The Thermal Wall

Every orbital data center architecture hits the same fundamental constraint: you can't get rid of heat fast enough.

Why Space Computing?

The demand for AI compute is growing faster than terrestrial infrastructure can scale. Space offers unlimited real estate, abundant solar energy, and natural vacuum cooling, in theory. The challenge is turning that theoretical advantage into engineering reality.

The Heat Problem

On Earth, you blow air over heatsinks or pump water through cooling towers. In space, there's no air. The only way to shed heat is thermal radiation, and radiation scales with surface area. More compute means more heat means more radiator surface means more mass.

Why Previous LDR Attempts Failed

Liquid Droplet Radiators have been studied since the 1980s. NASA and others demonstrated the physics works. The droplets radiate heat beautifully. The problem was always getting them back:

• Electromagnetic collection:requires conductive fluids, magnets, and significant power
• Electrostatic steering:charge leakage in plasma environment, complexity
• Mechanical scoops:moving parts, wear, alignment precision

Each approach added mass, complexity, and failure modes that eroded the theoretical advantage. After 30 years, no LDR system has flown operationally.

Centradiant's Approach

We use centrifugal force from a slowly spinning disk to passively collect droplets. No magnets, no charging, no precision alignment. The spinning creates the collection force. The dual-layer mesh at the rim creates the collection surface. Physics does the rest.

Built on Proven Heritage

The Centradiant system isn't built from scratch. Five of nine subsystems use flight-proven technology with direct ISS heritage:

• Ceramic bearing:ISS Solar Alpha Rotary Joint (operating since 2007)
• Rotary fluid coupling:ISS Thermal Rotary Radiator Joint
• Solar arrays:iROSA with XTJ Prime cells (ISS heritage since 2021)
• CB-DC705 silicone oil:DC-705 base (STS-77 heritage) + carbon black for near-unity emissivity

The remaining subsystems each have a clear path to qualification through a $237K ground prototype program (6 months to TRL 5). Even under conservative assumptions (>99.99% capture efficiency), the system is 6–9× lighter per kilowatt than conventional panel radiators.