Nvidia’s winning streak and robotics push

Nvidia shares closed at $198.87 on April 15 after 11 straight up days, the company’s longest winning streak on record. (gurufocus.com) The stock rose 1.2% that day and gained 20.4% over the 11-session run, according to GuruFocus. Yahoo Finance showed Nvidia at $198.35 at the April 16 close, with the stock up 89.83% over the past year. (gurufocus.com) (finance.yahoo.com) Investors are still buying Nvidia first for data-center chips, but the company has spent the past year widening its pitch to “physical AI” — software and chips that help robots perceive, plan and act in factories, warehouses and hospitals. (investor.nvidia.com) (blogs.nvidia.com) That push was on display at Nvidia’s March 16, 2026 announcements, where the company said ABB Robotics, FANUC, KUKA, Universal Robots, YASKAWA, Medtronic and Figure were building on its technology. Nvidia said those partners span industrial, surgical and humanoid robotics. (investor.nvidia.com) The basic idea is to train robots in simulation before they touch the real world. Nvidia’s Isaac GR00T stack combines robot “brains” for reasoning and control, synthetic training data, simulation tools and Jetson AGX Thor, the onboard computer meant to run the system inside a robot. (developer.nvidia.com) Nvidia started laying out that roadmap in March 2025, when it introduced Isaac GR00T N1, which it called an open foundation model for humanoid robot reasoning and skills. In May 2025, it followed with GR00T N1.5, GR00T-Dreams and Blackwell-based systems aimed at robot training and simulation. (nvidianews.nvidia.com 1) (nvidianews.nvidia.com 2) At GTC 2026, Nvidia added an “open physical AI data factory blueprint,” which it said is meant to speed development of robotics, vision AI agents and autonomous vehicles. The company also grouped robotics with broader industrial software and manufacturing announcements, not just chip launches. (nvidianews.nvidia.com) The market case for Nvidia is that each layer of that stack can sell more silicon: big graphics processors in the cloud for training, workstation and server chips for simulation, and Jetson systems at the edge where robots operate. Nvidia described that model in March as cloud-to-robot computing. (blogs.nvidia.com) (nvidianews.nvidia.com) The open question is how quickly those robotics bets turn into revenue on the scale of Nvidia’s data-center business. For now, the stock’s 11-day run shows investors are still willing to pay for the idea that Nvidia can extend its AI lead from servers into machines that move in the physical world. (gurufocus.com) (investor.nvidia.com)