How an e-scooter inventor raised $5 million to build space data centers

Here’s one metric to track SpaceX’s IPO later this week: The company has changed the commercial industry’s perception of long-term, capital-intensive space to the extent that a talented founder with no space experience could fund a space data center company.

Orbital, a new company that emerged in May from the a16z startup accelerator program Speedrun with a 5 million seed round, is the latest company to take on the challenge of space — as soon as Starship flies regularly. Other investors include Basis Set, Human Element, Wayfinder, Antler, Anti Fund, Ascent, Rubik, Zero Knowledge Ventures, LYVC, Feld Ventures, New Legacy, FNDR, UpHonest and Asterisk.

Founder and CEO Euwyn Poon previously founded e-scooter company Spin in 2017 and sold it to Ford a year later, joining the auto giant. When he was ready to start a new company, Speedrun’s a16z was eager to get on board, according to partner Andrew Chen, who told TechCrunch that Poon had worked on several ideas before landing on space data centers.

You know the tone. There is an unsatisfied need for AI computing, and its adoption is slow in the world. Why not head to space for unlimited sunlight and limited environmental reviews? The biggest problem is the brutal economics of launching things into orbit, which currently leaves the business case unable to close.

Orbital, like many of its competitors, is betting on SpaceX getting its Starship rocket and offering it to commercial customers. “We will reach full scale when Starship comes online,” Poon explained. The price of the Falcon 9, the current state of the art, “makes this economically unfeasible.”

Currently, Poon and the company — which includes about a dozen people in Los Angeles, with experience at Amazon LEO, SpaceX, and Northrop Grumman — are working on a demo flight that will see the company fly a partner’s Nvidia Blackwell satellite chip to test orbital radiation protection and thermal management technology. In 2028, the company hopes to launch its first data processing spacecraft with Nvidia’s Space-1 Vera Rubin-class GPUs.

Meanwhile, the company wants to start doing smart work, which will allow it to make money for each satellite launched. That’s the same way a rival data center startup, Starcloud, already has a GPU in orbit and plans to launch several more to generate revenue until Starship allows them to use their full constellation.

Orbital’s goal is to operate 10,000 satellites providing a distributed gigawatt of computing power, each satellite providing 100 kw of power. In comparison, Elon Musk said that SpaceX expects its AI satellites to generate up to 150 kw, while Starcloud expects to install a large spacecraft rated at 200 kw to run the chips.

Some companies are too impatient to wait for Starship. Cowboy Space Company, another space data center startup supported by a16z, recently decided to start building its own rockets. Jeff Bezos’ space company, Blue Origin, has also announced plans to launch data centers into space using its new Glenn launch vehicle.

Poon is confident that the breadth of demand for AI will allow more companies to succeed. “There are many paths that companies are following in our space,” he told TechCrunch, before laying out a number of decisions that include companies pursuing different AI workloads, designs, and concepts for what a space data center should look like.

Chen said Poon’s experience in scaling up a company with 250,000 scooters distributed in 100 cities shows that he can handle the difficult task of building a space company. In the long run, a project like this could take a decade and $5 billion or more, but Chen said business firms are more comfortable with such timelines.

“This kind of thing would have sounded crazy 10 years ago when we were building mobile apps,” he said. “Starting it in 2026 allows you to tap into all the energy and excitement that is happening in the capital markets.”

Poon found his way into the space data center business via a circuitous route. After leaving Ford, he bought an Nvidia A100 on a lark, put it together in a Santa Clara data center and offered open-weight models. That first experience convinced him of the importance of bringing computing into the AI ​​era.

Now he has to put several thousand GPUs into space.