Google will pay SpaceX $920M per month for compute
Summary
Google will pay SpaceX \$920 million per month from October 2026 through June 2029 for access to approximately 110,000 NVIDIA GPUs, CPUs, memory, and related components. This deal, announced in a regulatory filing, is similar in scope to SpaceX's earlier agreement with Anthropic, which involves \$1.25 billion per month for compute from its Colossus 1 data center. Google, despite being a major AI compute owner, cites unexpected demand for its recently launched AI products like Gemini Enterprise as the reason for securing this "bridge capacity." The agreement includes a cancellation clause allowing either party to terminate with 90 days' notice after December 31, 2026, and a performance clause for GPU delivery by September 30, 2026. This announcement comes just before SpaceX's anticipated Nasdaq IPO, aiming to raise \$75 billion at a \$1.75 trillion valuation, with Google's existing stake expected to exceed \$100 billion post-IPO.
Key takeaway
For Directors of AI/ML planning future infrastructure, this deal highlights the intense competition for high-end compute. You should proactively assess your organization's projected AI compute requirements and explore long-term capacity agreements with flexible terms. Relying solely on internal build-out may leave you vulnerable to demand surges, potentially delaying product launches or increasing operational costs. Consider diversifying your compute strategy to include external providers.
Key insights
Major tech firms are securing massive, long-term compute capacity to meet surging AI product demand.
Principles
- AI compute demand outstrips even large incumbents' supply.
- Compute deals often include performance and exit clauses.
- Strategic partnerships can bridge capacity gaps.
In practice
- Evaluate future AI compute needs against current capacity.
- Structure compute contracts with termination flexibility.
- Consider external compute providers for demand spikes.
Topics
- AI Compute
- NVIDIA GPUs
- Data Center Infrastructure
- Cloud Partnerships
- SpaceX IPO
- Gemini Enterprise
Best for: CTO, VP of Engineering/Data, Entrepreneur, Investor, Executive, Director of AI/ML
Related on AIssential
Editorial summary, takeaway, and curation by AIssential. Original article published by AI News & Artificial Intelligence | TechCrunch.