Hacker News new | past | comments | ask | show | jobs | submit login

It's worth noting that the SpaceX satellites only have a mission length of 5-7 years, and the Spectrolab solar panels are built to last for a typical 15 length mission (from the link you provided). Perhaps they will opt for (presumably) cheaper components given the shorter lifetime.

15 years is the nominal design lifetime for an "insured" geostationary telecom satellite... When SES or Intelsat launches a new 3500 to 6000 kilogram sized satellite it's insured by a third party company. There's specialists in this. I'd say it's much more likely the SpaceX program will be self-insured so they can take a risk with satellite bus technology and other tech that has never been flown before.

How this will affect their choice of PV cells I don't know.

With 4000 satellites it's likely that their approach to each satellite will be smaller and much less redundancy:

If a traditional 5000 kg geostationary telecom satellite can be compared in analogy to a big, expensive, 4U, quad socket xeon server that has multiply redundant everything.

These small satellites might be much more like a facebook open compute blade server, 1+0 and redundant nothing, but much, much cheaper to build and with a software architecture tolerant of entire nodes failing and disappearing from the network.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact