Sending AI data centers into space should dramatically increase revenue and profitability for Nvidia and SpaceX. Both stand to benefit as this industry expands. x.com 

SpaceX's path to an incremental 20 gigawatts of compute next year adds capacity-driven upside to its long-term case, according to Elon Musk. x.com 

SpaceX's AI-cloud ambitions face a severe two-sided constraint: finding 12 GW of Rubin capacity in one year looks implausible, and even that buildout would need customers willing to rent roughly $50B of compute annually. The scale of both the supply and demand challenges makes the plan highly skeptical. x.com 

SpaceX's AI compute buildout has more evidence of demand from big customers. Per Sawyer Merritt, Anthropic's IPO filing says it agreed to spend up to $120.5 billion on SpaceXAI compute through 2029, with payments of about $3.1 billion a month once all capacity is delivered, expected in December 2026. Either side may cancel on 90 days' notice after initial terms end in January and March 2027. x.com  x.com 

SpaceX's compute business is one the market refused to price. Musk said on the earnings call that SpaceX would end 2026 with over 2 gigawatts of compute and reach 10 gigawatts by end-2027, all on Nvidia chips, while most neoclouds are fighting to reach one. With attention on the lockup and the Starship timeline, Morgan Stanley noted the market was assigning zero or negative value to the AI business, on top of a launch monopoly and the Starlink cash machine. x.com 

SpaceX's space-based AI infrastructure is far from proven, so running it at near-perfect reliability will be a challenge on its own. SpaceX has filed plans and announced a design, but it is unclear when the system will be fully functional or what it can do, which makes Musk's "non-zero chance" of $1T in revenue by 2029, cited by @sv_techie, hard to lean on. x.com 

SpaceX's exclusive use of Nvidia GPUs and other supply constraints make Musk's 13-20 GW operational compute forecast for end-2027 too high. A substantial buildout is still incredibly enticing, but reaching that scale likely requires cheap capacity from struggling neoclouds, mostly through capacity deals and some acquisitions. Its attractive $40B off-balance-sheet GPU lease deal is only a drop in the bucket of the debt needed. Capacity can still grow quickly over time. x.com  x.com  x.com  x.com  … Show more

SpaceX's cheap capital and hardware execution give its compute buildout an edge over neoclouds and megacaps. It is quickly becoming the world's largest compute supplier: its 2033 bonds pay 5.65% versus 9.625% on CoreWeave's 2032 notes, and it deployed 110,000 GB300s at Colossus II in 64 days. It targets 10GW to "maybe" 20GW in 2027. One of Anthropic's biggest compute suppliers, SpaceX has up to $120.5B in estimated commitments through 2029, cancellable on 90 days' notice after early 2027. x.com 

SpaceX's IPO has to sell a vision whose weak point is Grok, presented as a success when most of the xAI team has left, so it will need more than faith that Musk will figure it out. The IPO is a funding round dressed as an exit, with proceeds going to AI compute, launch infrastructure and satellites to bring compute to space, and it cements Musk as an unfireable CEO. SpaceX does hold a genuine monopoly in launch, with over 80% of mass to orbit since 2023. x.com 

SpaceX's compute outlook improves as Anthropic commits more spending and institutions fund its GPU expansion. Anthropic's S-1 lists up to $120.5B in SpaceXAI compute purchases through 2029; Pimco bought $4B of the roughly $40B Apollo-led GPU debt package. The GPU lease is expected to close before year-end and entails about $40B in principal payments, not new shares. Elon Musk says Starlink Mobile will take 2-3 years to build out, versus the 4-6 years estimated by Verizon's CEO. x.com  x.com  x.com  … Show more