What happened
SpaceX facilities in Tennessee and Mississippi ran without backup cooling and power for months, interrupting AI model training, according to The Information.
Why it matters
It spent about $15.8 billion on AI capex in Q2 alone and scaled compute from 0.4 gigawatts to 1.4 GW in a year, but reliability lags its 99.9% uptime target.
What to watch
The Sept. 30 deadline is key—Google may terminate, accept fewer GPUs, or cut payments after a one-month grace period. Watch whether SpaceX delivers on time.
WHO IT HITSSpaceX data center executives are being replaced by leaders from its rocket and Starlink units; investors in SpaceX and its partners (like Alphabet and Nvidia) face risk if the Google deal is jeopardized.
Ask the AI about this article →
Summaries like this, in your inbox every morning.
SpaceX's rapid data center build-out has collided with the realities of physical infrastructure. The company pushed capacity to 1.4 GW from 0.4 GW in a year to cash in on AI demand, but the pace led to temporary power and cooling outages that interrupted model training. This puts the company in a bind: it needs speed to meet a Sept. 30 deadline for Google's GPUs, but also must prove it can run reliable facilities.
The company's response—delaying projects in Texas, manufacturing its own turbine components, and reshuffling leadership with execs from its rocket and Starlink units—suggests it is prioritizing reliability without sacrificing schedule. Elon Musk believes in-house manufacturing could bring gas turbines online 18 months faster, calling it a 'profound game changer.'
The stakes are enormous: roughly 90% of SpaceX's $28 trillion addressable market opportunity relates to AI, and Google's $920 million-a-month contract is a flagship deal. Success hinges on whether the new leadership and operational changes can stabilize uptime without blowing the deadline. If not, Google's exit clauses pose a direct risk to the company's AI revenue ambitions.
For example, today's edition would include:
AI-summarized, only the topics you pick — one digest a day via Email, Slack, or Discord.
Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →
Ask AI anything about this article. Q&As are published on this page for other readers too.
Michael Burry, famous for The Big Short, is short Nvidia, Palantir and Tesla, and in his Substack newsletter s…
Investors have three creative routes to Anthropic exposure before its expected IPO: buying Alphabet, Amazon, o…
Much of the attention on AI infrastructure buildouts is now tied to sheer compute power, with dominance define…
Barron's reported September 10 that Kepler Computing emerged from stealth with a memory architecture using fer…
Dynatrace acquired Arize AI, adding AI observability, evaluation and agent monitoring to its application obser…
Reuters reported September 10 that inference-chip startup d-Matrix will use Nvidia's NVLink Fusion to connect…
