NVIDIA has enhanced its Agent Toolkit by integrating re-architected PhysicsNeMo libraries and updated CUDA-X libraries, enabling developers to build autonomous AI engineers with advanced physics, simulation, and quantum chemistry capabilities. This expansion supports complex chip and system design workflows by providing AI physics skills and accelerated solvers, helping companies like Cadence, Siemens, and Synopsys speed up chip design, verification, and packaging. The new tools, including the Nemotron 3 Ultra open model, improve accuracy and efficiency in hardware design coding and allow enterprises to customize AI agents for their specific needs. These advancements mark a significant step toward automating and accelerating engineering processes in the semiconductor industry.
Silvaco and NVIDIA have teamed up to enhance semiconductor design and manufacturing by integrating NVIDIA's AI and accelerated computing with Silvaco's physics-based simulation technology. This collaboration aims to create high-fidelity digital twins...
Silvaco has teamed up with NVIDIA to enhance semiconductor design and manufacturing through AI-powered digital twins and accelerated computing. This collaboration integrates Silvaco's physics-based simulations with NVIDIA's AI and GPU technologies to...
Nvidia has successfully signed Toyota, Fanuc, Kioxia, and five other leading Japanese industrial companies into its physical AI coalition this week. This move highlights Nvidia's expanding influence in the AI sector and its strong partnerships with k...
Nvidia has signed agreements with eight leading Japanese industrial companies, including Toyota, Fanuc, and Kioxia, to join its Physical AI Coalition. This move strengthens Nvidia's position in the industrial AI market and confirms $1 trillion in dem...
Chinese AI startup Moonshot AI's rapid rise with its open-weight Kimi K3 model, which competes with US models, has sparked debate over the practice of AI distillation. Distillation uses outputs from advanced AI models to train others, raising concern...
