The Optimus robot is the part of Jeff Brown’s 70X AI Agent thesis that scales beyond the car. Musk has said he wants to build one million humanoids at a target price of roughly $25,000 each. If Tesla hits even a fraction of that target, the demand for the components inside each robot becomes the investment thesis. This piece covers the broader component stack a humanoid robot needs and where the supply chain pressure lands. The campaign-level framing and the optical bottleneck angle are covered in the parent 70X AI Agent analysis; this one stays on the humanoid stack.
The Optimus production target
Musk’s stated goal is one million Optimus robots at a $25,000 price point. The target matters because it sets the demand curve for every component inside the robot. One million units is automotive-scale production applied to a machine that, today, is hand-assembled in low volumes. The gap between current output and the stated target is where the supply chain thesis lives.
Musk’s production projections have historically run long. The Cybertruck missed its initial timeline by years. The original Model 3 production ramp hit what Musk called “production hell” in 2017 and 2018. The Semi, announced in 2017, did not begin deliveries until 2022. The Optimus timeline inherits that history. Whether the one-million-unit target is reached in 2026, 2027, or later is the variable that sets the pace of demand for the component stack.
The thesis requires Optimus to scale from prototype to production at any meaningful volume, because the components inside each robot are the same regardless of how many units ship. The supply chain logic rewards the component supplier whether Tesla builds 100,000 robots or one million, because the supplier ships into the ecosystem rather than to a single order.
The components inside a humanoid
A humanoid robot has four major component systems. Each has a supply chain, and each has a bottleneck layer where the investment thesis concentrates.
The vision system is the first. Optimus uses camera-based perception, the same architecture Tesla’s FSD uses on the road. The robot needs to see and navigate its environment, recognize objects, and track motion. The optical sensors, the image signal processors, and the inference compute that turns raw vision into decisions are the same component stack Brown identifies as the bottleneck for FSD. That is the connection between the Optimus thesis and the 70X AI Agent campaign. The same hidden supplier that gives FSD its eyesight gives Optimus its eyesight.
The actuator system is the second. A humanoid robot has roughly 28 degrees of freedom across its joints, each driven by an electric actuator. The actuators translate electrical commands into physical motion, and they are the component that determines how strong, how fast, and how precise the robot’s movements are. The supply chain for actuators is mature, because industrial robotics has used electric actuators for decades. The bottleneck for Optimus is not whether actuators exist, but whether they can be produced at the cost and weight the $25,000 price point requires. Tesla has discussed custom-designed actuators for Optimus rather than off-the-shelf parts, which is an indication that the existing supply chain does not meet the cost target at scale.
Memory and storage are the third. Every robot needs onboard memory for the inference model and storage for the operating system and task data. The memory supply chain is dominated by a small number of producers, and the demand from a million-robot buildout compounds with the demand from autonomous vehicles, data centers, and consumer electronics. Memory is a commodity input rather than a differentiated component, but the suppliers are a known list and the demand curve from robotics is incremental to the existing market.
Power is the fourth. Optimus runs on a battery pack, and the energy density of that pack determines how long the robot can operate between charges. Tesla’s battery expertise from the vehicle business transfers directly, and the battery supply chain is the same one feeding the automotive line. The bottleneck here is cell production capacity, because every Optimus battery competes for factory output with every vehicle battery Tesla builds.
Where the bottleneck sits
The four component systems have different supply chain structures. Memory and battery cells are commodity inputs produced by large, well-known manufacturers. The market already prices the demand from data centers and vehicles, and the incremental demand from robotics is a marginal addition to a large existing market. The investment thesis in these layers is a volume play, not a bottleneck play.
Actuators are a custom design problem. Tesla’s decision to engineer its own actuators for Optimus suggests that the existing robotics supply chain does not meet the cost, weight, and integration requirements for a $25,000 humanoid. The investment opportunity in this layer is in the component manufacturers who can supply the subcomponents, the magnets, the windings, and the precision gear sets that go into a custom actuator design. The market for these subcomponents is fragmented, and the supplier who wins the Optimus design cycle wins a meaningful volume contract.
The vision system is where Brown’s thesis concentrates. The optical hardware that gives Optimus its eyesight is the same component stack that gives FSD its eyesight, and the supplier who makes that hardware is the one position that scales across both the automotive and the robotics buildout. Every Cybercab and every Optimus needs the vision system, so the demand compounds across product lines rather than running in parallel, which is the structural argument for the hidden supplier as the picks-and-shovels position.
The competitor landscape
Tesla is not the only company building humanoid robots. Boston Dynamics has been in the space for years. Figure AI raised $675 million at a $2.6 billion valuation. Agility Robotics, Apptronik, and a growing list of startups are all developing humanoids for industrial and commercial deployment.
Each of them needs the same component categories. Vision, actuation, memory, and power are the inputs to any humanoid, regardless of who builds it. The supplier that solves perception for one robot can solve it for all of them. That is the picks-and-shovels logic at the industry level rather than the brand level. The supplier to the humanoid industry is the position that benefits from the entire category scaling, not from Tesla alone winning the market.
Brown’s track record on this method runs through the Nvidia call in 2016, when he identified the GPU maker as the hidden supplier behind the machine learning revolution before the market re-rated the stock. The structure is the same: find the enabling component, identify the manufacturer, and invest before the market prices in the demand. Past performance does not guarantee future results, and the returns cited in Brown’s published record are calculated on public market data from publicly stated entry points.
The investment angle
The Optimus supply chain thesis is a robotics industry buildout story, not a Tesla-specific story. The components inside a humanoid are the inputs to a category, and the suppliers of those components benefit from the category scaling regardless of which robot maker wins the deployment race.
The vision system is the layer where the 70X AI Agent campaign concentrates, because it is the component that scales across FSD and Optimus simultaneously. The actuator layer is a custom design problem where the supplier who wins the Optimus design cycle captures a meaningful contract. Memory and power are commodity inputs where robotics demand is incremental to large existing markets.
Musk’s production timelines run long, and the one-million-unit Optimus target inherits that history. The thesis does not require the full target on schedule. It requires Optimus to scale from prototype to production at any volume, because the component supplier ships into the ecosystem. The structural case for a vision system supplier to the humanoid category is the part of the thesis that holds independent of whether Tesla hits its specific production number. Which company occupies that position is the question the paid report answers, and the structural case is the part that survives whether that specific pick is right or not.
The Promo Watch board tracks the rest of the current campaigns we cover.