The author argues that SpaceX's earnings reinforce the case for continued AI investment by demonstrating accelerating demand and strong pricing for compute. SpaceX shares climbed 7% on Wednesday following strong third-quarter results and a favorable outlook through year-end and into 2027. Elon Musk wants to combine SpaceX with Tesla, Musk's electric vehicle and solar company.
Musk said semiconductor supply no longer represents the main obstacle to expanding compute capacity. Musk identified electricity, cooling and infrastructure as greater constraints on deploying what Musk called super intelligence. In September, President Donald Trump ordered every executive branch department and agency to replace the term AI with SI. Musk has adopted that terminology within SpaceX. Business Insider reported earlier this week that Amazon was rushing to purchase AI hardware despite lacking the power infrastructure needed to operate that hardware.
Musk said SpaceX had reassigned engineers responsible for rocket innovations to address electricity and cooling bottlenecks. Musk claimed SpaceX deploys power and cooling infrastructure much faster than any competitor and might eventually exceed all competitors combined. Musk said SpaceX plans to add between 2.5 and 3 gigawatts of compute power by year-end. Musk expressed reasonable confidence that SpaceX could add another 10 gigawatts in 2027. Musk also described a small possibility that SpaceX could have more than 20 gigawatts of super intelligence compute operating next year. A gigawatt can supply electricity to more than 750,000 homes for one year.
The author argues that SpaceX's targets suggest additional demand for GE Vernova and Eaton, even though SpaceX appears capable of producing some turbine components internally. GE Vernova manufactures heavy-duty gas turbines. Eaton supplies electrical and cooling equipment for power grids and data centers.
SpaceX's AI operations include social platform X, the recently purchased coding assistant Cursor and the Grok chatbot. SpaceX formally introduced Grok Bot in August as an agentic AI product comparable to Meta Platforms' Muse. Agentic AI systems can execute tasks involving several steps with minimal or no human involvement. Musk described Grok Bot as the fastest-growing product Musk had ever encountered. Musk said individual users increase Grok Bot usage over time as the product becomes more useful. President and COO Gwynne Shotwell said Grok Bot reached 800,000 paid weekly users earlier this month, more than doubling during the preceding month. Musk said Grok Bot might eventually account for most of SpaceX's super intelligence compute use.
Meta is due to report earnings on Wednesday night. Muse received a favorable reception after launching in September and rapidly reached the top of Apple's App Store download rankings. The author argues that continued usage matters more than initial downloads and wants Meta to disclose Muse usage, paid subscriptions and implications for compute requirements and spending.
The author argues that Grok Bot's usage supports continued demand for AI accelerators, central processing units and memory chips. Agentic AI needs substantially more central processing units and memory than earlier systems. The author identifies Nvidia's graphics processing units, Intel's central processing units, Micron's high-bandwidth memory and Broadcom's custom chips, including work on Google's tensor processing units, as potential beneficiaries. The author also expects Nvidia and Broadcom to benefit from networking demand as compute systems extend beyond individual chips and servers.
Investors have recently questioned whether AI infrastructure spending will produce sufficient returns to sustain further investment. CFO Bret Johnsen said new compute deployments recover their investment in under one year at current prices. Johnsen said pricing was slightly above the upper end of the $30 to $50 per watt range discussed last quarter. The author interprets those comments as evidence that demand is at least as strong as three months ago. The author argues that Microsoft, Alphabet and Amazon could recover cloud capital spending sooner than Wall Street expects. The author plans to examine those companies' earnings reports next week for comparable comments.
The author remains cautious about aggressively favoring AI investments because political headwinds persist. The author acknowledges Musk's tendency to announce highly ambitious targets and schedules, including the possibility of 20 gigawatts next year. However, the author argues that current demand, supplier pricing power and an unchanged assessment of chips' long useful lives support the AI investment cycle.
SpaceX's AI Compute Expansion and Strategic Shifts
FLASH AIShare
Summary
SpaceX's AI compute expansion is accelerating, with the company targeting 2.5–3 gigawatts of power by year-end and 10 gigawatts by 2027. Q3 results were strong, pushing shares up 7%. Elon Musk highlighted electricity and infrastructure as key constraints, shifting engineers to resolve bottlenecks. Grok Bot, part of SpaceX's AI operations, has 800,000+ paid weekly users and is expected to boost demand for hardware. Musk also mentioned a possible SpaceX-Tesla merger. On-chain news shows growing interest in AI + crypto news as demand for compute resources rises.
Source:Show original
Disclaimer: The information on this page may have been obtained from third parties and does not necessarily reflect the views or opinions of KuCoin. This content is provided for general informational purposes only, without any representation or warranty of any kind, nor shall it be construed as financial or investment advice. KuCoin shall not be liable for any errors or omissions, or for any outcomes resulting from the use of this information.
Investments in digital assets can be risky. Please carefully evaluate the risks of a product and your risk tolerance based on your own financial circumstances. For more information, please refer to our Terms of Use and Risk Disclosure.