Goldman Sachs Raises 2026 Humanoid Robot Shipment Forecast to 75,000 Units
nashnova research
Goldman Sachs raised its 2026 humanoid-robot base-case shipment forecast from 51,000 to 75,000 units in an 80-page physical-AI report, projecting 6.5 million units and a $138 billion market by 2035 — with warehouse logistics as the first commercial beachhead and a structural pull-through for the semiconductor supply chain.
How big is the forecast revision?
The 2026 base case rose from 51,000 to 75,000 units — roughly a 47% increase.
The 2030 figure jumped from 256,000 to 890,000 units; the 2035 figure from 1.4 million to 6.5 million units, representing a market worth roughly $138 billion.
This means → Goldman no longer treats humanoid robots as a distant concept. The bank now sees them as a commercially scalable category within five years.
Why does warehouse logistics come first?
Amazon has deployed over 1 million robots across more than 300 facilities. Walmart has automated freight operations in 3,100 U.S. stores and processes over half its e-commerce orders through automated facilities.
Goldman estimates that broader Amazon robot deployment by 2030 could unlock roughly 5.6% leverage on total cost-to-serve — up to $72 billion in cost savings and about 240 basis points of EBIT margin improvement in an upside scenario.
In plain terms = warehouses already run millions of simpler robots for moving and sorting. Humanoid robots are the next step — they handle more complex tasks, and the savings drop straight to the bottom line.
Can the auto industry keep pace?
Body welding is already highly automated, but final assembly and parts sorting still rely on human labor — exactly the gap humanoid robots target.
If automakers adopt humanoid robots — priced at $20,000–$60,000 — at penetration rates of 10%–50%, the gross-level impact is a 1%–6% operating-margin uplift.
This means → the auto sector doesn't just "might use" humanoids — it already has a clear economic case. The only open variables are unit price and reliability.
How does the semiconductor supply chain benefit?
Goldman estimates $3,000–$6,000+ in semiconductor content per humanoid robot: high-performance compute modules — the robot's "brain" chips — at $1,500–$4,000+, analog and mixed-signal chips — sensor chips that perceive the physical world — at $750–$1,050+, and edge storage at $600–$800+.
In plain terms = every additional robot sold means $3,000–$6,000 in chip revenue. At scale, this creates an entirely new demand line for the semiconductor industry.
Who gets disrupted by software-defined controllers?
Humanoid-robot mobility is accelerating a shift from traditional PLCs — programmable logic controllers, dedicated hardware boxes that run factory equipment — to vPLCs — virtual PLCs that replace the hardware box with software.
Goldman projects the vPLC market will grow at 20%–30% annually.
This reflects a direct challenge to Siemens, Rockwell Automation, and Schneider Electric — incumbents whose business models bundle hardware and software. Pure-software competitors are now targeting the same control layer.
Can these forecasts actually be met?
A concurrent Goldman enterprise survey found that roughly 40% of executives expect at least 10% of their workflows to be automated by general-purpose robots within three to five years.
This means → corporate purchase intent already exists, but whether warehouse deployments deliver scalable unit economics is the key validation point for this round of bullish forecasts.
Put simply = forecasts are forecasts. The market ultimately asks one question: once humanoid robots go into warehouses, do the numbers actually work?
市场有风险，内容仅供研究参考，不构成投资建议。