Cisco Executive Predicts Rise of “Space Data Centers” to Fuel AI Boom
Tech News, Davos,Switzerland, 20 January 2026: In a bold vision shared at the World Economic Forum 2026, Cisco’s President and Chief Product Officer, Jeetu Patel, declared that the next frontier for AI infrastructure lies beyond Earth’s atmosphere. As the demand for generative AI scales toward a projected 800 billion AI agents worldwide, Patel argues that terrestrial constraints on power and cooling will make “space data centers” a logical and necessary evolution.
The Infrastructure Breaking Point
Traditional data centers are currently facing a dual crisis on Earth:
-
The Cooling Tax: Patel noted that roughly 90% of the weight in a modern server rack is dedicated to cooling infrastructure rather than compute power.
-
Power Scarcity: AI models like GPT-6 and Llama 5 require multi-gigawatt clusters that are straining global energy grids to their limits.
The “Orbital Advantage”
By moving compute clusters into Low Earth Orbit (LEO), the industry can bypass these terrestrial bottlenecks:
-
Solar Intensity: Solar panels in space are significantly more productive due to constant exposure and lack of atmospheric interference.
-
Passive Cooling: The vacuum of space acts as a natural heat sink. Patel highlighted that cooling in orbit operates at a “very different level of economic proportion” compared to the water-intensive systems required on Earth.
-
Compression of Innovation: Patel emphasized that the industry is moving at breakneck speed, stating that “what used to happen in 10 years is now happening within six months.”
The Path to 2026 and Beyond
While acknowledging complexities—such as heat dissipation in a vacuum and maintaining high-speed connectivity—Patel revealed that these systems are “actually starting to get built.” Cisco is positioning itself as a key architectural partner, focusing on the Non-Terrestrial Networking (NTN) required to link these orbital factories back to Earth.
This vision aligns with other 2026 “moonshots,” such as Google’s Project Suncatcher, which is currently testing solar-powered TPU constellations to perform high-scale machine learning in orbit.
