160 square meters, which if ran in optimum conditions would reject between 100 to 150 kilowatts.
Micrometeorites is not a significant concern. There aren't that many of them and 100% reliability is not required. If one is hit then you end of life it or de-rate it (i.e. shut down some of the processors or underclock them) appropriate to the lost solar energy or lost radiating power.
Server clusters already replace racks every 5 years or so and dispose of them. (It used to be ever 3 years.) How is this any different?
160 square meters, which if ran in optimum conditions would reject between 100 to 150 kilowatts.
Micrometeorites is not a significant concern. There aren't that many of them and 100% reliability is not required. If one is hit then you end of life it or de-rate it (i.e. shut down some of the processors or underclock them) appropriate to the lost solar energy or lost radiating power.
Server clusters already replace racks every 5 years or so and dispose of them. (It used to be ever 3 years.) How is this any different?