Jensen Huang Just Said $40 Trillion. Here Are 5 Physical AI Stocks Wall Street Is Quietly Loading Up On Before the Rest of the Market Catches On
NVIDIA CEO Jensen Huang has called humanoid robots and labor automation a $40 trillion total addressable market, and on the Animal Spirits podcast, Derek Yan argued physical AI is "potentially bigger" than EVs or smartphones, with Waymo serving as live proof the underlying autonomy stack already works in the wild. Capital is rotating into this trade now, before the broader market has priced it in. Here are the five stocks I'm watching, starting with the name almost nobody on retail Twitter is yelling about. 1. Cognex (CGNX): The Eyes Behind Every Robot Robots without vision are paperweights. Cognex ( NASDAQ:CGNX ) is the machine-vision standard for factory floors and warehouses, and in Q1 it shipped the In-Sight 6900 AI vision platform powered by NVIDIA alongside the In-Sight 3900 powered by Qualcomm. That's the company plugging itself directly into Jensen's Isaac and Cosmos stack at the edge, exactly where humanoid and industrial robots need to see, sort, and decide in real time. The numbers tell you Wall Street is already on it quietly. Q1 FY26 revenue hit $268.44 million, up 24.3% year over year, with adjusted EPS of $0.34, beating the $0.25 estimate, and Q2 guidance points to adjusted EPS of $0.40 to $0.44, roughly 68% year-over-year growth at the midpoint. The stock is up 86% year to date as I write this, and the chart looks like a setup that the rest of the market hasn't fully understood yet. CEO Matt Moschner put it bluntly: "Our latest AI vision products reinforce our technology leadership and objective of becoming the #1 provider of AI-powered machine vision." If Cognex supplies the eyes, somebody supplies the brain. 2. NVIDIA (NVDA): The Brain You knew this name was coming. NVIDIA ( NASDAQ:NVDA ) is the operating system of physical AI: the DRIVE Hyperion partnerships across Hyundai, Kia, Uber, BYD, Geely, Isuzu, and Nissan, the Isaac GR00T humanoid foundation model, and Cosmos world models for synthetic training data. I've owned this stock for over 15 years, and the pivot from gaming GPUs to the central nervous system of the robotics economy is the most aggressive platform expansion I've seen from any company in my career. Act now: the analyst who called NVIDIA in 2010 just named his top 10 AI stocks — and Cognex didn't make the cut. Grab the names FREE today . On the most recent call, Huang told investors NVIDIA has "line of sight to projects requiring tens of gigawatts of NVIDIA Corporation AI infrastructure in the not-too-distant future" and said "billions of robots, hundreds of millions of autonomous vehicles, and hundreds of thousands of robotic factories and warehouses will be developed." Data Center revenue ran $39 billion, up 73% year over year, and Q2 guidance came in at $45 billion at the midpoint. The street is still constructive: 48 Buy and 10 Strong Buy ratings against just 1 Sell, with an analyst target of $295.69 versus the current $212.60. Polymarket traders are pricing a 0.65 probability NVIDIA hits $216 in June, with upside scenarios into the $240 range. The compute is in place. Now somebody has to put it on wheels. 3. Tesla (TSLA): The Robot Company Disguised as a Car Company Elon Musk is building Tesla into a robotics company in plain sight. Tesla ( NASDAQ:TSLA ) is the most vertically integrated physical-AI bet in public markets: Optimus humanoids, FSD, Robotaxi, Cybercab, Megapack, and the only real-world fleet generating training data at scale. Optimus production lines are being installed at Fremont (designed for 1 million robots per year) with a second-gen line at Gigafactory Texas designed for 10 million robots per year. Q1 FY26 delivered revenue of $22.39 billion, up 15.8% year over year, with non-GAAP EPS of $0.41 versus a $0.36 estimate and auto gross margin expanding to 21.1% from 16.2%. The really interesting number: FSD active subscriptions hit 1.28 million, up 51% year over year. That's a software annuity layered on top of a hardware business that's already throwing off cash. Prediction markets are skeptical on near-term Optimus and California robotaxi timelines, with Polymarket pricing only a 0.11 probability of a California robotaxi launch by June 30. That gap between Huang's $40 trillion thesis and traders' short-window pricing is exactly where asymmetric returns get made. Big trucks and big factories are where this thesis pays first. 4. Symbotic (SYM): Where Physical AI Is Already Cashing Checks Forget the demo videos. Symbotic ( NASDAQ:SYM ) already runs end-to-end robotic warehouse systems for Walmart and a growing roster of retailers, with the SoftBank-backed Exol joint venture targeting the warehouse-as-a-service market. This is the rare physical-AI play with revenue that scales as humans get pulled out of fulfillment centers. Q2 FY26 revenue came in at $676.48 million, up 23.1% year over year and beating consensus. Three numbers matter from this report: 70 active systems in deployment (up from 46 a year ago), adjusted EBITDA of $77.75 million, more than doubling year over year, and a contracted backlog around $22.70 billion. That backlog is multiple years of revenue locked in regardless of macro chop. CEO Rick Cohen said "customers across several verticals are now realizing tangible value from our end-to-end automation systems", and retail still has no idea this exists. Reddit chatter shows just 1 qualified mention per tracking period, with a brief bullish sentiment score of 68 appearing only once. The biggest names are deploying systems in stockrooms. The last mile is a different animal entirely. 5. Serve Robotics (SERV): The Punchline And here's the payoff. Serve Robotics ( NASDAQ:SERV ) is the pure-play physical-AI small cap that almost nobody is talking about, and its Gen3 sidewalk robot runs on NVIDIA Jetson Orin compute. After the Diligent Robotics acquisition, Serve now operates sidewalk delivery robots and Moxi hospital robots across 44 cities in 14 states, with roughly 2,000 outdoor robots and over 100 hospital robots in service. This is Waymo's proof-of-concept applied to the last mile and the hospital corridor. Q1 FY26 revenue was $2.98 million, up roughly 578% year over year, with management reaffirming ~$26 million in 2026 revenue, roughly 10x fiscal 2025's $2.7 million. Daily active robots jumped to 812 from 73 a year ago. CEO Ali Kashani framed it directly: "We are leading the development of Physical AI in the real world, operating across multiple physical domains while building towards a unified autonomy platform." The target is a sub-$1 per delivery cost versus $8 to $10 with human couriers in what management frames as a $450 billion robotic and drone delivery opportunity by 2030. The stock is down 15% year to date, sitting at a $8.84 market price against an $18.45 analyst target with 8 Buy ratings and zero Holds or Sells. That's the asymmetric setup. If Huang and Musk are even half-right about robots replacing labor at scale, a $750 million market cap on a company already deploying autonomous fleets in dozens of cities is the kind of mispricing that doesn't last. The Trade Vision (CGNX), compute (NVDA), vehicles and humanoids (TSLA), warehouses (SYM), last-mile autonomy (SERV). That's the full stack of Huang's $40 trillion thesis, in order of how the capital flows. Trillion-dollar CEOs are directing billions into this right now, before consensus catches up. The window between "quietly loading up" and "crowded trade" is closing.
Source: AOL.com