Sundar Pichai, CEO of Google, announced on Fox News Sunday that the company intends to begin building data centers in outer space, powered by solar energy, as early as 2027.
According to Pichai, this is a response to the growing energy demand generated by the development of artificial intelligence. “We’re taking the first step in ’27,” Pichai said on November 30. “We’ll send tiny, tiny racks of machines, put them on satellites, test them, and then we’ll start scaling from there”[4].
Earlier, at the beginning of November, Google unveiled Project Suncatcher – a research initiative aimed at advancing machine learning in space. As part of the project, the company is collaborating with Planet, a firm specializing in satellite imaging. Together, they plan to launch two test satellites that will evaluate hardware and optical communication systems in Earth’s orbit. Pichai predicts that within a decade, space-based data centers will become “a more normal way of building data centers”[4][6].
A New Era of Data Centers – From Earth to Orbit
Data centers are increasingly scrutinized for their environmental impact. According to a report from the U.S. Department of Energy from December 2024, in 2023 data centers consumed 4.4% of the total electricity in the United States, and by 2028 this share could rise to as much as 12%[2]. Over the past five years, Google has more than doubled the energy consumption of its data centers – from 14.4 million megawatt-hours in 2020 to 30.8 million in 2024[1].
Golestan Radwan from the United Nations Environment Programme points out: “We still don’t know much about the environmental impact of AI, but some of the data we have is concerning. We need to ensure that the net effect of AI on the planet is positive before we deploy this technology on a large scale”[4][6]. Pichai emphasizes that outer space offers access to “a hundred trillion times more energy than we currently generate on all of Earth” thanks to constant sunlight. Satellites in heliosynchronous orbits can collect solar energy almost continuously, up to eight times more efficiently than ground-based systems[6][7][8][9][10][4].
Environmental Impact and the Technology Race
Google is not the only company developing off-Earth computing infrastructure. The startup Starcloud, backed by Y Combinator and Nvidia, launched a satellite equipped with AI on November 2, featuring an H100 GPU processor that offers a hundred times the computing power of previous space systems.
According to Starcloud CEO Philip Johnston, space-based data centers could generate ten times lower carbon dioxide emissions than their terrestrial counterparts[1][2][3][4]. It is also worth noting that Google is heavily investing in ground-based infrastructure – at the beginning of November, the company announced a $40 billion investment by 2027 to build new data centers in Texas, marking Google’s largest investment in a single U.S. state[5][6][7][1].
The company also recently unveiled the Gemini 3 model, which quickly achieved high rankings in performance benchmarks. According to an April report by McKinsey, the technology race in AI could mean that by 2030, ground-based data centers will require over $5 trillion in capital expenditures[8][9][10][11][1].