NVIDIA Expands Agent Toolkit with PhysicsNeMo and CUDA-X Libraries, Deploys Vera CPU in Chip Design Workflows

NVIDIA has expanded its NVIDIA Agent Toolkit by integrating re-architected PhysicsNeMo and new and updated CUDA-X libraries, enabling developers to build autonomous AI engineers. The toolkit introduces AI physics skills, sparse solvers, and quantum chemistry capabilities. Industry leaders are already adopting these tools, with Cadence delivering up to 20x faster multiphysics performance and Siemens achieving more than 10x faster library characterization while reducing token costs by more than 10x. Additionally, Samsung, Synopsys, and TSMC are integrating the cuEST library into GPU-accelerated pipelines to achieve up to a 50x speedup in key quantum-chemistry workloads.

Agent Toolkit PhysicsNeMo expansion

• ▪The expanded NVIDIA Agent Toolkit enables software developers to build autonomous AI engineers with AI physics skills, accelerated solvers, and quantum chemistry capabilities
• ▪NVIDIA expanded its NVIDIA Agent Toolkit for engineering by adding re-architected NVIDIA PhysicsNeMo libraries and updated NVIDIA CUDA-X libraries

Nemotron 3 Ultra RTL coding

• ▪NVIDIA Nemotron 3 Ultra can be post-trained on proprietary data and deployed locally or on premises to provide enterprises with greater customization and data privacy
• ▪NVIDIA Nemotron 3 Ultra leads among open models in agentic register-transfer level coding with the ACE-RTL agent from NVIDIA Research on the comprehensive verilog design problems benchmark

Industry partner adoption speedups

• ▪Siemens is using NVIDIA NeMo Gym, Nemotron open models, and CUDA-X libraries with the Siemens Fuse EDA AI Agent to deliver over 10x faster library characterization while reducing token costs by more than 10x
• ▪Cadence is using NVIDIA Nemotron, accelerated computing, and CUDA-X libraries with the Cadence AuraStack AI Super Agent and the Cadence Millennium M2000 platform to deliver up to 20x faster multiphysics performance

AI physics capabilities

• ▪Samsung is applying NVIDIA PhysicsNeMo to perform chip-scale thermal-stress analysis with numerical solver-level accuracy across domains containing up to 10 billion cells
• ▪NVIDIA PhysicsNeMo libraries help agents train and deploy customizable AI physics models for complex design and simulation tasks, turning model architectures into callable tools

Quantum chemistry integration

• ▪NVIDIA cuEST brings high-accuracy quantum chemistry simulations to device-relevant scales, enabling density functional theory and post-DFT methods to be integrated into production workflows
• ▪Samsung, Synopsys, and TSMC are integrating NVIDIA cuEST into their GPU-accelerated pipelines to achieve up to a 50x speedup for key quantum-chemistry workloads