WHY SPACE COMPUTE

Some workloads are easier
to run above the atmosphere.

The thesis is not that computers belong in space. The thesis is that some of the hardest constraints on Earth-based computing are energy, heat, and land — and that orbit changes the shape of those constraints for specific future workloads.

Near-continuous sunlight, no grid queue

Energy

Orbital photovoltaic systems operate in sunlight for most of the duty cycle, without the diurnal swings, weather variability, or grid-interconnect delays that constrain terrestrial clean power. Power becomes a function of array area and orbital geometry — not land use or transmission capacity.

Radiative rejection to deep space

Cooling

There is no air in orbit. Heat leaves through radiation — engineered via large emissive surfaces designed to dump thermal load at the right wavelengths. This removes the need for evaporative water loops and local heat-island effects. It also demands serious thermal design; nothing about this is automatic.

No zoning fights, no community burden

Land & permitting

Orbital infrastructure does not compete with housing, agriculture, or ecosystems for space. It does not trigger the years-long permitting and interconnect processes that gate new terrestrial builds. For certain workloads, that alone changes the economics.

Removing terrestrial pressure

Water & grid strain

A share of hyperscale data-centre footprints consume significant volumes of water for cooling. Shifting a portion of workload off-planet can reduce pressure on local water systems and grid capacity, especially in regions where both are already constrained.

Training, batch, science

Workload fit

Orbital compute is not for real-time user traffic. It is for workloads that tolerate orbital propagation delays: large-model training, batch inference, scientific simulation, long-baseline data processing, archival pipelines, and compute tied to space-native data sources.

Not a replacement. A new tier.

Niche-first

Inorbii is explicit about this: we are not replacing Earth data centres. We are building an additional class of infrastructure for the workloads where orbital deployment may become economically and technically meaningful over the next decade.

THERMAL ENGINEERING

Radiative cooling is the
central engineering problem.

In orbit, the only outbound path for waste heat is electromagnetic radiation. That makes the radiator field as critical as the solar array. Its area, orientation, emissivity, and materials define how much compute the platform can sustain. Thermal design is not a constraint on the platform. It is the platform.

COMPUTEWASTE HEAT →RADIATIVE REJECTION TO DEEP SPACESOLAR INPUT →

Illustrative diagram. Actual rejection geometry depends on orbital regime, sun angle, payload density, and system architecture.

READ NEXT

How it fits together — power, compute, thermal, comms.