Silvaco and NVIDIA Partner to Accelerate Semiconductor Digital Twins with AI
Event summary
- Silvaco and NVIDIA announced a collaboration on July 26, 2026 to advance physics-based digital twins for semiconductor design and manufacturing using AI and accelerated computing.
- Silvaco completed a fully scaled 3D FDTD simulation of a photonic edge coupler with 3.2 billion mesh nodes on 32 NVIDIA GPUs in under four hours, achieving less than 0.15 dB difference between measurement and simulation.
- The partnership aims to reduce simulation cycles from weeks to days, improve accuracy and insight, and scale engineering and collaboration through cloud-based visualization and AI-driven workflows.
The big picture
This partnership aligns with the broader industry trend of integrating AI into semiconductor design and manufacturing processes. As chip complexity grows, the ability to simulate and optimize designs faster could become a critical competitive advantage. The collaboration also highlights the increasing importance of digital twins in enabling real-time collaboration across global engineering teams.
What we're watching
- Adoption Pace
- How quickly semiconductor manufacturers will integrate AI-accelerated digital twins into their design and manufacturing workflows.
- Competitive Response
- Whether competitors like Synopsys or Cadence will respond with similar AI-driven simulation solutions.
- Technical Validation
- The extent to which early adopters validate the accuracy and efficiency gains promised by the Silvaco-NVIDIA collaboration.
Related topics