SpaceX Faces Reliability Issues as AI Data Centers Expand Rapidly
SpaceX has encountered significant reliability challenges while rapidly expanding its AI data center infrastructure, with some facilities operating without backup power for months. Engineering and supply chain issues have affected sites in Tennessee and Mississippi, leading to outages that disrupted AI model training. The company is investing heavily—$15.8 billion on AI-related capital expenditures alone in Q2—to scale compute capacity from 0.4 gigawatts to 1.4 GW over the past year. The urgency stems from a $920 million-per-month contract with Google, which requires SpaceX to deliver around 110,000 Nvidia GPUs by September 30. Failure could result in reduced payments or termination of the agreement. To address these issues, SpaceX has delayed new data center projects outside its Memphis hub and reshuffled leadership. Despite significant losses—$3.7 billion for AI operations in H1—the company's aggressive expansion is driving rapid revenue growth, with second-quarter AI revenue exceeding $2.5 billion.
Tickers: $GOOG $GOOGL $NVDA $SPCX
Topics: AI Data Centers Alphabet/Google Contract Backup Power Failures SpaceX