📊 Full opportunity report: SpaceX Owns Every Layer of AI Now. The Model Is Still the Weak Link. on ThorstenMeyerAI.com — validation score, market gap, and execution plan.
TL;DR
SpaceX has acquired Cursor for $60 billion, gaining control over all AI stack layers, including compute, power, and applications. Despite this vertical integration, the AI model itself remains a weak point, highlighting ongoing limitations.
SpaceX has completed a $60 billion all-stock acquisition of Cursor, a leading AI coding application, consolidating control over every layer of the AI infrastructure. This move positions SpaceX as a uniquely integrated AI conglomerate, combining hardware, data centers, research, and applications. However, the core AI model remains a weak link, raising questions about the company’s dominance in the space.
On June 16, SpaceX announced it had exercised its option to acquire Cursor, a profitable AI coding startup, for $60 billion in stock, with the deal expected to close in the third quarter of 2026. This acquisition gives SpaceX ownership of all critical AI layers: from the silicon and compute infrastructure to the application level. Cursor, founded in 2022 by MIT graduates, generated approximately $4 billion in annual revenue by June, primarily from its AI coding services that are actively used by paying clients.
SpaceX’s integration includes its supercomputers in Memphis, the Grok AI research team, and its vast compute capacity, which includes the Colossus supercomputers with around 555,000 Nvidia GPUs. The company also owns or leases significant power and data infrastructure, and it has plans to deploy AI satellites as orbital data centers. The purchase of Cursor not only secures a profitable application but also a team of AI developers and a distribution channel, directly linking AI research with hardware and deployment capabilities.
Despite this vertical integration, the core AI model, known as Grok, is considered a weak link. Internal reports reveal that the model’s training efficiency is low, with only about 11% utilization of available GPU FLOPs, far below the 35-45% typical for production models. This inefficiency is attributed to hardware architecture limitations and the complexity of parallelizing training across the existing infrastructure. As a result, SpaceX and its AI team are exploring alternative training arrangements, including leasing compute to rival labs like Anthropic and Google, which pay billions annually for access to the hardware.
SpaceX owns every layer
of AI now
The $60B Cursor buy completes the stack: power, compute, research, model, app, distribution. But owning every layer isn’t winning every layer — and the model is the weak one.
(Anysphere)
You can buy a coding app and a model team. You can’t buy the research lead that makes your foundation model the one everyone else builds on — which is why Anthropic pays Musk $1.25B/month, not the other way around. Owning every layer bought SpaceX the right to attempt the hard thing. It hasn’t done it yet.
Implications of SpaceX’s Fully Integrated AI Empire
This development signals a major shift in AI industry dynamics, with SpaceX becoming the only company to fully own and control every layer of the AI stack. Such vertical integration could lead to increased control over AI development, deployment, and commercialization, potentially giving SpaceX a competitive edge. However, the persistent weakness of the core AI model suggests that owning infrastructure alone does not guarantee superior AI performance. The company’s significant investments highlight both the ambition to dominate AI infrastructure and the ongoing challenge of improving model quality and efficiency.
For industry observers and competitors, this move intensifies the race for AI dominance, raising questions about the sustainability of such integrated models and the importance of model innovation versus infrastructure control. It also underscores the strategic value of owning profitable AI applications, even if the underlying models are not yet optimal, as they generate substantial revenue and user engagement.
AI Systems Performance Engineering: Optimizing Model Training and Inference Workloads with GPUs, CUDA, and PyTorch
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Background on SpaceX’s AI Infrastructure and Cursor Acquisition
SpaceX’s recent push into AI began with its development of the Colossus supercomputers, which drastically accelerated AI training timelines and set new industry benchmarks. The company built the Colossus cluster in Memphis, reaching over 555,000 GPUs in record time, at an estimated cost of billions. Simultaneously, SpaceX’s research arm, xAI, developed the Grok model line, and the company entered into large-scale compute leasing agreements with rivals like Anthropic and Google, which pay billions for access to SpaceX’s hardware.
The acquisition of Cursor, announced on June 16, marks a strategic move to consolidate AI development, combining profitable applications with proprietary hardware and research. Prior to the deal, Cursor had rebuffed offers from OpenAI and Microsoft, emphasizing independence and rapid growth. The move to purchase Cursor reflects SpaceX’s ambition to control the entire AI ecosystem, from silicon to application, positioning itself as a fully integrated AI player in the West.
“This acquisition accelerates our AI ambitions by integrating top-tier hardware, research, and applications under one roof.”
— SpaceX spokesperson
ASUS Ascent GX10 AI Supercomputer, DGX Spark, NVIDIA GB10 Superchip, 128GB LPDDR5x, 1TB PCIe Gen4 NVMe SSD, Wi-Fi 7 & BT5.4, Agentic AI Ready, Supports OpenClaw, NemoClaw, Stackable Chassis
Extreme AI Performance: Powered by NVIDIA GB10 Grace Blackwell Superchip delivering 1 petaFLOP of AI performance and 128GB…
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Unresolved Questions About AI Model Performance
It is not yet clear how effectively SpaceX can improve the Grok model’s training efficiency and overall performance in the near term. Internal reports indicate low GPU utilization, but the company has not publicly detailed plans for significant model upgrades or breakthroughs. Additionally, the long-term impact of leasing compute to rivals and the strategic value of owning profitable applications versus developing superior models remain uncertain.
Hands-On Generative AI with Transformers and Diffusion Models
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Next Steps in SpaceX’s AI Strategy and Development
SpaceX is expected to focus on optimizing its AI models, potentially through new training techniques or hardware configurations, to address current inefficiencies. The company will likely integrate Cursor more deeply into its ecosystem, expanding its AI applications. Regulatory and competitive responses also remain to be seen, especially as other tech giants and AI labs respond to SpaceX’s expanding infrastructure and vertical integration. The deal’s closing in Q3 2026 will mark a new phase of operational and strategic planning.
Corsair AI Workstation 300 Desktop PC – AMD Ryzen AI Max 385 CPU – AMD Radeon 8050S iGPU (Up to 48GBs vRAM) – 64GB LPDDR5X 8000MHz Memory – 1TB M.2 SSD – Black
AI-Optimized Compact Workstation: Experience AI performance out of the box with the compact 4.4L form factor, built for…
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Key Questions
Why did SpaceX buy Cursor for $60 billion?
SpaceX purchased Cursor to own a profitable AI application, a developer distribution channel, and a team of AI experts, integrating these assets with its hardware and research capabilities to build a fully integrated AI ecosystem.
Does owning all AI layers guarantee better AI models?
Not necessarily. While vertical integration provides control and potential cost advantages, the current AI model, Grok, remains inefficient, showing that infrastructure alone does not ensure superior AI performance.
What are the risks of SpaceX’s approach?
The main risks include over-reliance on existing hardware and models that are not yet optimized, potential regulatory scrutiny, and the challenge of continuously improving AI capabilities while maintaining profitability.
How does this affect the broader AI industry?
This move sets a precedent for full-stack AI ownership, intensifying competition and possibly prompting other tech giants to pursue similar vertical integration strategies.
What is the significance of leasing compute to rivals?
Leasing compute to competitors like Anthropic and Google generates substantial revenue, but it also reveals the underlying inefficiency of the current AI models and the strategic importance of hardware infrastructure.
Source: ThorstenMeyerAI.com
