TL;DR
Google has entered a deal to pay SpaceX $920 million per month for AI computing capacity at xAI data centers, starting October 2023 and lasting until June 2029. This marks a major infrastructure investment amid Google’s AI expansion and SpaceX’s push into AI services.
Google has officially signed a deal to pay SpaceX $920 million per month for AI compute capacity at SpaceX’s xAI data centers, starting in October 2023 and extending through June 2029. This agreement underscores Google’s significant investment in AI infrastructure and SpaceX’s entry into the AI data center market, with potential implications for the cloud and AI industries.
The deal, as disclosed in a regulatory filing, involves Google utilizing approximately 110,000 Nvidia GPUs, along with other processors and memory, housed within SpaceX’s data centers. The contract spans nearly six years, with capacity ramping up through September 2024 at a reduced fee, and includes provisions for early termination if SpaceX fails to deliver the committed GPU capacity by September 2026.
Google’s spokesperson confirmed that the agreement aims to support the surge in demand for its AI platform, Gemini Enterprise, launched in October. The deal is part of Google’s broader strategy to expand its AI infrastructure amid fierce competition from other tech giants and hyperscalers. SpaceX’s data centers, originally built for Grok, its AI chatbot, are now being leveraged to monetize capacity through third-party agreements, including this one with Google.
Why It Matters
This deal signals a substantial shift in the AI infrastructure landscape, with Google investing heavily in dedicated data center capacity to meet rising demand. It also highlights SpaceX’s strategic move into AI data services, leveraging its existing infrastructure to generate revenue amid its ongoing IPO preparations. For the AI industry, it underscores the importance of scalable compute resources and the increasing role of non-traditional data center providers.
NVIDIA DGX Spark™ – Personal AI Desktop Supercomputer – Desktop GB10 Grace Blackwell Chip
Supercomputer performance directly to your desk in a compact, energy-efficient design, enabling enterprise-scale AI and high-performance computing right…
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Background
Google has been rapidly increasing its AI infrastructure spending, revising its capital expenditure forecast upward to $180-$190 billion in 2023. The company introduced Gemini Enterprise in October, aiming to compete with OpenAI and others. Meanwhile, SpaceX, after merging with Elon Musk’s xAI in February, has been investing heavily in AI, despite facing legal and technical challenges with its Grok model. The recent agreement with Google is part of SpaceX’s strategy to monetize its data centers, which were initially built to support Grok and other AI workflows.
Previously, Google and SpaceX had collaborated in 2019, with Google providing networking and computing resources for Starlink, the satellite internet service. This new deal marks a reversal in roles, with SpaceX now providing infrastructure for Google’s AI needs, reflecting evolving industry dynamics.
“The deal was made to ensure we have bridge capacity to meet surging customer demand for our agent platform, Gemini Enterprise, which has been even higher than we expected.”
— a Google Cloud spokesperson
“This partnership underscores the increasing importance of dedicated AI infrastructure and could reshape how major tech firms approach data center utilization.”
— an anonymous researcher
Data Center Electrical Design: high-performance computing (HPC) facilities
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
What Remains Unclear
It remains unclear how much of the capacity SpaceX will deliver initially, whether the deal will be fully executed as planned, and how this will impact SpaceX’s overall revenue from AI data center services. Additionally, the long-term competitive implications for other cloud providers are still developing.
AI Systems Performance Engineering: Optimizing Model Training and Inference Workloads with GPUs, CUDA, and PyTorch
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
What’s Next
Next steps include SpaceX scaling up its GPU capacity to meet contractual obligations, potential updates on capacity ramping, and further announcements from Google regarding its AI infrastructure expansion. Monitoring SpaceX’s ability to deliver on GPU commitments and Google’s deployment of the capacity will be key.
HP NVIDIA Tesla M60 16GB Server GPU Accelerator Processing Card 803273-001
16GB
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Key Questions
Why is Google investing so heavily in AI infrastructure now?
Google aims to support its expanding AI services, including Gemini Enterprise, and stay competitive with other hyperscalers investing heavily in AI compute capacity.
What does this mean for SpaceX’s business strategy?
It marks a shift toward monetizing its data centers for AI workloads, diversifying revenue streams beyond space launch and satellite services.
Will this deal impact other cloud providers?
Potentially, as it emphasizes the importance of dedicated AI infrastructure, possibly prompting other providers to seek similar partnerships or develop their own capacity.
How does this relate to SpaceX’s IPO plans?
The revenue from this deal could improve SpaceX’s financial outlook, supporting its upcoming IPO and demonstrating its AI infrastructure capabilities.
Source: Hacker News
