Elon Musk said on Tuesday that demand for supercomputer data centers remains strong, but warned that raising money for massive AI infrastructure does not guarantee companies can actually build and operate the systems.
"It’s not that supercomputer data centers are not in high demand, as they definitely are," Musk wrote on X. "However, getting money to build complex hardware systems with massive power, cooling & networking is not the same as doing it."
Musk was responding to a post highlighting stress in AI-related credit markets, where investors have increasingly demanded higher yields to finance data-center expansion. Reuters reported that rising rates and worries about returns have begun pressuring AI-linked debt, including financing tied to CoreWeave Inc. (NASDAQ:CRWV), Oracle Corp. (NYSE:ORCL) and large hyperscaler projects.
That pressure comes after an extraordinary borrowing wave. CoreWeave and Oracle have already emerged as two major borrowers in the AI infrastructure buildout, while Reuters has highlighted growing investor scrutiny over whether AI projects can generate enough cash flow to justify their financing costs.
Musk framed the problem as one of execution, not demand. Space Exploration Technologies Corp. (NASDAQ:SPCX) “has a super cracked team that is outperforming the rest of the world by an order of magnitude in mass to orbit and even we find it very difficult!" he wrote.
His own companies are pushing aggressively into data-center infrastructure. SpaceXAI says its Colossus supercomputer was built in 122 days and later doubled to 200,000 Nvidia H100 GPUs, with a roadmap toward one million GPUs. Musk has said Colossus 2 could eventually run more than 1.2 million Nvidia GB200 and GB300 chips.
Tesla Inc. (NASDAQ:TSLA) is also expanding on-site AI compute. Its second-quarter filing said Cortex 1 and Cortex 2 in Texas provide more than 205 megawatts of production capacity, with further expansion planned through year-end.
Musk’s ambitions extend beyond Earth. SpaceX says orbital data centers could eventually use continuous solar power, radiative cooling and Starlink laser links to avoid some of the land, grid and cooling constraints facing terrestrial facilities. Musk has argued that space is "the only way to scale at scale".
Orbital computing also faces major engineering and economic challenges. Reuters reported in April that experts warned space-based data centers could face extreme launch costs, difficult cooling requirements, limited hardware-upgrade options and the harsh space environment.
Photo courtesy: Shutterstock