|

SpaceX Plows $40B Into Nvidia Chips For AI Cloud

ByNoel Saldanha
Illustration for SpaceX Plows $40B Into Nvidia Chips For AI Cloud

SpaceX locked in a $40 billion lease backed by Apollo, the investment firm, to buy NVIDIA Vera Rubin chips for AI cloud services starting in 2027.

Key Takeaways

  • SpaceX reached a Binding Commitment for an approximately $40 billion master finance lease for GPUs with an entity managed by Apollo Global Management
  • The lease will deliver next-generation NVIDIA Vera Rubin chips for SpaceX’s terrestrial compute buildout
  • Delivery schedules are expected to commence in the first half of 2027, subject to closing conditions
  • SpaceX said the Binding Commitment includes institutional managers and a Goldman Sachs structuring agent, with additional terms to appear in an 8-K

Binding Commitment, Why SpaceX Is Betting On Terrestrial Compute

SpaceX said the lease will deliver next-generation NVIDIA Vera Rubin chips for its terrestrial compute buildout and framed the purchase as a step to deploy the world’s largest compute infrastructure to support Grok and enterprise AI customers, tying the move to near-term payback projections and institutional financing partners.

How The nvidia vera rubin chips Lease Works And When It Closes

SpaceX’s update says the lease is a master finance commitment with multiple financing parties, is subject to closing conditions, and that delivery schedules are expected to commence in the first half of 2027.

The company also said additional terms will be disclosed in an 8-K filed with the SEC and described the arrangement as a Binding Commitment that includes institutional managers and a Goldman Sachs structuring agent.

Also Read: Kazakhstan Telecom Giant Fires Up 880 Nvidia Chips To Build Local AI

What This Means For The AI Compute Market

A $40 billion committed GPU lease would be the largest GPU financing on record and will shift equipment supply dynamics, vendor contract power, and data-center buildout pacing. The scale boosts demand for NVIDIA’s Vera Rubin family and increases competition for skilled datacenter rack space, power, and networking, and companies that sell cloud GPU time or colocate pods now face a deeper well of capacity from a nontraditional provider with enterprise sales ambitions.

From Hyperscaler Builds To New Compute Entrants

SpaceX GPU expands a trend where vertically integrated firms finance and deploy hyperscale AI compute outside traditional cloud providers.

The transaction resembles recent bank-and-asset-manager-backed deals that underwrite long-lived hardware for AI companies, but at a larger scale. Watch for the upcoming Form 8-K for legal terms and for vendor scheduling that will show how much capacity arrives in H1 2027.

What To Watch Next

Monitor SpaceX’s SEC filings for the 8-K describing final lease terms and any equipment schedules, and watch NVIDIA statements about Vera Rubin shipments.

Also track announcements from Apollo, Goldman Sachs, and anchor investors for allocation details and any carve-outs for cloud customers.

Read Next: Honeywell Gives Factories AI That Catches Breakdowns Before Shutdowns

Related: Oracle Hands Google AI Chips To Cloud Buyers Eyeing Nvidia

Similar Posts