On October 26, 2026, Dunamu & Partners — the investment arm of South Korean crypto-exchange operator Dunamu — disclosed that it had taken part in Starcloud's $250 million funding round, becoming the round's only Korean investor. The disclosure put orbital computing back in the headlines. But careful readers should note what the news is, and what it is not: the $250 million round itself was raised in August 2026, not in October. The October news is a new investor's disclosure about that existing round. Understanding that distinction is the key to understanding orbital compute itself — a field where the vision is enormous, the engineering is real, and the timeline is longer than the headlines suggest.
What orbital compute is
A data centre is, at heart, a building full of computers that needs three things: power, cooling, and a network connection. On Earth, all three are increasingly constrained. Grid connections for large data centres can take years to secure, sites near population centres are scarce, and cooling consumes vast amounts of water and electricity. Orbital compute proposes moving the computers to space, where sunlight is available almost continuously, waste heat can be radiated into the cold of deep space without cooling water, and solar energy per square metre is roughly eight times more efficient than on Earth's surface, free of weather, clouds, and night.
Starcloud, founded in 2024 by engineers formerly with SpaceX's Starlink and Airbus, is the most prominent company pursuing this idea. Its plan is a constellation of data-centre satellites in low Earth orbit, communicating with each other by laser, that could eventually deliver large-scale computing capacity from space. In February 2026, the company asked the US Federal Communications Commission (FCC) for permission to fly up to 88,000 satellites as a distributed data centre in space, in orbits 600 to 850 kilometres high. The FCC accepted the application for filing in March 2026 — an important procedural step, but not an approval.
The funding, correctly dated
Starcloud's financing history is short and steep. In March 2026, the company announced a $170 million round led by Benchmark and EQT Ventures. In August 2026, it closed a $250 million Series A extension at a $2.3 billion post-money valuation, led by Manhattan West, with Nvidia, Cisco Investments, Benchmark, and EQT among the participants. That August round took Starcloud's total raised since its 2024 founding to $450 million. Dunamu & Partners' October 26 disclosure confirms its participation in that same August round; it does not announce a new round, a new valuation, or new total.
Investors evaluating the sector should anchor on the verified sequence: $170 million in March, $250 million in August, $450 million cumulative, $2.3 billion valuation.
What has actually flown
Starcloud is not only a slide deck. Its first satellite, Starcloud-1 — a 60-kilogram spacecraft — launched in November 2025 carrying an Nvidia H100 graphics processing unit, and ran AI inference in orbit using Gemma, Google's AI model. It has since been used to analyse data from satellite radar company Capella Space to detect ships in distress and wildfires.
The next steps are more ambitious and further away. Starcloud-2 satellites, each around 8 kilowatts, were reported in August 2026 as planned for shared launches in 2027, doing work for customers including US government agencies. Starcloud-3, a much larger spacecraft designed to be released from SpaceX's Starship, is being developed on production lines at a new 100,000-square-foot factory in Woodinville, Washington, with estimates ranging from 2 tons and 100 kilowatts to 3 tons and 200 kilowatts. Separately, on September 30, 2026, Firefly Aerospace announced it will carry a Starcloud payload around the Moon on its Elytra orbiter, on a mission planned for no earlier than 2028, where it will process data from an onboard vision system after the lander separates.
The physics problem, and the business problem
Orbital compute's pitch rests on real physics. In a vacuum, there is no air to carry heat away, so every watt a chip produces must be radiated into space — a process governed by the Stefan–Boltzmann law, under which a hotter radiator sheds heat dramatically more efficiently. Starcloud circulates coolant through its hardware and transfers heat to deployable radiator panels.
But the business problems are just as real. Launch costs, radiation-hardening, component reliability, and the sheer scale of an 88,000-satellite constellation all remain to be proven. Regulatory approval for tens of thousands of satellites, space-debris management, and competition from terrestrial data-centre builders who are also innovating on power and cooling all stand between today's demonstrations and a working orbital cloud.
Why it matters for founders and investors
For founders, Starcloud is a lesson in narrative discipline: a genuinely novel technology, demonstrated in orbit, funded by serious investors — and still best understood through verified milestones rather than headline numbers. For investors, orbital compute is a classic frontier bet: enormous potential upside, a long path to revenue, and a valuation ($2.3 billion) that already prices in substantial success. The right diligence questions are concrete: How many satellites, at what power, by when? What revenue do early customers pay? And what does the FCC actually approve? The answers to those questions, not the size of the last round, will decide whether computing's next home is in the sky.
Angeled.com connects founders with investors.
Seoul Economic Daily (SIGNAL), "Dunamu Unit Joins $250 Million Round for U.S. Space Data Center Startup," October 26, 2026
Source: datacenterdynamics.com
Data Center Dynamics, "Starcloud closes $250m Series A extension at $2.3bn valuation," 2026
Source: geekwire.com
GeekWire, "Firefly will give Starcloud's AI data center a ride to lunar orbit," October 1, 2026
Source: spacedew.com
SpaceDew, "Meet Starcloud, the Company Trying to Put Data Centers in Orbit," October 2026
Source: wisevoter.com
Wisevoter, "Space Data Startup Starcloud Secures $250 Million Funding," October 26, 2026
Read Next
General Translation Raises $27 Million Series A Led by a16z to Build Agent-First Localization Infrastructure
LangChain Raises $300 Million Led by CapitalG at a $3 Billion Valuation to Build the Agent Stack
Mount Thor Raises $45 Million to Run Cloud Fleets of Macs for AI Agents
OpenAI Reportedly Raises $15 Billion at a $1.25 Trillion Valuation
Q-CTRL Raises $290 Million Series C Led by Salesforce Ventures, Reaching Unicorn Status
SE3 Raises $30 Million Series A Led by Accel



