Elon Musk Replaces Executives Over SpaceX AI Data Center Reliability Issues
Elon Musk restructured SpaceX's AI data center construction team and replaced several senior executives following persistent engineering and reliability issues at facilities in Tennessee and Mississippi. Executives from SpaceX's rocket production and Starbase operations, as well as civil engineers from Neuralink, have been reassigned to stabilize the infrastructure. The reshuffle highlights the severe engineering trade-offs between rushing AI hardware deployment and maintaining critical infrastructure stability as compute scales to gigawatt levels. It also threatens to delay SpaceX's planned data center expansions in Texas and other regions, which could impact compute leasing agreements with partners like Google and Anthropic. To accelerate early deployment, SpaceX operated AI facilities for months without installing backup cooling and redundant power supply systems, significantly increasing outage risks. Despite these operational setbacks, SpaceX reached approximately 1.4GW of operational compute capacity by June with plans to exceed 2GW by year-end, generating $2.6 billion in Q2 compute leasing revenue.
## BACKGROUND
Modern AI data centers housing tens of thousands of GPUs require immense electrical power and advanced liquid cooling infrastructure to run uninterrupted. Building compute clusters at gigawatt scale pushes power grids and thermal management systems to their physical limits. Redundant backup generators, uninterruptible power supplies (UPS), and secondary cooling loops are standard industry safeguards required to prevent hardware damage and catastrophic service outages when primary systems fail.