Connectivity from satellites is now a reality and a surprising number of systems already depend on this, but what are the benefits and limitations and where is edge in space heading — maybe to the moon
Space is the quintessential edge location: remote, poorly and intermittently connected, extremely resource challenged and uniquely environmentally challenged. Using satellites for connectivity has already transformed a range of edge applications on earth and they’ll be necessary for 6G; the next step is moving compute into space. After several years of promises, that’s just starting to become a reality while NASA’s Artemis missions aim to bring the edge to the Moon, with a distant eye on Mars. The challenge of reliable, high performance computing in space is far from solved though — and geopolitical questions add another dimension of complexity. In the meantime, the miniaturization and reliability work done for space may have payoffs for the most challenging environments here on Earth.
For the last few years I've been contributing to the State of the Edge report from LF Edge and this year I got to go back to the predictions I'd made in a previous year about edge compute in space - just as it became somewhere between timely and political.
I spent a couple of months digging into why you might want to do compute in space, what the physics really says about cooling, radiation and the other restrictions, how much edge compute is already happening in satellites and whether that's useful.
I also look at whether lunar exploration and industrialising the moon will take space edge to the next level. The tensions over astronomy and the ever increasing number of satellites aren't going away, but what do we get out of compute in space? Download the report and dig in
edge
space
radiation
cooling
satellites
Arm
Raspberry Pi
local models