NVIDIA at CES: Defining the Future of Agentic Systems and Physical AI

Google Cloud Next '26 keynote, Blueprint for the Agentic Enterprise

Punchlines: * "The era of the pilot is over. The era of the agent is here." * Nearly 75% of customers now leverage AI in production, and internally at Google, AI generates nearly 75% of all new code. * We are firmly in the agentic Gemini era, shifting from asking if we can build an agent to how we can securely manage thousands of them. * "Compute is no longer defined by chip. Compute is the entire data center," powered by the new AI Hypercomputer and 8th-generation TPUs. * "Reasoning without context is just a guess," which is why the new Agentic Data Cloud turns unstructured information into decisive action. * Cybersecurity must now operate at "machine speed," as human analysts can no longer keep up with AI-weaponized threats and shrinking exploit windows.

NVIDIA GTC 2026 Keynote: The Era of AI Factories, Agentic Systems, and the Vera Rubin Platform

Key Punchlines: * The Inference Inflection: AI has evolved from perceiving to reasoning and taking action, driving massive compute demand that is projected to reach at least $1 trillion by 2027. * Tokens as a Commodity: Data centers are transforming into power-constrained AI factories where maximizing token throughput directly equals revenue. * Vera Rubin Architecture: NVIDIA introduced the Vera Rubin platform, integrating the new Vera CPU, Rubin GPU, and Groq LPU to boost inference token generation performance by 35x. * The OpenClaw Revolution: Acting as the "operating system" for AI agents, OpenClaw and the enterprise-secure Nemo Claw framework will turn every software provider into an Agentic-as-a-Service company. * NVIDIA DSX: A new Omniverse digital twin platform enables companies to virtually design, simulate, and optimize the energy efficiency of giant AI factories. * Physical AI: The generative AI moment for robotics and self-driving cars has arrived, powered by platforms like Alpamo, Isaac Lab, and Cosmos world models.

The Terafab Project and the Path to a Galactic Civilization

Punchlines: * Becoming a multi-planetary species is essential to harness the sun's massive energy and effectively scale human civilization. * SpaceX, Tesla, and xAI are collaborating to build the "Terafab" in Texas, an unprecedented facility aiming to produce a terawatt of AI compute per year with a rapid iteration loop. * Deploying AI compute in space will soon be cheaper and easier than on Earth due to constant solar power, infinite scaling room, and zero atmospheric interference. * Achieving this ambitious goal requires launching 10 million tons of payload to orbit annually, which SpaceX plans to accomplish with its Starship. * The long-term vision includes building an electromagnetic mass driver on the moon to reach a petawatt of compute. * This massive scale of AI and robotics aims to unlock a post-scarcity economy of total abundance, where money becomes obsolete and trips to Saturn are available to anyone.

NVIDIA at CES: Defining the Future of Agentic Systems and Physical AI

• "There is a bubble in media headlines, but not in AI infrastructure adoption." • "At Snowflake, the rule for enterprises is simple: your data is your data." • "AI flips the healthcare script, moving from 80% clerical work to 80% clinical reasoning." • "Coding has become the gateway to actuation and the software of the future." • "You no longer program software; you train it on GPUs instead of CPUs." • "The ChatGPT moment for physical AI—where AI understands the laws of nature—is nearly here." • "Vera Rubin is in full production, representing a giant leap engineered through extreme co-design." • "We are cooling this supercomputer with hot water to maximise global data centre efficiency." • "The agentic system is the new interface; we are moving beyond the era of the spreadsheet." • "AI is a five-layer cake being reinvented from the land and power up to the applications."

Winning Against the Macro: C.H. Robinson's Lean AI Strategy Delivers Exponential Results

Punchlines We are a technology-generated company servicing a logistics industry. Lean is the engine and AI is the accelerator. AI is not a buzzword; it is how we are running the company. We have fundamentally changed the processes to decouple head count growth from volume growth. We have achieved over 40% productivity increase since the end of 2022. Our core strategy is to outgrow our markets and expand our operating margins simultaneously. Automated quoting now covers 100% of requests in about 32 seconds, up from 65% manually. We are migrating from generative AI to the game-changing potential of agentic AI. We are nearly at mid-cycle margins in the worst freight market ever. The technology makes the operating model better, and the operating model makes the technology better. The next two years will be more exciting than the last two years because we are still in the early innings. We do no AI for hobby; every investment must show a specific business return.