Data Centers? In Space?! Startup Raises $250 Million
Newsweek · C · trust 47/100
0 Share Newsweek is a Trust Project member See more of our trusted coverage when you search. Prefer Newsweek on Google to see more of our trusted coverage when you search. A startup seeking to move AI data centers into space has raised a further $250 million from investors, including Nvidia, as it bets that some of the world's future computing infrastructure could operate beyond Earth.
Starcloud, a Washington-based company founded in 2024, announced a $250 million extension to its Series A funding round, bringing its total funding to $450 million and giving the company a valuation of $2.3 billion.
The company's long-term vision is to create an orbital computing network capable of handling AI workloads and, eventually, the more demanding computing tasks that are currently handled by large data centers on Earth.
New investors include not only the AI chipmaker Nvidia but also Cisco Investments, the venture capital arm of the networking technology company Cisco. Existing backers include Benchmark and EQT.
Supporters of space-based data centers argue that moving computing power into orbit could help address some concerns around AI's energy demands. However, experts caution that orbital computing is far from a proven solution, highlighting major obstacles such as cooling.
Newsweek has contacted Starcloud co-founder and CEO Philip Johnston for comment by email.
Starcloud develops satellites designed to perform AI computing tasks in space rather than relying solely on earth-based data centers.
According to a company news release, Starcloud's goal is to build a fleet of 88,000 satellites that provide 20 gigawatts of orbital computing capacity. While the company hasn't translated that figure into consumer-facing terms, 20 gigawatts is broadly comparable to the output of multiple large power stations. The company is also building a 100,000-square-foot manufacturing facility in Woodinville, Washington, to expand production of the satellites it hopes will eventually form a large orbital computing network.
Starcloud says it has already achieved several milestones, including flying an Nvidia H100 chip—a specialized computer chip widely used to train and run AI models such as ChatGPT—into orbit and running AI workloads in space.
"Last November we put the first NVIDIA H100 in orbit," Johnston said in the company's funding announcement. "Today, this fresh capital empowers us to build the infrastructure to launch many more of NVIDIA's most advanced GPUs into space."
Nvidia's participation is one reason the funding round has attracted attention.
Starcloud says it was the first company to place an Nvidia H100 graphics processing unit into orbit in November 2025. The companies are now collaborating on Nvidia's Space-1 Vera Rubin Module, a computing platform designed to withstand radiation exposure and other challenges associated with operating in space.
According to Nvidia and Starcloud, the Space-1 Vera Rubin Module is specifically "built for extreme conditions in space." It's expected to provide significantly more computing power than the H100 chip that Starcloud previously flew into space.
The pitch for orbital data centers comes as concerns grow over the impact of AI infrastructure on electricity grids, water supplies and local communities.
According to the U.S. Department of Energy, data centers consumed about 4.4 percent of U.S. electricity in 2023. That figure is projected to rise to between 6.7 percent and 12 percent by 2028, with annual electricity use increasing from 176 terawatt-hours to between 325 and 580 terawatt-hours, meaning data center power demand could roughly double or, in some scenarios, more than triple.
The International Energy Agency has separately projected that global data center electricity demand could grow from about 460 terawatt-hours in 2024 to more than 1,000 terawatt-hours by 2030 as AI adoption accelerates.
The U.S. Government Accountability Office (GAO) said in an April report that space-based data centers could reduce the amount of land, electricity and water required by data centers on Earth. The report also noted that some low-Earth orbits could provide satellites with near-continuous solar energy.
The idea could also appeal to communities that have raised concerns about the impact of data center construction. Loudoun County, Virginia, one of the world's largest data center hubs, identifies noise, power demand, water use and emissions among the major issues communities consider when evaluating new facilities.
In theory, shifting some computing infrastructure into orbit could reduce pressure on local power grids, decrease water consumption used for cooling and avoid some of the noise concerns associated with large facilities and backup generators.
While space is cold, heat generated by computer systems does not dissipate easily in a vacuum. According to the GAO, large-scale cooling systems capable of supporting orbital AI facilities remain unproven.
The agency's report also warned that a major increase in satellite numbers could create congestion in Earth's orbit and increase the risk of collisions—"including with crewed missions"—raising potential safety concerns for astronauts.
Launch costs are another significant hurdle.
Starcloud's long-term plan depends heavily on SpaceX CEO Elon Musk's next-generation Starship rocket, which the company believes could dramatically reduce the cost of placing heavy payloads into orbit.
Johnston, discussing one of the company's biggest priorities, told TechCrunch earlier this month, "One of the biggest costs is now on securing your launch capacity."
"We're going to need to book an enormous amount of launch," the CEO said.
Starcloud is not alone in pursuing orbital computing.
According to public filings cited by industry publications, SpaceX has explored plans for a project known as "Starmind," which would support computing infrastructure in orbit.
A number of other companies are also working on related…
Read the original at Newsweek →
Open in TruthVane →