Google has announced a research initiative called Project Suncatcher to put artificial intelligence data centers into space to meet the demand for AI computing power.
Scientists and engineers believed that tightly packed constellations of about 80 solar-powered satellites could be placed in orbit about 400 miles above the Earth’s surface stocked with the powerful processors required to meet increasing demand for AI.
According to Google research data released on Tuesday, prices of space launches are falling so quickly that by the middle of the 2030s the operating expenses could be comparable to one on Earth.
Meanwhile, using satellites could mitigate the impact on the land and water resources required to cool previous datacenters.
The orbiting data centers envisioned under Project Suncatcher would report through optical links, which commonly use light or laser beams to transmit information.
The giant tech companies accelerating development in AI are projected to spend $3tn on earthbound datacenters from India to Texas and from Lincolnshire to Brazil.
The spending has intensified concern about the impact on carbon emissions, given the lack of clean energy found to power the sites.
Google said, “In the future, space may be the best place to scale AI companies.”
The co-founder of the startup, Philip Johnstron said, “In space, you get almost unlimited, -low-cost renewable energy.”
He further explained, “The only cost on the environment will be one the launch and there will be 10 times carbon dioxide savings over the life of the datacenter compared with powering the datacenter terrestrially.”
Google is planning to launch two prototype satellites by early 2027 to achieve a significant milestone and present a sustainable and scalable solution to the exponential growth of AI’s increasing energy demands.
