Building Data Centers in Space
Watch on YouTube ↗  |  February 13, 2026 at 16:14 UTC  |  7:55  |  Bloomberg Markets
Speakers
Jim Cantrell — CEO of Phantom Space (Co-Founder of SpaceX)

Summary

  • The next frontier for data centers is space, driven by power constraints and environmental impact on Earth; individual satellites currently produce kilowatts, but the goal is megawatts.
  • Launch capacity is the single critical bottleneck ("the moat") for the space economy; demand for launch far exceeds supply, even with the advent of SpaceX's Starship.
  • Hardware in space follows a 5-year replacement cycle (similar to iPhones) due to chip obsolescence, creating a recurring revenue model for launch providers rather than a one-time infrastructure build.
  • Hyperscalers are actively seeking "unique data" residing in space to train Large Language Models (LLMs), as terrestrial data sources become saturated.
Trade Ideas
Ticker Direction Speaker Thesis Time
LONG Jim Cantrell
CEO, Phantom Space (Co-Founder SpaceX)
"The real restriction bottleneck is launch... Launch is a very, very high moat... Investors as they're looking at the future of space and where to invest, you look at those who control launch." In a gold rush (Space Data/AI), the bottleneck is the most valuable position. Launch providers are the "bridge" across the river. Furthermore, Hyperscalers are the primary customers with the capital to pay for this access to secure unique data for AI models. Long the owners of launch infrastructure (SpaceX, Phantom Space) and the capitalized clients driving demand (Hyperscalers). High capital intensity, regulatory delays (FAA/Federal ranges), and technical failure risks inherent to rocketry. 2:03
LONG Jim Cantrell
CEO, Phantom Space (Co-Founder SpaceX)
"What we hear from the HYPERSCALERS is they're looking for unique data... that unique data resides in space... We think that's one of the killer apps [AI]." The AI trade is evolving from "chip manufacturing" to "data acquisition." As terrestrial data becomes commoditized or exhausted, the premium shifts to companies that can harvest and process unique physical-world data from orbit. Long the AI value chain that extends into physical infrastructure and data acquisition. Overvaluation in the AI sector; the speaker notes the industry recently went through an "asset bubble" in 2021/2022 that has since burst and is recovering. 0:05