THE APEX TIMES
Shaun Maguire, a Sequoia partner, criticized big tech’s approach to AI infrastructure siting near SpaceX and pointed to a model that places most of SpaceX’s future growth in revenue from renting AI compute.
A Sequoia partner on Tuesday used blunt language to criticize how large cloud and AI players are positioning themselves around Elon Musk’s SpaceX, arguing that they are “absolute jokes” for building infrastructure next to the rocket company.
In commentary circulated by Yahoo Finance, Shaun Maguire said SpaceX’s commercial trajectory is increasingly driven not just by launch services or hardware, but by the ability to monetize artificial intelligence compute through rental arrangements. The remarks were tied to a financial model that frames SpaceX’s revenue growth in the late 2020s.
The model, as described in the report, projects SpaceX revenue rising from roughly $51.5 billion in 2026 to about $731 billion in 2028. The key driver of that growth, Maguire’s framework suggests, is “renting out AI compute,” implying a large-scale shift in how value is generated and captured in AI supply chains.
The criticism also specifically names Microsoft, Amazon, and Google as companies building at or near the same physical footprint as SpaceX. Maguire’s characterization is less about product rivalry and more about strategy, arguing that the companies’ approach to infrastructure deployment is mismatched against what he views as SpaceX’s dominant advantage in AI compute monetization.
For Microsoft, the contrast is notable because the company is already heavily associated with AI through Azure, its cloud platform, and through its work to integrate AI models into enterprise software and developer tools. However, the report does not claim that Microsoft has done anything wrong operationally, nor does it detail what, if any, location or project-specific connection exists between Microsoft facilities and SpaceX beyond Maguire’s general point.
The larger context for the cloud and AI sector is that “compute” has become a commodity with expanding demand, and competition is increasingly about who can supply capacity at scale, at competitive cost, and with flexible terms that let customers run training and inference workloads. In that environment, rent-like business models for compute resources can be attractive to customers because they can avoid large upfront commitments, although the practical economics depend on contract terms, capacity utilization, and power costs.
Still, important details are missing from the report itself. The comments and the described model do not provide evidence of specific contracts, disclosed customer commitments, capex plans, or the precise assumptions used to translate “AI compute rentals” into hundreds of billions in revenue by 2028. They also do not provide named data points or citations that would let outsiders verify the magnitude of the projections.
Investors and industry watchers will likely watch whether SpaceX, Microsoft, and other AI infrastructure providers make any more explicit statements about compute rental economics, capacity expansion, or partnerships. Until then, Maguire’s remarks should be read as a high-level strategic opinion paired with an illustrative projection, not as a confirmation of near-term financial outcomes.
The company reported stronger-than-expected fiscal fourth-quarter results, citing iPhone 18 Pro demand and pricing strength, while warning that rising memory costs could weigh on holiday-quarter margins.
The enterprise software company said it intends to fund its expanding artificial intelligence and cloud efforts with a combined equity and debt offering totaling $20 billion.
A new set of requirements, described in a market report, would shift some AI computing demand away from China by restricting how Nvidia-backed capacity is sold or made available through overseas data centers.
In a move aimed at institutional clients, Alphabet is aligning its planned crypto custody platform with growing concern about cybersecurity, including the possibility that future quantum computers could undermine protections used in cryptocurrencies such as bitcoin.
Louis Mosley, who oversees Palantir’s UK and Europe business, said many large enterprises have not yet moved to deploy AI at scale, a delay he linked to slower regional momentum for data and software-driven analytics.
A widely followed Reddit forum’s early trading buzz pointed to opposing tape reads for Intel and Palantir Technologies ahead of Tuesday’s session, underscoring how quickly social sentiment can shape short-horizon expectations.
A recent market discussion points to Apple as the only “Magnificent 7” stock still managing to outperform the broader S&P 500 this year, while several peers including NVIDIA and Tesla are described as lagging.
A market report says Cathie Wood reduced her exposure to Advanced Micro Devices Inc. by selling $32.6 million of AMD stock shortly before the company’s next earnings release.
Investors took a cautious stance in Tuesday trading as Apple’s outlook on margins and Palantir’s reported pace of contract growth raised questions about near-term earnings momentum. With midterm elections looming, risk appetite stayed muted.
The two mega-cap companies that often trade places by valuation are set to do so again after Apple’s guidance failed to impress and Nvidia’s shares held up.