Technology is already working with examples like digital twins and AI on shop floors. Scaling requires making deployment easier for customers.

Announcing major partnership: accelerating Siemens' EDA and simulation software, integrating physical AI and agentic AI into Siemens' systems. Technology will be used in Nvidia's AI factories and with partners like Foxconn.

Vera Rubin GPU is 10x more energy and cost efficient. Accelerating EDA/simulation tools and using digital twins enables creating more complex systems efficiently, doing the impossible right the first time.

AI creates real-world economic impact, not just in data centers but also at the edge with low-latency inference, creating huge potential for customers.

Factories face labor shortages, especially unskilled labor. Automation leads to higher yields and energy efficiency. US manufacturing ramp-up needs to be digital and AI-supercharged.

Factories are robotic systems orchestrating robots. Hard to deploy robots due to programming complexity. AI makes them easier to teach - showing demonstrations allows AI to learn by itself.

Every industrial revolution is energy constrained. From Hopper to Blackwell to Rubin, improved energy efficiency by 10x each generation, directly improving customer revenues within power constraints.

Energy demand scales with GDP growth but decoupling due to efficiency. Data centers demand high-quality energy, creating bottlenecks across supply chain from generation to transformers.

Memory bottleneck is severe but Nvidia works with all three HBM suppliers and has long-term relationships with plans in place.

Haven't spoken directly to Chinese government. Communication is through companies - if companies are allowed to buy, there will be strong demand, which is already being seen.

Siemens invested nearly $30B in software competence. Can build comprehensive physics-based digital twins but needs operations software for specific industries like life sciences.

Hired Groq engineers and licensed their technology. Groq architecture focused on low-latency token generation and inference. Excited about inventing new segment together for future use cases.

Can't discuss specific conversations. Space has abundant energy and cooling. System design would be radically different but chips would be same.

Manufacturing in space challenging to bring physical goods to Earth, but tokens/intelligence can be transferred easily - that's where to start in space.

Tesla has most advanced AV stack and operations. Nvidia's approach also vision-based with addition of radar and lidar. Similar approaches overall. Tesla doing great job.

Works in Silicon Valley for talent pool. Has offices worldwide. Fine with whatever taxes apply - never crossed mind as concern.