Cadence Expands Sovereign AI in Europe: From Chips to AI Factories

3 Nov 2026 • 5 minute read

European engineering teams are advancing increasingly complex chip and system design while building the infrastructure to run AI at scale. Both depend on substantial computing resources. Both involve sensitive intellectual property. And both require greater control over where engineering data is stored, processed, and managed.

To bring computing scale and regional data control together, Cadence is expanding its sovereign AI data center in Munich, becoming the first electronic design automation (EDA) vendor to deliver a comprehensive sovereign AI platform for semiconductor, system, and AI factory design in Europe.

The deployment combines European-hosted Cadence OnCloud services with an expansion roadmap that includes NVIDIA Vera and Grace CPU systems, the Cadence Millennium M2000 AI Supercomputer powered by the NVIDIA Blackwell platform, and Cadence Reality Digital Twin Platform instances. The platform enables customers to design chips, advanced packaging, systems, physical AI applications, and AI infrastructure while maintaining data residency within European jurisdiction.

Sovereignty Begins with Engineering Data

Data residency is one part of the sovereign foundation, and sovereign AI extends this principle to the development and operation of AI capabilities within a defined jurisdiction, with greater control over infrastructure, data, and applicable governance requirements.

For semiconductor and engineering organizations, this control must extend beyond AI models to the proprietary information used to create future products: chip architectures, advanced packaging designs, simulation results, thermal and mechanical models, and customer-specific system configurations.

Moving these assets to the cloud requires confidence in access management, identity controls, encryption, auditability, and the location of compute and storage resources. Cadence OnCloud in Munich provides European teams with a hosted design environment built around these requirements.

Connecting Compute-Intensive Workflows from Silicon to Systems

Alongside control over engineering data, teams need computing capacity to support increasingly large, interconnected electronic and system design workloads. Modern engineering programs increasingly combine EDA with 3D-IC and advanced packaging exploration, signal and power integrity analysis, thermal simulation, computational fluid dynamics (CFD), electromagnetic analysis, and AI-enabled optimization. AI factory projects add another layer of complexity, with design decisions tied to power delivery, cooling capacity, airflow, rack density, and operational resilience.

This convergence increases the demand for scalable compute across design exploration and simulation. NVIDIA accelerated computing and NVIDIA CUDA-X libraries power Cadence’s design and simulation software, helping engineering teams evaluate more design options and tackle compute-intensive workloads across semiconductor design, systems engineering, and AI infrastructure.

The planned Munich expansion combines NVIDIA Grace and Vera CPU systems with Millennium M2000 capacity powered by NVIDIA Blackwell, providing CPU and GPU resources for demanding design, simulation, and AI workloads while keeping engineering data within European jurisdiction.

Digital Twins for AI Factory Design and Validation

AI factories tightly integrate compute, networking, storage, power distribution, and cooling. Engineers must assess rack density, equipment placement, airflow, and heat removal alongside power capacity, redundancy, serviceability, and operational resilience. Validating these interactions under realistic operating conditions before installation helps identify constraints and refine configurations before deployment.

The Cadence Reality Digital Twin Platform supports this systems-level assessment by modeling power delivery, cooling, airflow, thermal behavior, and equipment placement before physical installation. Working alongside the NVIDIA AI stack, including NVIDIA Omniverse libraries, it helps teams compare configurations against physical and operational constraints. It also supports the NVIDIA Omniverse DSX Blueprint for incorporating applicable blueprint capabilities into AI factory design and operational workflows.

One important capability is the Cadence Reality DC Elements model for the Cadence Millennium M2000 AI Supercomputer powered by NVIDIA Blackwell GPUs. This validated digital representation helps customers evaluate AI factory and data center environments before installation, lowering deployment risk and accelerating time to operation.

In a recent deployment, the Cadence Reality Digital Twin Platform helped increase data center capacity utilization from 71% to 89% and operational availability from 90% to 98%, enabling operators to extract more value from existing infrastructure while improving reliability.

Cadence also plans to use the Reality Digital Twin Platform within the Munich facility to improve energy efficiency and reduce operational uncertainty by connecting engineering compute infrastructure with simulation-driven decisions across planning, deployment, and operations.

Security Assurance and Collaboration Must Scale Together

The Munich deployment is certified to the ISO 27001 and ISO 27017 standards and SOC 2 Type II attestation, covering information security management, cloud service controls, and independent assurance over controls. For customers reviewing cloud-based engineering infrastructure, these certifications provide a baseline for assessing how the environment is governed and operated.

This assessment is particularly important when projects involve distributed teams, foundries, suppliers, ecosystem partners, and customers. Centralized cloud delivery can simplify access to shared engineering resources, but collaboration still requires clear permissions, project boundaries, logging, and policies for retaining and deleting data. The Munich deployment brings regional infrastructure and security assurance together to support this model of governed collaboration.

Implications for European Engineering Teams

For European customers, the Munich data center expansion creates a path to scale cloud-based design, simulation, AI, and digital twin workloads within a sovereign environment. The facility expands Cadence's European cloud footprint and can also support startups participating in the EuroCDP program, complementing existing European cloud resources, including leading public cloud provider infrastructure in Frankfurt.

Together, these capabilities reflect the convergence of semiconductor design and AI infrastructure engineering. Developing next-generation AI systems requires teams to design complex silicon and validate the physical environments that support it. Cadence’s European expansion brings sovereign cloud delivery, NVIDIA accelerated computing, and digital twins together to address both, from protecting design IP to assessing power, cooling, and deployment readiness.

Read the press release to learn how Cadence is collaborating with NVIDIA to expand sovereign AI infrastructure in Europe for semiconductor, system, and AI factory design.