Orbital Compute's Near-Term Business Takes Shape with Kepler and Sophia Space Partnership

The near-term business of orbital compute is beginning to take shape as the number of GPUs in space starts to increase.

Canada's Kepler Communications launched the largest compute cluster currently in orbit in January. This cluster features approximately 40 Nvidia Orin edge processors across 10 operational satellites, all interconnected via laser communication links.

Kepler Communications now serves 18 customers and announced a new partnership with Sophia Space, a startup focused on testing its unique orbital computer software on Kepler's constellation.

Experts anticipate that large-scale data centers, similar to those envisioned by SpaceX or Blue Origin, will not be a reality until the 2030s. The initial phase of orbital compute will involve processing data collected in orbit to enhance the capabilities of space-based sensors used by both private companies and government agencies.

Kepler Communications views itself as a provider of infrastructure for space applications, rather than a data center company. CEO Mina Mitry explained to TechCrunch that the company aims to offer a network services layer for other satellites in space, as well as for drones and aircraft.

Sophia Space, conversely, is developing passively cooled space computers. This technology aims to address a major challenge for large-scale orbital data centers: preventing powerful processors from overheating without the need for heavy and costly active-cooling systems.

Under the new agreement, Sophia will upload its proprietary operating system to one of Kepler's satellites. The startup will then attempt to launch and configure it across six GPUs on two spacecraft. While this is considered standard practice for terrestrial data centers, it represents a first for orbital operations. Successfully validating the software's performance in orbit is a crucial de-risking step for Sophia before its planned satellite launch in late 2027.

For Kepler, this partnership validates the utility of its network. Currently, Kepler handles and processes data uploaded from the ground or collected by hosted payloads on its own spacecraft. As the sector matures, the company expects to integrate with third-party satellites to offer networking and processing services.

Mitry noted that satellite companies are now designing future assets with this model in mind, recognizing the advantages of offloading processing for power-intensive sensors, such as synthetic aperture radar. The U.S. military is a significant customer for such applications, particularly as it develops a new missile defense system that relies on satellites for threat detection and tracking. Kepler has already conducted a successful demonstration of a space-to-air laser link for the U.S. government.

This type of edge processing—handling data at its collection point for faster responsiveness—is where orbital data centers are expected to demonstrate their initial value. This focus differentiates Sophia and Kepler from established space companies like SpaceX and Blue Origin, as well as startups like Starcloud and Aetherflux, which are raising substantial capital for large-scale data centers utilizing traditional data center processors.

Mitry stated, “Because we have the belief it’s more inference than training, we want more distributed GPUs that do inference, rather than one superpower GPU that has the training workload capacity. If this thing consumes kilowatts of power and you’re only running at 10% of the time, then that’s not super helpful. In our case, our GPUs are running 100% of the time.”

Furthermore, the increasing viability of these orbital technologies could have broader implications. Sophia CEO Rob DeMillo pointed out that a recent ban on data center construction in a Wisconsin city, and similar proposals in Congress, could make space-based alternatives more attractive.

“There’s no more data centers in this [city],” DeMillo remarked. “It’s gonna get weird from here.”