Goldman Sachs just released a bombshell forecast that rewrites the investment thesis for humanoid robots, and the implications stretch far beyond the robots themselves. The bank raised its 2035 projection for humanoid robot shipments to approximately 6.5 million units, up from about 1.4 million in its previous estimate. That nearly fivefold increase signals that what was once science fiction is now becoming an industrial reality that Wall Street is willing to bet serious money on.

What Changed in Goldman Sachs' Humanoid Robot Outlook?

The shift reflects three major developments: faster advances in artificial intelligence, falling hardware costs, and growing investment in physical automation across industries. Goldman's revised 2035 estimate values the humanoid market at roughly $138 billion, compared with its previous $38 billion projection. The bank also raised its near-term forecasts significantly, expecting 75,000 units by 2026 (up from 51,000) and 890,000 units by 2030 (up from 256,000).

Goldman still acknowledges real obstacles standing in the way. Reliability concerns, limited real-world training data, autonomy challenges, and cost barriers remain significant hurdles. However, the bank's framing has shifted: these are now viewed as engineering problems to solve rather than reasons to dismiss the entire market.

Where Will Humanoid Robots Actually Get Deployed?

Goldman expects humanoids to gain traction in specific environments where they solve real business problems. Warehouses and logistics facilities represent the primary near-term opportunity because these are structured environments where robots can handle repetitive tasks, yet they contain enough variation that traditional fixed automation struggles. Automotive manufacturing is expected to follow as a second major deployment wave.

Amazon offers a particularly compelling case study. The company already operates more than 1 million robots across more than 300 facilities. Goldman estimates that Amazon's broader automation efforts could generate about $72 billion in cumulative service-cost savings from 2026 through 2030, potentially adding as much as 240 basis points to operating margins in an upside scenario. Tesla also has direct exposure through its Optimus humanoid program, positioning the company to benefit from this industrial shift.

How to Evaluate the Real Winners in Humanoid Robotics

• Semiconductor Content: Each humanoid robot could contain $3,000 to more than $6,000 of semiconductor content, including computing chips, analog and mixed-signal processors, and edge storage. At 6.5 million units, this translates to roughly $19.5 billion to $39 billion of annual semiconductor demand by 2035.
• Broader Ecosystem Play: The opportunity will not belong to a single chipmaker. Goldman's estimates point toward a diverse ecosystem of processors, memory, sensing, and control hardware, making physical AI potentially more valuable to semiconductor investors than betting on any single humanoid manufacturer.
• Infrastructure Suppliers: Companies supplying processors, sensors, motors, and software to the robotics industry may see more stable, predictable growth than the humanoid robot makers themselves, which face winner-take-most dynamics.

This picks-and-shovels approach mirrors how investors profited during the gold rush by supplying miners rather than mining themselves. Nvidia is an obvious beneficiary because physical AI requires increasingly powerful computing at the edge, but the semiconductor opportunity extends across multiple chipmakers and hardware suppliers.

Goldman's nearly fivefold increase to 6.5 million humanoid shipments by 2035 represents a watershed moment in how Wall Street views physical AI. The forecast is not a guarantee; other analysts have made even bolder predictions. Morgan Stanley has predicted 1 billion humanoid robots by 2050, while RBC Capital Markets estimates the market could reach $9 trillion by then. But the numbers are large enough to reshape investment priorities across multiple industries.

For investors, the takeaway is clear: the humanoid robot race is accelerating from speculation to industrial deployment, and the real wealth may flow to the companies supplying the components and infrastructure rather than those betting everything on a single robot design winning the market.