SpaceX computing power deal with Reflection worth $6.3B
SpaceX computing power deal worth up to $6.3 billion: the company will supply Nvidia GB300 GPU capacity to open-source AI startup Reflection AI via its Colossus 2 data center. The SpaceX computing power deal is structured as $150 million per month starting July 1, 2026, running through 2029.
Reflection gets access to Nvidia GB300 chips and the Colossus 2 facility, originally built to train and run xAI’s Grok models. The contract includes a termination clause with 90 days’ notice after the first three months, but it still implies at least a quarter of payments are locked in before any exit.
Reflection AI was founded in 2024 by former DeepMind researchers Misha Laskin and Ioannis Antonoglou. The startup focuses on open-source AI models and coding automation, has raised about $800 million, and is valued up to $25 billion, with Nvidia as a notable backer.
Broader compute market impact: at $150 million per month, this SpaceX computing power deal creates a clear pricing benchmark for large-scale GPU access. Decentralized GPU marketplaces such as Render, Akash, and io.net have argued they can undercut centralized providers; this contract gives them a concrete target. Traders should watch whether Reflection can sustain its valuation while absorbing heavy compute costs—its need for strong revenue or continued fundraising could shape sentiment around AI infrastructure plays.
Neutral
This is primarily a corporate AI-compute deal, not a direct crypto protocol or token-driven catalyst. Still, it can indirectly affect crypto-related “AI compute” narratives. In the short term, traders may price in expectations that centralized GPU pricing benchmarks ($150M/month) force decentralized compute platforms to sharpen cost/throughput competitiveness—potentially supportive for AI-compute token sentiment, but not strong enough to be clearly bullish without follow-on demand signals.
In the medium/long term, the key risk is Reflection’s economics: sustaining a valuation up to $25B while burning heavily on infrastructure could require continued fundraising or strong product revenue. If fundraising tightens or margins disappoint, sentiment toward AI infrastructure plays could soften. Historically, large enterprise AI compute announcements tend to create brief narrative spikes in related markets, but sustained price effects usually require visible monetization (contracts, usage growth, or network profitability) rather than headlines alone.
Overall, the event is informative for the AI compute sector and marginally relevant to AI-compute crypto themes, but it doesn’t directly change broader crypto market stability (e.g., liquidity, regulation, or major exchange flows). Hence the expected impact is neutral.