The surging demand for AI computing power is driving data center innovators toward space. A major hurdle remains: insufficient rockets to launch data centers into Earth orbit, coupled with high costs.

Many in the field are banking on SpaceX's Starship - slated for its twelfth test flight possibly this weekend - to address this. Yet, even after becoming operational, commercial availability could be years away due to SpaceX's own satellite priorities. Blue Origin's New Glenn faces similar issues, having failed to deploy a satellite on its third launch in April.

This scenario pushes space data center projects toward the mid-2030s, such as Google's Suncatcher, or initial edge processing for orbital sensors, like Starcloud.

A different approach: "We're launching our own rocket program," said Baiju Bhatt, CEO and founder of Cowboy Space Corporation, in an interview. He anticipates the inaugural launch by late 2028.

The firm today revealed a $275 million Series B funding round, closing at a $2 billion post-money valuation. Index Ventures led, with participation from Breakthrough Energy Ventures, Construct Capital, IVP, and SAIC.

Bhatt, who co-founded Robinhood, started this venture in 2024 as Aetherflux, aiming to harvest solar energy in space and transmit it Earthward. The concept of orbital data centers prompted a shift to utilizing that power on-site. Practical challenges then spurred rocket development and a rebrand.

Bhatt consulted various launch providers but found no viable path for scaling orbital data centers solely via satellites - capacity was lacking, and costs couldn't match ground-based options.

"Plenty of new rockets are emerging, but looking three to four years ahead, supply stays tight. I expect first-party providers to focus on their own payloads," Bhatt noted.

Vertical integration makes sense but is ambitious. In the West, only a few private firms like SpaceX, Rocket Lab, and Arianespace reliably launch commercial rockets. Blue Origin and United Launch Alliance have battled development delays for years. Startups such as Stoke Space, Firefly Aerospace, and Relativity Space have toiled without yet achieving operational flights.

Cowboy Space now directly challenges SpaceX and Blue Origin, the sector's leaders.

"The market opportunity is vast enough for multiple winners," Bhatt asserted. "AI demand intensifies while Earth-based capacity dwindles."

A key edge: laser-focus on data centers with bespoke design. Traditional rockets feature a booster to reach space's edge and an upper stage for orbital insertion. Cowboy Space integrates data centers into the upper stage - echoing Explorer 1, America's first satellite, which served as a rocket's final stage with radios and instruments.

Tailoring the rocket for data-center payloads streamlines engineering. Each satellite will weigh 20,000 to 25,000 kilograms, produce 1 MW to power nearly 800 GPUs. The rocket's power exceeds Falcon 9's but trails Starship. Bhatt envisions a reusable booster long-term.

The team includes space veterans like ex-Blue Origin propulsion engineer Warren Lamont and former SpaceX launch director Tyler Grinne. They plan proprietary engines - the priciest, trickiest component. Facilities for testing, production, and launches are in progress.

The refreshed strategy pairs with a new name underscoring the goal to "power humanity from the high frontier," though Bhatt concedes it justifies his cowboy hat and mustache.