Synopsys and NVIDIA Partnership Demonstrates Need for Accelerated GPU and AI Applications to Meet Simulation Demand

On December 1, 2025, Synopsys and NVIDIA announced an expansion of their strategic partnership to accelerate semiconductor engineering and design using NVIDIA Graphics Processing Units (GPUs). Synopsys will use NVIDIA CUDA to transfer more simulation and engineering workloads from Central Processing Units (CPUs) to GPUs, accelerating a decades-old transition, with additional intention toward developing Artificial Intelligence (AI) and digital twins. Synopsys will integrate AgentEngineer with NVIDIA’s Agentic AI solutions—namely NIM microservices, NeMo Agent Toolkit, and Nemotron models—to accelerate the development of Agentic AI, while leveraging NIVIDA Omniverse and Cosmos to build complex digital twins across industries to enhance design, testing, and validation throughout product lifecycles. Signaling the deepness of the commitment, NVIDIA also purchased US$2 billion of Synopsys common stock. The partnership underlines how the trend toward simulation-driven design is putting stress on existing simulation infrastructure, prompting providers to partner with GPU and AI companies to manage more complex and computationally expensive simulations.

Simulation workflows are growing both in breadth and computing power—expanding from product to scenario and system simulation with higher granularity and accuracy. Meanwhile, increasing product complexity is forcing designers and engineers to lean heavily on simulation and digital twins to visualize, test, and validate product designs. Both customers and suppliers recognize the potential for AI to make simulation more efficient and for GPUs to cover entire workflows instead of individual solvers.

Vendors have strengthened partnerships with NVIDIA for similar purposes over the past 12 months. Siemens expanded its partnership to connect Siemens Xcelerator with NVIDIA Omniverse to build physics-based digital twins for manufacturing. Cadence works with NVIDIA to build digital twins with Omniverse and develop Agentic AI stacks with NVIDIA NeMo, all underpinned by its GPUs.

NVIDIA’s partnership with Synopsys is unique due to a direct US$2 billion investment—consistent with NVIDIA’s aggressive spending to expand its influence in the AI market and ensure demand for its hardware. The investment in Synopsys, in particular, signals NVIDIA’s belief in the comprehensive Electronic Design Automation (EDA)/Computer-Aided Design (CAD)/simulation solutions stack Synopsys is building after acquiring Ansys in July 2025 for approximately US$35 billion.

NVIDIA is the clear partner of choice to meet increasing simulation demand, and Synopsys competitors should (and likely will) pursue similar partnerships to provide the computing power needed for more complex simulations. In order to take full advantage of such partnerships, suppliers must:

• Invest in new High-Performance Computing (HPC) hardware. Current applications of GPUs are broad, but CPUs still account for a significant amount of simulation workflows. To apply GPUs to full workflows instead of just individual physics solvers, which is necessary as simulations become significantly more complex, suppliers must spend now.
• Leverage AI to reduce burden on simulations. With increasing demand for simulations, suppliers must invest in ways that reduce time and required simulation runs. AI should be used for simulation set-up, optimizing simulation feedback, and reuse of data to train AI models. Foundational AI models that can be fine-tuned to apply across projects remain a frontier for innovation.
• Support designers and engineers with resources to reduce simulation runs. As suppliers develop AI tools and Reduced-Order Models (ROMs) to approximate simulation results, they must also develop capabilities in copilots and user interfaces that can guide users in deciding whether a product design requires full simulation or can be appropriately approximated. This ensures that the AI models are utilized and will help manage simulation demand.