SpaceX Owns Every Layer of AI Now. The Model Is Still the Weak Link.

📊 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 purchased Cursor for $60 billion, gaining control over all AI infrastructure layers. However, the company’s AI model is still considered the weak link, highlighting ongoing challenges in AI development.

SpaceX has acquired Cursor for $60 billion, giving it ownership of every layer of the AI stack, from hardware to application. This move makes SpaceX a uniquely integrated AI company, but the company’s AI model itself remains a significant weakness, according to industry experts. The deal, announced on June 16, follows a series of strategic investments and acquisitions that position SpaceX as a notable player in AI infrastructure.

On June 16, SpaceX exercised its option to buy Cursor, a profitable AI coding company, for $60 billion in all-stock. Cursor, founded in 2022 by MIT graduates, had rapidly grown to generate approximately $4 billion in annual revenue by June, primarily from its AI coding application, Cursor. The acquisition includes Cursor’s team, its model, and its distribution channels, integrating them into SpaceX’s broader AI ecosystem.

With this purchase, SpaceX controls the entire AI infrastructure: from the supercomputers in Memphis with over 555,000 Nvidia GPUs to its satellite-based data centers, and from its own research labs to the applications deployed by Cursor and other partners. The company’s compute capacity alone is estimated at around 2 gigawatts, making it one of the most extensive AI infrastructure providers globally.

Despite this, industry analysts note that SpaceX’s AI model, Grok, is still considered a weak link in its AI stack. The model’s training efficiency is reportedly low, with internal reports indicating only about 11% utilization of GPU FLOPs, below the typical 35–45% for production-grade models. This inefficiency has led SpaceX’s AI team to lease out its most powerful supercomputers to rivals like Google and Anthropic, generating revenue but also highlighting the model’s ongoing performance challenges.

At a glance
breakingWhen: announced June 16, 2026; deal expected…
The developmentSpaceX completed its $60 billion all-stock acquisition of Cursor, integrating it into its AI infrastructure, but the core AI model still faces performance limitations.
SpaceX owns every layer of AI — the stack, the rentals, the weak link
AI Dispatch · Infrastructure & Strategy

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.

$60B
all-stock · Cursor
(Anysphere)
The stack, layer by layer
06
Distribution
X · Tesla · Optimus · Cursor’s developer base
Strong
05
Application — Cursor
~$4B annualized revenue · just acquired
Bought
04
Model — Grok  ← the weak link
Underdelivered vs compute; training moved to Colossus 2
Weak
03
Research — xAI
Folded into SpaceX, Feb 2026
Mid
02
Compute — Colossus 1 & 2
~555K GPUs · orbital data-center plans filed
Dominant
01
Power
On-site gas generation, built faster than utilities interconnect
Dominant
The landlord pivot — renting Colossus 1 to rivals
Colossus 1 · Memphis
220,000+ GPUs · 300 MW
xAI couldn’t parallelize Grok on its mixed H100/H200/GB200 build, so it moved training to Colossus 2 and leased the rest out.
⚠ ran at ~11% utilization — “embarrassingly low”
Anthropicthru May 2029
$1.25Bper month
Googlethru June 2029
$920Mper month
combined ≈ $26B / year in compute revenue
122
days to build the first 100K-GPU cluster
~555K
Nvidia GPUs across the Memphis site
~2 GW
total power capacity
~$18B
in silicon (phase 1 alone ~$4B)
The take

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.

Sources: SpaceX S-1 & SEC filings; WSJ; Reuters; CBS; TechCrunch; Forbes; Business Insider; Introl; Built In (Feb–Jun 2026). Lease figures per SpaceX filings; utilization per a reported internal xAI memo.
thorstenmeyerai.com

Why Controlling All AI Layers Matters

SpaceX’s control over every layer of the AI stack—hardware, data centers, research, and applications—positions it as a comprehensive player in the AI industry. This vertical integration allows for coordinated development and potential cost efficiencies. However, the ongoing performance limitations of its AI model suggest that infrastructure control alone may not be sufficient for establishing leadership in AI. The leasing of compute resources to competitors underscores the technical challenges associated with optimizing the model’s performance.

For industry observers and competitors, this development indicates a trend toward more integrated AI ecosystems, but it also emphasizes that owning infrastructure does not automatically translate into superior AI capabilities. The performance of the AI model remains a critical factor in determining competitive advantage.

NVIDIA RTX PRO 6000 Blackwell Server Edition

NVIDIA RTX PRO 6000 Blackwell Server Edition

As an affiliate, we earn on qualifying purchases.

As an affiliate, we earn on qualifying purchases.

Background on SpaceX’s AI Expansion

Over the past year, SpaceX has expanded its AI ambitions through strategic investments, including the integration of xAI, its research arm, and the development of the Grok model line. The company built the Colossus supercomputers in Memphis, which have become central to its AI training efforts. These supercomputers are among the largest and fastest in the world, capable of training large models at high speeds, with initial build costs estimated in the tens of billions of dollars.

Prior to the Cursor acquisition, SpaceX had been leasing its supercomputers to major AI labs like Anthropic and Google, generating significant revenue. Internal reports indicate that these labs often lease the hardware because SpaceX’s models are not yet fully optimized, with low GPU utilization rates. The company’s move to acquire Cursor reflects an effort to integrate application development more directly into its infrastructure, but the core model’s performance remains a concern.

“The acquisition of Cursor is a strategic move to enhance our AI capabilities and support innovation across all layers.”

— SpaceX spokesperson

NVIDIA DGX Spark™ - Personal AI Desktop Supercomputer – Desktop GB10 Grace Blackwell Chip

NVIDIA DGX Spark™ – Personal AI Desktop Supercomputer – Desktop GB10 Grace Blackwell Chip

Supercomputer performance directly to your desk in a compact, energy-efficient design, enabling enterprise-scale AI and high-performance computing right…

As an affiliate, we earn on qualifying purchases.

As an affiliate, we earn on qualifying purchases.

Unresolved Questions About Model Performance

It remains uncertain how much SpaceX’s AI model, Grok, will improve following the acquisition of Cursor. Industry insiders suggest that current training inefficiencies limit the model’s capabilities, but the extent of future improvements is not yet clear. Whether SpaceX can address these technical challenges to achieve higher performance levels is an open question.

AI Systems Performance Engineering: Optimizing Model Training and Inference Workloads with GPUs, CUDA, and PyTorch

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.

Next Steps for SpaceX’s AI Strategy

SpaceX is expected to continue integrating Cursor into its AI infrastructure, with plans to deploy the jointly trained model across its applications. The company may also focus on improving Grok’s training efficiency and scaling its AI models. Future regulatory filings and industry disclosures will clarify how SpaceX addresses the model’s current limitations and whether it can translate infrastructure control into AI leadership.

Local AI on Linux in Practice: Build Private LLM Servers, GPU Workstations, Ollama Apps, Dockerized AI Services, and Self-Hosted AI Infrastructure with CUDA, ROCm, vLLM, and Open WebUI

Local AI on Linux in Practice: Build Private LLM Servers, GPU Workstations, Ollama Apps, Dockerized AI Services, and Self-Hosted AI Infrastructure with CUDA, ROCm, vLLM, and Open WebUI

As an affiliate, we earn on qualifying purchases.

As an affiliate, we earn on qualifying purchases.

Key Questions

What does SpaceX’s acquisition of Cursor mean for the AI industry?

This move indicates a trend toward more integrated AI ecosystems, where a single company controls infrastructure, data, and applications. It may influence industry strategies, prompting competitors to consider similar vertical integration or to focus on enhancing their own models.

Internal reports indicate that SpaceX’s Grok model has low GPU utilization, which limits its training efficiency and overall performance. This bottleneck affects the company’s ability to fully leverage its extensive infrastructure for AI development.

Will owning all AI infrastructure layers guarantee success?

Ownership of infrastructure provides certain advantages, but the effectiveness of the AI model itself remains critical. Without improvements in model performance, the potential for leadership in AI may be limited.

What are the implications of leasing supercomputers to rivals?

Leasing hardware to competitors like Google and Anthropic generates revenue but also indicates that SpaceX’s models are not yet fully optimized, highlighting ongoing technical challenges.

What are SpaceX’s future plans for AI development?

Future efforts are likely to include further integration of Cursor, initiatives to improve Grok’s training efficiency, and potential deployment of new models utilizing SpaceX’s extensive compute resources.

Source: ThorstenMeyerAI.com

You May Also Like

Why AI Chatbots Aren’t Winning Over Americans for Money Advice

As someone who is constantly seeking ways to make informed financial decisions,…

Revolutionizing Finance: The Future of AI Integration

The finance industry is undergoing a major shift, fueled by the seamless…

Unveiling the Magic of Machine Learning: An Insightful Guide

Welcome to our comprehensive guide on machine learning, the innovative technology reshaping…

Forward-Deployed Engineer Economics 2.0: The Unit Economics Math, Six Months Later

Six months after initial analysis, FDE economics reveal profitability at scale but challenges at lower levels, impacting enterprise AI deployment strategies.