Bezos Says Musk's 2-3 Year Space Data Center Plan Is Too Ambitious
Updated
Updated · CNBC · May 20
Bezos Says Musk's 2-3 Year Space Data Center Plan Is Too Ambitious
1 articles · Updated · CNBC · May 20
Jeff Bezos said orbital data centers are realistic but Elon Musk’s 2-3 year timeline is “probably not right,” arguing the technology will take longer to become viable.
Energy is the main hurdle, Bezos said, with chip costs needing to fall and launch prices needing to drop before space-based computing can make economic sense.
Musk said in February that building orbital data centers was a key reason for merging SpaceX with xAI, underscoring how central the concept is to his AI strategy.
Blue Origin is pursuing its own push: it filed plans in March for 51,600 data center satellites under Project Sunrise and aims to start deploying its TeraWave constellation in Q4 2027.
Can orbital data centers become cost-effective before terrestrial AI becomes more energy-efficient?
Will solving AI's energy crisis on Earth create an unsolvable debris crisis in space?
Who will govern our data in the lawless frontier of space?
The 2026 Orbital AI Infrastructure Race: Technical, Economic, and Regulatory Frontiers in Space-Based Data Centers
Overview
As of May 2026, the race to build space-based data centers is accelerating, with major players competing to deploy orbital AI infrastructure. This competition is driven by the promise of unprecedented computing power and faster communication, which are essential for the next generation of AI applications and global connectivity. The push for distributed, low-latency, and resilient computing is shaping both the pace of technological progress and the diverse strategies companies use to reach this complex goal. Blue Origin, among others, is advancing ambitious plans, highlighting how this intensifying race is redefining the future of computing in space.