Morgan Stanley Details Deepening Tesla-SpaceX Physical AI Convergence Across Compute, Chips, and Starlink
A research report from Morgan Stanley analysts led by Adam Jonas highlights accelerating operational and technological convergence between Tesla ($TSLA) and SpaceX ($SPCX), arguing the two companies are uniquely positioned to jointly capture the emerging "Physical AI" market across robotics, automated transport, and distributed energy.
The note frames the two entities as operating under a unified mission of "converting energy into intelligence," noting that as pure-play large language models (LLMs) confront scaling and performance limits, the next phase of enterprise value creation will be determined by deploying AI into physical-world machines and industrial environments.
Key operational touchpoints and cross-company synergies detailed in the report include:
- Semiconductor & Compute Infrastructure: Tesla and SpaceX are partnering with Intel ($INTC) on a joint semiconductor fabrication initiative in Texas called "Terafab," with total project investments estimated up to $119 billion. On the compute front, Morgan Stanley projects SpaceX could scale dedicated compute capacity to 4.9 gigawatts by late 2027 and reach 15.3 gigawatts by 2031 to support training workloads for xAI and autonomous fleets.
- Starlink Connectivity & Edge Robotics: Tesla is integrating Starlink low-Earth-orbit satellite terminals across its vehicle ecosystem, starting with the Cybercab, providing low-latency edge connectivity to support Robotaxi networks in cellular dead zones.
- Cross-Procurement & Agent Platforms: SpaceX and xAI procured over $506 million in Tesla Megapack utility storage and $131 million in Cybertruck utility fleets in 2025 to power the Colossus compute cluster. The two entities are also co-developing "Macrohard," an agentic software platform for robotics and autonomous execution.
- Equity & Manufacturing Integration: Tesla holds approximately 18.99 million shares of SpaceX Class A common stock (originating from its earlier xAI positioning), while providing a massive industrial manufacturing base of over 135,000 employees to scale hardware designs.
From an equity perspective, Morgan Stanley reiterates Overweight ratings on both names ($400 price target on Tesla, $300 on SpaceX), arguing that while stopping short of proposing a formal corporate merger, the tight integration of energy, orbital comms, compute, and volume manufacturing creates a virtually irreplicable vertical moat in physical robotics.
