electronics lifecycles, from high-level synthesis with Catapult, verification with Questa One and Veloce, and custom IC design and verification with Solido, through to physical implementation with Aprisa, signoff verification with Calibre and design-for-test with Tessent. It also supports advanced 3D IC integration using Innovator3D IC and PCB design with Xpedition.

Accelerating custom integrated circuit (IC) design with agentic AI

The expanded collaboration enhances Siemens’ Solido Characterization Suite with agentic AI capabilities. These agentic workflows automate library characterization by automatically generating and verifying Liberty files by running Solido Characterizer with Solido LibSPICE, Solido Generator and Solido Analytics. The solution delivers production-proven results for advanced-node standard cell, memory and custom IP libraries while reducing characterization turnaround times by more than 10X and achieving over a 5X to 10X reduction in token costs.

These agentic capabilities extend further into the custom IC design flow with the new Solido Layout Analyzer, an AI-powered measurement-aware solution for visualizing, exploring and analyzing parasitic and layout-dependent effects in post-layout designs. Integrated with Fuse EDA AI Agent, Solido Layout Analyzer dramatically boosts design quality and engineering productivity by enabling natural language prompting with results analysis, fixing recommendations, and report generation.

“For STMicroelectronics’ non-volatile memory team, the ability to relate layout insight directly to electrical behavior is critical. Rather than discovering layout dependent effects late in the process, Solido Layout Analyzer brings this analysis earlier into the flow, improving design confidence and helping cut down the time we spend debugging complex design blocks by weeks. We are planning to test and validate the advantages in our on-going design activity,” said Gianbattista Lo Giudice, Non-Volatile Memory Design manager, STMicroelectronics.

Accelerating digital verification with agentic AI

With verification consuming up to 70 percent of design time and complexity, the industry faces an unprecedented productivity crisis. Building on the recently announced Questa One Agentic Toolkit, Siemens is extending these domain-scoped agentic AI workflows with NVIDIA's Nemotron 3 Ultra reasoning model, designed for complex, long-running agents. In agentic RTL benchmarking with the ACE-RTL agent, Nemotron 3 Ultra leads among open models. Its reasoning and efficiency help autonomous agents evaluate engineering trade-offs while continuously validating designs against golden test harness. This enables teams to identify issues earlier, accelerate verification closure, and deliver trusted outcomes faster.

"We're at an inflection point where the complexity of AI chips, chiplets, and 3D ICs has outpaced traditional verification methodologies,” said Abhi Kolpekwar, senior vice president and general manager Digital Verification Technologies, Siemens EDA, Siemens Digital Industries Software, “agentic AI is the natural path forward that scales with this complexity—agents that can orchestrate multi-domain verification, reason across billions of test scenarios, and deliver production-quality results in timeframes that were impossible just two years ago."