The daughter of NVIDIA's CEO made a rare public appearance in Beijing. On the opening day of the 2026 World Robot Conference, Madison Huang visited the Yizhuang district, touring robotics companies such as Dobot, Lightwheel AI, and UBTech Robotics. The post-90s generation "NVIDIA heiress" now serves as Senior Director of Product and Technical Marketing for NVIDIA's physical AI platform, earning an annual salary of $1.2 million, roughly 8 million RMB. What is less known is that she initially studied culinary arts, worked as a chef, and was previously employed at LVMH. After several career pivots, she eventually joined her father's company as an intern. Her arrival in Beijing coincides with a peak moment for China's robotics sector, with Unitree Robotics having just gone public and a wave of Chinese robotics IPOs on the horizon. The second half of the robotics race has begun.

Jensen Huang's Daughter Visits Beijing to Scout Robotics

Global attention in robotics has once again turned to Beijing. On August 19, the 2026 World Robot Conference opened in Yizhuang, where a familiar face appeared in the crowd—Jensen Huang's daughter, Madison Huang. She moved between multiple robotics booths throughout the exhibition hall. At the Dobot Robotics booth, Madison observed a demonstration of the "one brain, multi-body" embodied AI platform and watched robots playing table tennis. At the Lightwheel AI booth, she focused on the company's progress in human data collection, physical simulation, large-scale evaluation, and real-world deployment feedback. During her visit, the NVIDIA delegation personally tested physical simulation equipment and engaged with the team on details such as simulation scenario generation and parameter calibration. She subsequently visited booths including UBTech and Huazhijian, where UBTech's commercial service humanoid robot Walker C1 approached her holding a bouquet of flowers, offering them to her with a handshake. She then spent considerable time experiencing the ultra-bionic humanoid robot U1. At the Kepler Robotics booth, Madison tried the heavy-duty Kirin series quadruped robot, and at the JD.com exhibition area, she continued exploring robot real-world operations and data training systems, watching action demonstrations in specific scenarios and exclaiming "Amazing" several times. Many attendees recognized the "NVIDIA heiress," with some taking photos on their phones and others approaching for pictures. Unlike her father Jensen Huang, who frequently visits China, Madison has rarely appeared in the public eye. Just one day before her Beijing trip, Madison led a team to tour LG's robot data factory in Seoul, South Korea. Her itinerary has consistently revolved around robotics, which aligns with her current position at NVIDIA.

Madison currently serves as Senior Director of Product and Technical Marketing for NVIDIA's physical AI platform, with responsibilities covering products such as Omniverse, Cosmos, Isaac, and the humanoid robot foundation model GR00T. Over recent years, these business segments have grown increasingly significant within NVIDIA's portfolio, representing the next wave repeatedly highlighted by Jensen Huang: physical AI. In simple terms, large language models enable AI to understand and generate text, while physical AI aims to help machines comprehend the real world and take actual action. Robots serve as the most critical carrier for this vision. Consequently, her Beijing visit carries deeper implications. The companies Madison observed represent some of China's most active robotics firms today, forming an ecosystem that NVIDIA cannot ignore in its physical AI ambitions. Rewinding more than a decade, her life trajectory was far removed from chips, AI, and robotics. Back then, she chose to enter the kitchen.

From Chef to an $800,000 NVIDIA Salary

Madison Huang is a post-90s generation member. When NVIDIA was founded, she was only one or two years old. Growing up in an engineering family, she initially considered following her parents' path and applying for an electrical engineering major upon high school graduation. Unexpectedly, it was Jensen Huang who first advised her to reconsider. In his view, his daughter lacked genuine passion for engineering—her real interest lay in cooking. Madison then withdrew her original application and enrolled at the Culinary Institute of America. After graduating in 2012, she worked at restaurants in New York and San Francisco, then traveled to Paris to study pastry, followed by wine studies at Le Cordon Bleu in London. Later leaving the culinary industry, Madison joined LVMH, where she gained exposure to marketing and brand management. These seemingly unrelated experiences share a common thread: the kitchen demands precision in timing, heat, and detail, while luxury goods similarly emphasize product quality and experience. Madison discovered that what truly captivated her was the process of refining a product to excellence—a near-obsessive attention to detail and a deep enjoyment of perfecting every aspect. The turning point that brought her back to the technology sector came in 2019, when Madison and her brother Spencer participated in MIT's artificial intelligence course together. She subsequently pursued an MBA at London Business School. The following year, while still studying, Madison entered NVIDIA as an intern. Initially, she struggled to adapt, often feeling lost amid the technical jargon in meetings. She once shared that she created a document to record every unfamiliar term, researching and studying each one after meetings. Over time, the list of unfamiliar words grew shorter, and her work became increasingly seamless. Within a few years, Madison advanced from intern to senior director. Reflecting on her journey, she noted in a recent speech that the AI era does not require a linear career path—detours ultimately become valuable assets. Interestingly, her brother Spencer also followed a similarly unconventional route. After college graduation, Spencer opened a cocktail bar with a friend. The bar had no fixed menu; guests described their tastes and the bartenders created drinks on the spot, eventually earning a spot on the "Asia's 50 Best Bars" list. Madison joined NVIDIA first, while Spencer later closed the bar, returned to academia, completed his MBA, and joined NVIDIA as an intern in 2022. He now primarily works on robotics software and simulation-related products, including robot learning simulation and synthetic data generation. According to NVIDIA disclosures, Madison's total compensation for fiscal year 2026 was approximately $1.232 million, equivalent to roughly 8.3 million RMB, while Spencer's total compensation was approximately $1.32 million. This inevitably raises discussions about nepotism. At an NVIDIA all-hands meeting in 2025, Jensen Huang did not sidestep the issue, acknowledging that the company does employ a number of employees' children, but noting that parents rarely recommend children who would bring them embarrassment. He even joked that some second-generation employees had outperformed their parents. After their respective journeys, both siblings have returned to NVIDIA, converging on the company's next battleground—robotics.

China's Hottest Track

This year's World Robot Conference remains as vibrant as ever. Nearly 400 companies brought over 3,000 exhibits, with more than 300 new products unveiled collectively. Beyond dancing, boxing, and table tennis, an increasing number of robots are now handling material transport, item sorting, and equipment operation, appearing in real-world scenarios such as automotive manufacturing, logistics, and commercial services. The changes on the exhibition floor mirror the broader industry. Over the past few years, China's humanoid robots have rapidly completed the journey from zero to one. Startup companies have emerged in rapid succession, complete machine iterations are accelerating, and core component supply chains are gradually taking shape. Capital has followed suit, with financing records and valuations continuously being refreshed. Now, the first batch of companies is beginning to go public. On the very opening day of the World Robot Conference, Unitree Robotics officially listed on the STAR Market, with its stock price surging more than 600% on the first day. The IPO pipeline is accelerating. AGIBOT has announced the initiation of its Hong Kong listing process, CloudMinds' STAR Market IPO application has been accepted, and companies including Lehuzhihang, Skyline Robotics, Different AI, LimX Dynamics, Zhongqing Robotics, Astribot, and Galbot are also advancing their capitalization efforts. According to incomplete statistics, more than 50 robotics-related companies are queuing for listings at the Hong Kong Stock Exchange alone. However, the ringing of the bell is not the finish line. While capital markets can assign valuations, the industry ultimately must provide the answers. In the first half of this year, China's humanoid robot shipments have already exceeded 40,000 units. Following the rapid growth in volume, more critical questions are emerging: Can robots adapt to more diverse tasks and still understand instructions and complete actions in unfamiliar environments? The true competition in robotics is thus shifting from the "body" to the "brain." Wang Xingxing noted at this year's World Robot Conference that the real "ChatGPT moment" for robotics should be when a robot enters an environment it has never seen before and completes the vast majority of tasks based solely on voice or text commands. He estimated that this day may still be two to ten years away. This pushes the competition into deeper territory. China already possesses one of the world's most complete manufacturing systems. From components to complete machines, robots undergo research, development, production, and iteration at remarkable speed. More importantly, the vast industrial scenarios in automotive, consumer electronics, logistics, and retail provide robots with a rare testing ground. Robots are built here and then deployed into the real world. When global tech giants turn their attention to China's robotics industry, they see not just a proliferation of startups, but a genuine industrial competition. Moving forward, the true test is only just beginning.