Voyager’s new product family brings composable, open compute to orbit so mission teams can stop designing a new satellite computer for every program.
Nearly every mission flying today commissions its onboard computer from a blank sheet. Months, sometimes years, of non-recurring engineering paid for by a single program. One box, sized to one power budget, locked to one vendor’s design and married to one mission for life. That is the blank-sheet tax, and every customer pays it.
Voyager is introducing Space Edge™, a family of composable on-orbit compute products built to end it. Customers select cards from a proven catalog, configure the system for their mission and fly. Space Edge allows users to compose rather than commission.
The other half of the orbital compute story
Orbital data centers have captured the industry’s imagination over the past year. But spaceborne computing can go one of two ways: wholesale terrestrial data centers in space or, the lesser explored option, what Voyager is calling the Space Edge.
“Orbital data centers are solving a very specific problem: taking the capability of a terrestrial data center moving it to space,” said Paul Tilghman, chief technology officer, Voyager. “But Space Edge is a secondary approach to spaceborne computing that is about integrating processing with sensors and doing the computation right where the data lives.”
Three forces are converging to make something like Space Edge possible. Launch costs continue to fall, spaceborne sensors are proliferating faster than downlink capacity can carry their data to the ground and optical inter-satellite links are turning individual spacecraft into connected fabrics. Missions such as missile defense make the point plainly: infrared, radio frequency and electro-optical data from multiple spacecraft must be fused in seconds, with no time to wait ninety minutes for a ground station pass.
Build from a catalog, not a blank sheet
Space Edge launches with three foundational processing cards: a general-purpose processor card, a field-programmable gate array card and a graphics card. Computer vision for missile defense calls for one mix, whereas cascading ARM processors that run large language models in deep space call for another. Space Edge can be uniquely composed to work for both situations.
“The old model was to spend one year on the design for one processing problem that wouldn’t make it to space for another four to five years. But the industry is moving too fast for that now. Space Edge lets us custom build a solution out of pre-existing blocks,” said Tilghman.
The composability means the cost is shared across every program that flies it rather than borne alone. Each card added and each mission flown makes the next one faster, more affordable and more capable. An open-standards ecosystem also means customers are never wedded to a single supplier, and allied partners can build sovereign capability on a common platform.
“You might build a solution that is seventy-five percent Voyager and twenty-five percent from another vendor,” said Tilghman. “That is the beauty of evolving away from building bespoke electronics.”
From prototype to platform
A prototype Space Edge platform is already operating aboard the International Space Station. It has been used to work through the software layer of on-orbit computing: how payloads developed on the ground are orchestrated, placed and given access to sensors in flight. The knowledge gained from the prototype is what determined which processing cards the product family needed.
“It is active, flying and we talk to it every day. We consider it a community asset,” said Tilghman.
The longer arc points beyond defense applications. As missions travel farther from Earth, the round trip to mission control stretches from three seconds at the Moon to roughly half an hour at Mars. The radiation-tolerant processors of previous generations cannot carry the load.
“A processor with the same speed as the 386 I had as a child is not going to give you ‘Houston in a box’ on the way to Mars. To enable humans to traverse the cosmos, we need to bring the computing substrate needed for modern AI to space,” said Tilghman.
Space Edge is built for primes carrying avionics non-recurring engineering into every bid, satellite and bus manufacturers redesigning compute for every customer, constellation operators who cannot afford a custom computer per spacecraft and the government programs that need one compute standard across a proliferated architecture. The blank sheet has had a long run, but with Space Edge, your next mission can start at the finish line.