SpaceX is restructuring its data center infrastructure in Tennessee and Mississippi following operational failures. Facilities previously lacked complete backup power and cooling, leading to outages that disrupted AI training for clients including Google and Anthropic. The company failed to maintain its target uptime of 99.9 percent.
The technical issues stemmed from prioritizing rapid scaling over stability, utilizing temporary solutions like gas turbines and Tesla Megapacks. SpaceX is now mandating the installation of permanent backup systems during construction and implementing rigorous pre-operational testing. These changes follow the departure of several data center executives.
To address the shortfall, SpaceX reassigned over 300 engineers to the data center division from a pool of 1,300 internal applicants. As of June 2026, the company reported 1.4 GW of compute capacity and aims to exceed 2 GW by the end of the year, marking a 43 percent increase.
Financial stakes remain high, with external compute rental revenue reaching $2.6 billion in Q2 2026. A contract with Google involves $920 million per month for access to 110,000 Nvidia GPUs, with a delivery deadline of September 30, 2026. Supply chain constraints have already forced the postponement of certain expansion projects.
