Three configurations
Order the cube that matches the job.
All three are the same 5 cm CubeSat-standard module on the same J1 / J2 fabric, and all three join a cluster the same way. Boards can be added later.
CubeSat standard
CubeSat standard by construction, not by adaptation. The cube that ran on your roof is the cube that flies.
Open source
No proprietary bus to design around: J1 and J2 are published.
Powered by Linux
The same module carries a project from bench prototype to field unit to flight article without a redesign.
One module, many machines
What it becomes depends on which boards you stack.
Every card attaches through the same two 64-pin headers — J1 carrying power, reset and four UARTs, J2 carrying the CAN, SPI and I²C data fabric. Add and remove boards to match the mission; the cube, the standard and the network membership never change.
Board index
Five cards, and what is on each.
The full J1 / J2 pin assignment is in the datasheets. Everything below is the short form.
The same cube, read three ways
What it is depends on who is holding it.
NODE IO is a component, an infrastructure and a membership at the same time. Which of the three you came for changes nothing about the hardware.
If you build machines
The core of the thing you are building.
Open source, CubeSat standard, and built on silicon from the top of each category rather than whatever was cheapest. No proprietary bus to design around: J1 and J2 are published, the OS is Linux, and the same module carries a project from bench prototype to field unit to flight article without a redesign.
For business & government
Distributed infrastructure, not a fleet of gadgets.
Multi-domain swarms — land, sea, air and orbit — as clusters of sensors that compute and decide locally and coordinate over the network. Distributed computing and distributed AI, with no single point of failure to lose and no single point of control to seize. Sovereign by architecture, because the nodes belong to whoever bought them.
If you are one person
Being part of something considerably bigger.
Every cube you connect joins your cluster and adds capacity the network did not have before — compute, storage, and links across ground and orbit. There is no central build-out to wait for and no permission to request. The infrastructure reaches exactly as far as the people building it.







