Starcloud, a pioneering space compute company, has achieved a valuation of $1.1 billion following its latest funding round, positioning it among the fastest startups to reach unicorn status after graduating from Y Combinator.
This Series A funding round, which concluded a mere 17 months after the company's demo day presentation, was spearheaded by prominent investors Benchmark and EQT Ventures. This significant investment underscores the escalating interest in migrating data centers to orbit, particularly as terrestrial development faces increasing resource and political hurdles. However, it's acknowledged that this business model relies on unproven technology and necessitates substantial capital expenditure.
To date, Starcloud has successfully raised a total of $200 million. The company marked a significant milestone by launching its inaugural satellite, equipped with an Nvidia H100 GPU, in November 2025. This year, Starcloud plans to deploy an even more powerful iteration, Starcloud 2, featuring multiple GPUs, including an Nvidia Blackwell chip, an AWS server blade, and a dedicated bitcoin mining computer.
Further demonstrating its ambitious vision, Starcloud is actively developing a specialized data center spacecraft, named Starcloud 3, designed for deployment from SpaceX's reusable heavy-lift rocket, Starship. This 200-kilowatt, three-ton spacecraft is engineered to seamlessly integrate with the "pez dispenser" system that SpaceX employs for deploying its Starlink satellites from Starship.
Philip Johnston, CEO and founder of Starcloud, expressed his expectation that Starcloud 3 will represent the first orbital data center capable of achieving cost-competitiveness with its terrestrial counterparts. He projects power costs to be around $.05 per kilowatt-hour, contingent on commercial launch costs stabilizing at approximately $500 per kilogram.
A primary challenge, however, remains Starship's operational readiness. Johnston anticipates commercial access to Starship opening up between 2028 and 2029. This timeline highlights a broader reality for all major space data center initiatives: powerful orbital computing will remain prohibitively expensive until a new generation of rockets achieves a high operational cadence, a development that may not materialize until the 2030s.
"If it ends up being delayed, we’ll just carry on launching the smaller versions on Falcon 9," Johnston stated, acknowledging contingency plans. He further emphasized, "We’re not going to be competitive on energy costs until Starship is flying frequently."
Johnston elaborated on Starcloud's dual business model: "There’s kind of two business models." The first involves providing processing power to other spacecraft in orbit; for instance, the company's initial satellite analyzes data gathered by Capella Space’s radar spacecraft. The second, more ambitious model, envisions that as launch costs decrease, more powerful, distributed orbital data centers could eventually attract workloads from their terrestrial equivalents.
This emerging industry is undeniably in its infancy. A telling example is Nvidia CEO Jensen Huang's recent unveiling of the company’s Vera Rubin Space-1 chip modules at its annual GPU Technology Conference. Notably, he did not mention that these modules had yet to be produced or shared with Nvidia's development partners.
Indeed, the current number of advanced GPUs in orbit is estimated to be in the mere dozens, a stark contrast to the nearly 4 million units Nvidia is projected to have sold to terrestrial hyperscalers in 2025 alone.
Furthermore, consider the energy consumption disparity: SpaceX’s Starlink, the largest satellite network with 10,000 spacecraft, generates approximately 200 megawatts of energy. In comparison, terrestrial data centers with over 25 gigawatts of power are presently under construction in the U.S., according to Cushman and Wakefield.
Johnston asserts that Starcloud holds a significant competitive advantage, having deployed the first terrestrial GPU in orbit. This achievement was leveraged to train an AI model in space—a claimed first—and to run a version of Gemini. Beyond performance metrics, Johnston highlights the invaluable data Starcloud has accumulated regarding the operational requirements of powerful chips in the space environment.
"An H100 is probably not the best chip for space, to be honest, but the reason we did it is we wanted to prove that we could run state of the art terrestrial chips in space," Johnston told TechCrunch. This hard-won expertise, gained even after an Nvidia A6000 GPU failed during a previous launch, will critically inform future design iterations.
A multitude of technical challenges remains to be addressed, including the efficient generation of power and the effective cooling of high-performance chips. Starcloud-2 is slated to feature the largest deployable radiator ever flown on a private satellite, and Johnston anticipates at least two more versions of this spacecraft will eventually reach orbit.
Another significant hurdle is synchronization. The most demanding data center workloads, particularly for AI training, necessitate hundreds or even thousands of GPUs operating in perfect concert. Achieving this in space would either require exceptionally large spacecraft or highly reliable and powerful laser links between satellites flying in formation. Most companies active in this sector anticipate that such complex workloads will only become feasible long after simpler inference tasks are routinely performed in orbit.
Beyond Starcloud, other notable players developing space data center solutions include Aetherflux, Google’s Project Suncatcher, and Aethero, which successfully launched Nvidia’s first space-based Jetson GPU in 2025.
The unspoken, yet significant, competitor in this domain is SpaceX itself, which has sought U.S. government approval to construct and operate a network of a million satellites specifically for distributed compute capabilities in space.
While directly competing with SpaceX presents a formidable challenge for any entrepreneur, Johnston believes there is ample room for coexistence.
"They are building for a slightly different use case than us," he conveyed to TechCrunch. "They’re mainly planning on serving Grok and Tesla workloads. It may be at some point that they offer a third party cloud service, but what I think they are unlikely to do is what we’re doing [as] an energy and infrastructure player."
The Editorial Staff at AIChief is a team of professional content writers with extensive experience in AI and marketing. Founded in 2025, AIChief has quickly grown into the largest free AI resource hub in the industry.