Key points
- Nvidia (NVDA) expects physical AI revenue to hit $100 billion in a decade, up from $10 billion today.
- Nvidia booked $96.2 billion in revenue last quarter.
- Goldman sees humanoid robots alone at about $138 billion by 2035, a narrower market than physical AI.
Deepu Talla told the Financial Times he expects Nvidia's (NVDA) revenue connected to physical AI to grow tenfold over the coming decade, from $10 billion to $100 billion. Talla is the company's vice-president of robotics and edge AI. Nvidia recorded $96.2 billion in revenue for the quarter that ended July 26, 2026, while revenue across the latest twelve months was approximately $303 billion. On that basis, today's $10 billion physical AI figure is a little over 3% of total revenue. The $100 billion projection for ten years from now slightly exceeds what Nvidia presently earns in one quarter.
"Ninety per cent of the world's actions need to happen in the physical world, not just in the digital world, that's why we believe the opportunity for physical AI is arguably an order of magnitude larger," Talla told the paper.
Physical AI covers robots, autonomous vehicles, and other systems that act in the world rather than generating text or images. Chief executive Jensen Huang has said every industrial company will become a robotics company as AI moves from the digital to the physical realm. Elon Musk is directing tens of billions of dollars into Tesla (TSLA) as it shifts from electric vehicles toward robotaxis and humanoids, and manufacturing start-ups selling what the industry calls robotics-as-a-service have drawn valuations closer to those of software companies.
| Nvidia (NVDA) revenue | Period | Amount |
|---|---|---|
| Latest quarter | Ended July 26, 2026 | $96.2 billion |
| Trailing twelve months | Through July 26, 2026 | About $303 billion |
| Physical AI today | Current annual run rate | About $10 billion |
| Physical AI target | Within ten years | $100 billion |
Washington is watching the same number
The prospect has fed hopes in Washington that American strength in technology will lift a manufacturing sector in long decline. "The only way the US is going to regain its role in manufacturing is by applying AI more effectively than China," said Chris Miller, an economic historian at Tufts University. The Pentagon's interest centers on the defense industrial base, which has struggled to expand production against a shortage of munitions and growing competition from China.
China's state venture fund is intended to channel about $137 billion into AI, robotics, and advanced manufacturing, while the flagship US robotics institute runs on about $30 million and a robotics executive order remains unsigned, as we covered in our look at state spending on robotics. On the demand side, Goldman Sachs recently raised its 2035 humanoid robot forecast nearly fivefold to 6.48 million units, a market it sizes at about $138 billion. The two figures are close in size but they are not the same measure. Goldman is sizing humanoid robots only, while Nvidia's $100 billion covers the wider physical AI category, including autonomous vehicles and other robotics.
Nvidia guided to revenue of $108.0 billion for the current quarter, plus or minus 2%.
Sony is moving on the field as well. Its chip unit is bringing about 8,000 workers back to the office, and Sony believes in-person collaboration will help develop new businesses such as physical AI, Nikkei reported.
AMD made its own physical AI bet on September 28, agreeing to buy Fei-Fei Li's World Labs for $8.2 billion in stock.
