Meta’s Space-Based Solar Power Deal Sounds Like Science Fiction, Wall Street Is Treating It Like a Revenue Stream

Mark Zuckerberg’s agreement to use uncollected sunlight from unbuilt satellites to power data centers that are being built faster than the grid can keep up with has an almost theatrical quality. And yet, here we are. On April 27, Meta announced a partnership with Overview Energy, a four-year-old Ashburn, Virginia startup that aims to place a thousand spacecraft in geosynchronous orbit, collect solar energy in areas without night, and transmit it back to Earth as a broad, low-intensity infrared light. It sounds like a proposal made in a writers’ room. Most investors are not amused.

The extent of Meta’s desire for power is worth considering. The company’s data centers used over 18,000 gigawatt-hours in 2024 alone—roughly enough to power 1.7 million American homes. Because each new AI feature, trained Llama model, and improved recommendation engine draws power the same way a furnace draws oxygen, the number is rapidly increasing. Speaking with industry insiders, it seems like traditional grid build-out is unable to keep up. Hyperscalers have begun to sign contracts for hydrogen experiments, geothermal pilots, and nuclear power. Meta has now literally ascended to the geosynchronous belt, which is located 22,000 miles above the equator.

The cleverness of Overview’s pitch sets it apart from earlier space-solar ideas that used to appear in DARPA white papers and then vanish into thin air. The company is not attempting to launch powerful lasers or microwaves. According to its CEO, Marc Berte, you could look straight into the beam without getting hurt. The infrared light strikes ground-based solar farms, which are the same panels we currently use, and is transformed into electricity by infrastructure that is already operating in the Texas plains or the Arizona desert. No new lines of transmission. No new disputes over permits with cattle ranchers. Just additional hours of production from solar farms that are now in the dark from dusk to dawn.

Wall Street is focused on that aspect. A solar farm’s unit economics are significantly altered if it can operate for twenty-four hours as opposed to eight. Even though no money has reportedly changed hands yet and the first orbital test isn’t scheduled until 2028, analysts at companies tracking Meta have started modeling the deal as if it were a real revenue and cost lever for 2030 and beyond. Markets pricing in a technology that has only flown on an aircraft, never from space, is a little out of the ordinary.

It is reasonable to be skeptical. Hardware in space malfunctions. Launch expenses skyrocket. Orbits that are geosynchronous become congested. Furthermore, there is currently no legal framework in place for transmitting energy across international borders. SpaceX, Blue Origin, Google, and other Silicon Valley companies are investigating various iterations of the same concept, indicating both competitive risk and validation. It’s still unclear if Meta is merely purchasing a cheap option on a future that might or might not materialize, or if Overview will be the business that truly delivers.

It’s difficult to ignore how quickly the unthinkable turns into a press release line item as you watch this play out. Ten years ago, energy conferences continued to make fun of terrestrial solar. In many areas, it is currently the least expensive power on the planet. Perhaps the satellites are operational. Perhaps they don’t. However, the fact that a publicly traded company can announce a constellation of a thousand spacecraft while its stock hardly flinches speaks volumes about the times we live in.