Written by: Rita
Production of Tesla Optimus has begun in Fremont, but the production guidance of 10,000 to 20,000 units is below the most optimistic market expectations. J.P. Morgan believes that opportunities in the humanoid robot supply chain are shifting from OEMs to component suppliers, with Hengli Hydraulic and Sanhua Intelligent Control as the top choices for Tesla’s robot supply chain.
In a report released on October 22, 2026, J.P. Morgan noted that during Tesla’s Q3 earnings call, Musk confirmed that Optimus has begun production in Fremont and may start shipping to other companies as early as the end of next year. However, Musk explicitly stated that production ramp-up will be gradual, with a production target of 10,000 to 20,000 units before external sales—below the most optimistic market expectations. He described the manufacturing process as an S-curve, emphasizing that “ten thousand things” must be done correctly simultaneously, with the slowest component or process determining the overall ramp-up pace.
Mass production guidance came in below expectations
Elon Musk’s production guidance of 10,000 to 20,000 units is below prior market expectations. JPMorgan noted that this guidance covers a 12-month period and refers to internal production prior to external sales. Musk compared the manufacturing process to an S-curve, where the slowest component or process determines the overall ramp-up speed. Tesla is deliberately reducing its reliance on the Chinese supply chain, which adds complexity to the ramp-up.
JPMorgan believes Tesla's production ramp-up is more cautious than some investors expected. Production has begun in Fremont, but scaling up will take time. Musk did not provide a precise timeline for external commercial deliveries, only indicating it would be approximately one year away. Consumer applications are further off and require first proving reliability, safety, and cost-effectiveness.
The hand is the primary hardware bottleneck.
Hand design, durability, and haptic sensing need improvement, which are critical barriers for industrial and consumer applications. Musk has explicitly identified the hand as the primary hardware bottleneck for Optimus. JPMorgan noted that achieving human-like dexterity, reliability, and long-duration operation remains challenging.
The AI5 chip is scheduled to be integrated into Optimus in April 2027. JPMorgan noted that the AI5’s computing power is approximately 10 times that of the current AI4, enabling more complex and generalized behaviors. Next steps include hardware integration, migration of the software stack, and validation on real-world tasks. Musk also mentioned that videos demonstrating Optimus performing non-predefined tasks will be shown within the next one to two months.
JPMorgan noted that Tesla did not provide specific updates on its vision system, cameras, or battery technology. The lead engineer for Optimus expressed high confidence in the body system, but details on reducers, vision, and batteries remain limited. The hands, onboard intelligence, and computational power roadmap were the focus of this update.
On-premises preferred
Tesla is building Optimus Academy to enable robots to play against themselves and generate training data. Optimus will initially be deployed within Tesla factories to perform dangerous or repetitive tasks. JPMorgan notes that internal deployment allows Tesla to collect proprietary real-world data, validate reliability, and iterate on hardware and software before offering it to external customers.
External commercial shipments are targeted for the end of 2027, with consumer applications to follow later. JPMorgan noted that Musk emphasized there is still a long way to go before humanoid robots achieve industrial-grade automation. Tesla’s deployment strategy is internal first, then external: internal deployment accumulates data and experience, while external sales require proving reliability, safety, and cost-effectiveness first.
JPMorgan Chase’s research at the China International Industrial Expo (CIIF) supports this assessment. The clearest demand comes from suppliers of sensors, 3D vision, data acquisition devices, reducers, and dexterous hands; robot system integrators have not yet become the central focus of demand. Leader Drive’s management stated that physical AI has contributed over 35% of its business and plans multi-year capacity expansion. Orbbec reported a significant increase in robot-related orders, particularly for data acquisition hardware and 3D vision modules.
Supply chain opportunities shift to components
Demand remains in the stages of R&D, pilot testing, and data collection, and has not yet progressed to large-scale end-user deployment. JPMorgan Chase notes that there remains a significant gap between humanoid robot demonstrations and industrial-grade automation. The bottlenecks lie in perceptual consistency, operational reliability, and delivering repeatable ROI within real-world workflows.
J.P. Morgan has become more selective regarding China’s humanoid robot supply chain. Recent gains in Chinese robot system manufacturers, led by Unitree and UBTECH, are viewed by J.P. Morgan as sentiment-driven. As the market absorbs Tesla’s more gradual production outlook, J.P. Morgan remains cautious on system manufacturers in the near term, citing a risk of disappointment over the next 1 to 3 months—particularly for Unitree, Estun, and UBTECH.
JPMorgan continues to favor Hengli Hydraulic and Sanhua Intelligent Control, both confirmed suppliers in Tesla's robot supply chain, as their stock prices have not yet fully priced in optimistic expectations for Tesla's robot. After the pullback since the first half of 2025, their P/E valuations are attractive. Both companies have solid underlying businesses, with valuations near or below their historical cyclical averages.
Green Harmonics is the preferred component choice, with stronger moat, scalability, and physical AI contributions. Orbbec follows closely, with a strong positioning in 3D vision and data acquisition technologies, but recent profitability risks warrant attention. J.P. Morgan’s preference ranking is: Hengli Hydraulic equal to Sanhua Intelligent Control (both overweight) > Green Harmonics (overweight) > Orbbec (overweight) > Inovance Technology (neutral) > UBTECH (neutral) > Estun Robotics equal to Unitree Robotics (both underweight).
Component valuations are more attractive
The P/E ratios of Hengli Hydraulics and Sanyou Intelligent Control are close to or below their historical cyclical averages. J.P. Morgan believes that the recent rally among OEMs has been driven by sentiment, and expectations need to be recalibrated following production guidance that fell short of forecasts. Valuations for component suppliers are more attractive. Green Harmonic’s physical AI contributions are concrete, while Orbbec’s technological positioning is strong, but it carries profitability risks.
On the risk side, Tesla’s production ramp-up has been slower than expected, hand technology bottlenecks remain unresolved, AI5 chip integration has been delayed, external commercial shipments have been postponed, and localizing the supply chain in China has added complexity. JPMorgan has adopted a cautious stance toward OEMs recently but remains relatively positive toward component suppliers.
JPMorgan maintains overweight ratings on Hengli Hydraulics, Sanhua Intelligent Control, Green Harmonic, and Orbbec; neutral ratings on Inovance Technology and UBTECH; and underweight ratings on Estun Automation and Unitree Robotics. Tesla’s Optimus production guidance fell short of expectations, shifting supply chain opportunities from complete units to components. Hengli Hydraulics and Sanhua Intelligent Control are the top picks, followed by Green Harmonic and Orbbec.
Can Optimus be delivered externally by the end of 2027, and when will the hand bottleneck be resolved?
Disclaimer
This article is a compilation and interpretation by Chaoxiang Research of a third-party brokerage research report (J.P. Morgan, October 22, 2026), combined with publicly available market information. The ratings, price targets, earnings forecasts, and related judgments cited herein reflect the views of the brokerage’s analysts and represent the position of their respective institution only; they do not reflect the views of Chaoxiang Research nor constitute any investment advice.
The market carries risks; make decisions independently. This article should not be used as a basis for buying or selling any securities.