Cool idea and quite amazingly short concept to implementation timeline, really. But one does wonder what the perceived demand is for "the growing demand for real-time computing in space". Are downlinks really that scarce? Is it that China has a relatively small global footprint so downlink from an arbitrary point above Earth is harder, or is there some killer application that simply works so much better in orbit than in a far faster larger, cheaper, more flexible and upgradeable ground-based datacentre?

Or is it a military thing where thousands of ASAT hits is either infeasible or would also take out all other nearby friendly orbital systems and thus be less vulnerable than a datacentre and/or power grid being hit by cruise missiles?