NVIDIA
Supplies the robot 'stack': Jetson Thor onboard compute, Isaac simulation and GR00T humanoid foundation models
Public · No humanoid status · Robotics leader · Santa Clara, California, United States · Founded 1993 · NVDA (Nasdaq) · nvidia.com
Also listed as a component supplier. The GTC 2026 press release date is approximate (GTC week).
Profile
NVIDIA is the default compute and simulation supplier for Western humanoid makers, and Figure, Agility, Boston Dynamics and 1X use Jetson Thor. It publishes open GR00T humanoid foundation models and Cosmos world models. It is also an active investor in Figure, Skild AI, FieldAI and NEURA.
Also listed as a component supplier.
People
- Jensen Huang FounderCo-founder, President and Chief Executive Officer
Listed in NVIDIA's filings as Jen-Hsun Huang.
- Deepu TallaVice President and General Manager of Robotics and Edge AI
- Jim FanCo-lead of NVIDIA's embodied AI research group (GEAR)@DrJimFanHandle linked from personal site
- Yuke ZhuDirector, NVIDIA AI; co-leads GEAR
Also an associate professor of computer science at UT Austin.
@yukezRole confirmed in their X bio
Sources: NVIDIA annual report on Form 10-K for fiscal 2026 (executive officers) · An interview with Nvidia's Deepu Talla about physical AI and robotics (Chipstrat) · NVIDIA GEAR research group page · X profile @yukez (bio) · As of . X handles are shown only when we could confirm they belong to the person.
Leadership on X
What leadership has posted about robotics, AI and tech in the last 30 days.
- Argued that a robot able to clean a home as well as a person is further off than major math breakthroughs, calling it the 'Physical Turing Test'.1
Chatter on X
Social posts, not verified facts. What X is saying about this company, rated for how far it can be trusted. Shows the last 14 days, plus anything still unresolved.
Items shown have cleared our internal credibility screen.
- ConfirmedProduct demoFirst seen Oct 7 · Updated Oct 10
NVIDIA's Seattle Robotics Lab is building robots to assemble GB300 server trays
NVIDIA shows its own robot stack tackling delicate hardware assembly such as fastening busbars and inserting connectors. It is a company demo, not a customer deployment.
Why this label- Posted by NVIDIA Robotics, the company's official account
Primary post on X · @NVIDIARobotics, Oct 7Backing: NVIDIA Robotics post on X, Oct 7
- ConfirmedSupply chain / customerFirst seen Oct 6 · Updated Oct 10
Unity releases an OpenUSD toolkit with NVIDIA for robotics and simulation teams
The toolkit targets industrial teams in manufacturing, robotics and simulation. No financial terms were disclosed.
Why this label- Announced by Unity's official account
Primary post on X · @unity, Oct 6Backing: Unity's announcement on X, Oct 6
- LikelyFunding / dealRumorFirst seen Oct 7 · Updated Oct 10
Rumor: NVIDIA reportedly discussing another $1B investment in Figure AI
The Information reports the talks as Figure seeks funding at about a $38B pre-money valuation. Neither company has confirmed it.
Reported, not yet confirmed by the company.
Why this label- Single anonymous source, repeated by other outlets
- No confirmation or denial from NVIDIA or Figure
Primary post on X · @wlthxyz, Oct 8Reporting: Humanoids Daily
Labels
- Confirmed
- The company, an SEC filing or a major outlet backs it.
- Likely
- A credible first-hand source or named outlet reports it; nothing official yet.
- Unverified
- One source, or a source with a thin track record. Treat as a rumor.
- Disputed
- A credible party contests it.
Last checked Oct 10.
Flagship products
- Jetson Thor Component
Blackwell-based onboard robot computer - Isaac GR00T N1.x / N2 AI model
Open humanoid foundation models; N2 previewed at GTC 2026 - Isaac Sim / Cosmos Software
Simulation and world models
Market & financials
Competitors
Companies with overlapping products. Direct = same customers and same job; Adjacent = overlapping technology or market, different focus.
Skild AI
Direct competitorBoth build general robot AI models that other companies' robots can run: Skild AI builds Skild Brain, a foundation model meant to run on any robot body, while NVIDIA supplies Jetson Thor compute and the open GR00T humanoid foundation models.
Physical Intelligence
Direct competitorBoth build general robot AI models that other companies' robots can run: Physical Intelligence builds the general-purpose π robot foundation models, while NVIDIA supplies Jetson Thor compute and the open GR00T humanoid foundation models.
Rhoda AI
Direct competitorBoth build general robot AI models that other companies' robots can run: Rhoda AI is building a robot brain pre-trained on internet video, while NVIDIA supplies Jetson Thor compute and the open GR00T humanoid foundation models.
Show 3 more
FieldAI
Direct competitorBoth build general robot AI models that other companies' robots can run: FieldAI builds Field Foundation Models for robots working in unstructured sites, while NVIDIA supplies Jetson Thor compute and the open GR00T humanoid foundation models.
Genesis AI
Direct competitorBoth build general robot AI models that other companies' robots can run: Genesis AI builds a robot-brain model and now its own hand and the Eno robot, while NVIDIA supplies Jetson Thor compute and the open GR00T humanoid foundation models.
Google DeepMind (Gemini Robotics)
Direct competitorBoth build general robot AI models that other companies' robots can run: Google DeepMind develops the Gemini Robotics models, used by partners such as Apptronik, while NVIDIA supplies Jetson Thor compute and the open GR00T humanoid foundation models.
Recent news
- — Nvidia and Foxconn report robot assembly results on AI server lines Taipei Times
- — GR00T N2 previewed at GTC 2026 alongside expanded open model families NVIDIA
From SEC filings
Robotics-related passages from SEC filings, newest first.
Scope Quarterly report for Q2 fiscal 2027; only robotics and physical AI passages are summarized.
NVIDIA's cloud commitments partly fund research on its open models, including the GR00T robot models.
These commitments provide the cloud infrastructure to support our research and development of our open models, such as NVIDIA Nemotron, Cosmos, and GR00T, and our autonomous vehicle software.
Item 1. Financial Statements
Scope Q2 fiscal 2027 results announcement; highlights come from the attached press release.
NVIDIA announced the Isaac GR00T Reference Humanoid Robot, an open humanoid reference design built on Jetson Thor.
Announced the NVIDIA Isaac™ GR00T Reference Humanoid Robot, the first open humanoid robot reference design built on NVIDIA Jetson Thor™ and the NVIDIA Isaac GR00T open development platform.
Exhibit 99.1, press release (Item 2.02 Results of Operations and Financial Condition)NVIDIA announced Halos for Robotics, a safety system for physical AI.
Announced NVIDIA Halos for Robotics, the industry’s first full-stack safety system for physical AI, unifying AI compute and safety in a single comprehensive platform.
Exhibit 99.1, press release (Item 2.02 Results of Operations and Financial Condition)NVIDIA launched Cosmos 3, an open model for physical AI.
Launched NVIDIA Cosmos™ 3, the world’s first fully open frontier omnimodel for physical AI.
Exhibit 99.1, press release (Item 2.02 Results of Operations and Financial Condition)
Show 2 earlier filings
Scope Q1 fiscal 2027 results announcement; highlights come from the attached press release.
NVIDIA announced new Cosmos and Isaac GR00T N robot models, new simulation tools, and general availability of IGX Thor.
Announced new NVIDIA Cosmos™ and NVIDIA Isaac™ GR00T N models, new Isaac simulation frameworks, the general availability of NVIDIA IGX Thor™ and physical AI leaders building on NVIDIA technology.
Exhibit 99.1, press release (Item 2.02 Results of Operations and Financial Condition)NVIDIA now reports robotics within its Edge Computing category.
Edge Computing highlights data processing devices for agentic and physical AI including PCs, game consoles, workstations, AI-RAN base stations, robotics and automotive.
Exhibit 99.1, press release (Item 2.02 Results of Operations and Financial Condition)
Scope Annual report for fiscal 2026; only robotics and physical AI passages are summarized.
NVIDIA describes an end-to-end physical AI platform for training, simulating and deploying robots.
For physical AI, we provide an end-to-end platform spanning data center infrastructure, open models, systems, embedded compute modules, and software stacks to train, simulate, and deploy advanced automation and robotics solutions.
Item 1. BusinessNVIDIA says it sped up releases of its open model platforms, including Cosmos for physical AI.
More recently, in support of market development, we have accelerated the release cadence of our open AI model platforms including NVIDIA Nemotron for agentic AI and Cosmos for physical AI.
Item 1. Business
Source: SEC EDGAR · Quotes checked word for word against the filing · As of .
Earnings calls: what they said about robotics
As of .
What the company said on its earnings calls and keynotes, by topic, oldest call first. Each item shows its type: Forecast, Aspiration, Design capacity (what a plant is built to handle, not what it makes) or Fact.
What NVIDIA reports for robotics and automotive
- Q4 FY2026Fact statementColette Kress, CFO
Automotive revenue was $604 million in the fourth quarter of fiscal 2026, up 6% from a year earlier and up 2% from the prior quarter, which NVIDIA attributed to demand for self-driving solutions. For the full fiscal year, automotive revenue was $2.35 billion, up 39%. NVIDIA stopped reporting automotive as its own line after this quarter; it now sits inside Edge Computing.
“Automotive revenue of $604 million was up 6% year-over-year”
- Q4 FY2026Fact statementColette Kress, CFO
NVIDIA said physical AI contributed more than $6 billion of NVIDIA revenue in fiscal 2026. The surrounding remarks cover robotaxis and robotics, but NVIDIA did not define exactly what the figure includes or break it down by product.
“Physical AI is here, having already contributed north of $6 billion in NVIDIA revenue”
- Q1 FY2027Fact statementColette Kress, CFO
Starting with the first quarter of fiscal 2027, NVIDIA changed how it reports revenue. Robotics and automotive no longer appear as separate lines; they sit inside a new Edge Computing platform that also covers PCs, game consoles, workstations and AI-RAN base stations. Edge Computing revenue was $6.4 billion, up 29% from a year earlier and 10% from the prior quarter, driven mainly by workstation demand, so the figure says little about robotics on its own.
“Edge Computing highlights devices for agentic and physical AI”
- Q1 FY2027Fact statementColette Kress, CFO
NVIDIA said its physical AI revenue exceeded $9 billion over the last 12 months, compared with more than $6 billion in fiscal 2026. The two figures cover different periods, and NVIDIA did not split either between robotics and autonomous vehicles.
“exceeding $9 billion in revenue over the last 12 months”
NVIDIA's view of physical AI
- Q1 FY2027Aspiration statementJensen Huang, CEO
Huang said he hopes that within the next five years the physical AI and robotics segment will grow incredibly fast. He gave no revenue figure or target.
“physical AI and robotics segment is going to grow incredibly fast”
Named robotics customers and partners
- Q4 FY2026Fact statementColette Kress, CFO
NVIDIA named Boston Dynamics, Caterpillar, Franka Robotics, LG Electronics and NEURA Robotics as leading companies whose robots and autonomous machines are powered by NVIDIA, built on its new Cosmos and Isaac GR00T open models and frameworks.
“Boston Dynamics, Caterpillar, Franka Robotics, LG Electronics and NEURA Robotics”
- Q1 FY2027Forecast statementColette Kress, CFO
NVIDIA said its partnership with Uber will power a robotaxi fleet across nearly 30 cities on four continents by 2028. In the same passage it said leading companies in industrial, surgical and humanoid robotics are building on its technology to develop and deploy at scale, without naming them.
“robotaxi fleet across nearly 30 cities and 4 continents by 2028”
- Q2 FY2027Fact statementColette Kress, CFO
NVIDIA said Amazon will adopt its full physical AI stack, meaning Omniverse, Cosmos, Isaac and Jetson, to power Amazon's fleet of warehouse robots. NVIDIA gave no timing or volume.
“Omniverse, Cosmos, Isaac, and Jetson, to power its fleet of warehouse robots”
Simulation and world models
- GTC Taipei 2026 keynoteFact statementJensen Huang, CEO
At its Taipei keynote, NVIDIA announced Cosmos 3, an open foundation model for physical AI. Huang said it can understand and reason about the physical world, generate and simulate it, and can even act as the robot's policy. NVIDIA releases the model, the data and how it was trained.
“we are announcing Cosmos 3. Cosmos 3 is the frontier of physical AI”
Show 3 more topics
Robot computers and reference robots
- GTC Taipei 2026 keynoteFact statementJensen Huang, CEO
Huang announced a reference humanoid robot for university and higher-education researchers, fully integrated and running NVIDIA's new Thor robot computer and its full software stack for data generation, simulation and runtime. He described it as 31 degrees of freedom, with 25-degree-of-freedom hands made by Sharpa, about 6 feet tall and 150 pounds. NVIDIA's keynote slide says it is built on a Unitree H2 Plus and available in October 2026.
“reference humanoid robot, all fully integrated”
Watch at 1:47:40 on the replayDeck
Note: The Unitree H2 Plus base and the October 2026 availability come from NVIDIA's keynote slide, not from the spoken words.
Sovereign and industrial AI tie-ins
- Q2 FY2027Fact statementColette Kress, CFO
NVIDIA announced a partnership with Noetra, Japan's national AI company, to build an NVIDIA DSX AI factory that will create open models for AI agents, digital twins, robotics and physical AI applications. It also said South Korea's LG and Hyundai Motor Group are partnering with NVIDIA to build and scale AI.
“digital twins, robotics, and physical AI applications”
Supply and capacity
- Q2 FY2027Forecast statementColette Kress, CFO
NVIDIA said it expects supply to remain a bottleneck at least through the end of fiscal 2028, which ends in January 2028, and guided fiscal 2028 revenue growth of about 70% as a supply-constrained outlook. The remarks concern data-center supply, with memory prices the main cost pressure, and do not address robot components.
“supply to remain a bottleneck at least through the end of fiscal year 2028”
Statements are NVIDIA's own, from its earnings calls and keynotes, checked against the recordings and, for the earnings calls, NVIDIA's posted transcripts. NVIDIA's fiscal year ends in late January, so Q2 FY2027 is the quarter that ended July 26, 2026. NVIDIA no longer reports robotics or automotive revenue as separate lines, and "physical AI" is NVIDIA's own grouping, so these figures are not comparable with other companies' robot sales. Forecasts and aspirations can change.
Sources
- NVIDIA Blackwell-Powered Jetson Thor Now Available (NVIDIA Newsroom) (Aug 25, 2025)
- NVIDIA Expands Open Model Families to Power the Next Wave of Agentic, Physical and Healthcare AI (NVIDIA IR) (Mar 16, 2026)
- Nvidia and Hon Hai say their robots improving at building AI components (Taipei Times) (Oct 6, 2026)
Confidence: high.