Most of the internet rides on thin glass fibers resting on the ocean floor, and hauling one to the surface for repairs is exactly as awkward as it sounds. Meanwhile hyperscalers keep spinning up data centers in every corner of the map, and the traffic between them has to travel somehow.
Endeavor Optical Networks reckons the fix is to aim lasers upward instead.
Founded in May, the startup emerged from stealth today holding $10.75 million in seed funding, backed by General Catalyst and Andreessen Horowitz. Chief executive Charlie Horowitz and chief technology officer Tyler Presser envision a fleet of laser-carrying spacecraft stitching data centers together from orbit.
The number that matters is 2.4 terabits
Consider the chasm EON has set out to bridge. Fiber laid beneath the sea carries 200 terabits per second or better. The bulk of satellite communications networks, the broadband providers among them, are nowhere near that figure.
Optical links themselves aren’t new. NASA relied on laser comms to send data home from its latest Moon mission, and private outfits such as York, Kepler and Cailabs have proven links running between Earth orbit and the surface.
Those demonstrations, however, were built around 2.5 Gbps. EON’s opening target is 2.4 terabits a second of throughput, according to Horowitz — approximately three orders of magnitude beyond anything demonstrated so far. Getting there means beating the perennial headache of optical communications: signals get warped as they pass through the atmosphere, and cloud cover stops them cold.
The company claims to have a method for handling that. What the method is, it won’t say.
Twenty satellites, not 5,048
Roughly 20 satellites are envisioned, each dedicating a link to a pair of continents, with the opening fleet supplying 24 hour coverage to early customers. Ground stations are to be sited on purpose close to nearby data centers and CDNs, backed by redundant locations and weather data to keep links from dropping.
Set that beside Blue Origin. Jeff Bezos’ space venture has unveiled TeraWave, a 5,048 satellite constellation chasing speeds as high as 6 Tbps for large-scale users. The ambition is greater, and so is the time it will take to launch and deploy. EON’s far leaner fleet ought to reach orbit sooner, though a smaller constellation doesn’t make the physics any easier.
Selling to the people who move the most data
Conversations are underway with hyperscalers and AI labs — logical enough, given nobody shifts more bits. The sales case zeroes in on routes that are pricey or poorly served: lengthy hauls such as France to Australia, or crossings thin on existing infrastructure, Africa to South America among them.
Dedicated capacity is the intended product, pitched at buyers who want complete control over how their data moves.
Nothing on that list arrives quickly. Seed money is earmarked for an optics lab, additional engineers and ground testing, all leading up to a demonstration satellite the company aims to put in orbit around the close of 2027. Horowitz anticipates that craft will post the strongest optical downlink throughput on record, no less than 800 gigs and possibly a full terabit.
Buy the bus, build the terminal
Restraint defines EON’s engineering approach. Its attention goes to building the optical communications terminal, with money spent judiciously on the pieces that must be exquisite — the gimbals steering the laser, for instance. Everything else it intends to source as capable off-the-shelf satellite busses, Apex Space’s included.
Apex is where Horowitz used to work. He was chief of staff to chief executive Ian Cinnamon before becoming the firm’s director of special projects.
“Charlie is a force of nature—he can move seamlessly from strategy to the details required to make something real,” Cinnamon said. “Charlie is the ideal founder, and I invested personally because I believe deeply in Charlie and what he’s building at EON with Tyler.”
Presser holds a PhD in astronautical engineering and has mapped out frontier missions for NASA. Rounding out the technical roster are Michael David Francois, a longtime Google executive whose work centered on global network infrastructure, and Wesley Baxter, an optics engineer whose most recent role was on Amazon’s LEO satellite network.
The investor isn’t worried about demand. She’s worried about the schedule
Jeannette zu Fürstenburg, the General Catalyst partner behind the deal, said EON brings together the fund’s two central themes: AI and resilience.
“I don’t worry about demand,” she said. “I think all of that will solve for itself. It’s really all about can you actually get this thing into space in the time that we discussed? We really think about founder-product fit, [Horowitz] is just the right caliber of guy to go after a problem like this.”
That is the risk stated plainly, and it deserves a moment’s thought. Nobody is really betting on whether the world wants terabit connections between continents. The wager is whether a company that came into existence in May can deliver working orbital hardware by late 2027.
A researcher’s caution
Caleb Henry, research director at Quilty Space, won’t wave the concept away, but he is blunt about what it has to clear.
“Data centers have high standards for quality and redundancy,” Henry said. “Satellite internet is just now progressing from a technology of last resort to dependable, high-bandwidth infrastructure. That’s not to say it will be impossible to make satellites optimized for data center connectivity, just that it will be harder and take longer than most entrepreneurs suggest.”
A reasonable point. Still, measured against the current enthusiasm for parking the data centers in orbit, firing bits through the atmosphere on a laser beam starts to look positively grounded.
One phrase keeps resurfacing in Horowitz’s remarks, and it explains more about how EON chooses its fights than anything else he offered.
“We have one rule at the company: no physics problems,” Horowitz said. “There’s a market that exists today that we can go serve. Down the road, we’ll go and take on more as it comes, but we know that this is a problem that exists today, that’s only getting worse. That’s our bet—more data is moving terrestrially than ever.”
Keep an eye on the demo satellite. Should EON clear 800 gigs on a downlink near the end of 2027, that 2.4 terabit figure stops reading like a slide in a pitch deck. Should the launch drift beyond that window, everything downstream drifts with it.