This is just what an orbital data center needs

[John Clark]

 On Tue, Jun 30, 2026 at 1:26 PM spike jones via extropy-chat <extrop...@lists.extropy.org> wrote:

> we need a subcategory: those who think the orbiting data center is a good idea, but who know it won't work.  I derived the math model for it 34 years ago and found there wasn't enough power available and even if there is enough power from a nuclear reactor, there isn't nearly enough heat rejection capacity.  It would be cool as all hell if we could, but... analogous to all hell... it gets too hot

Just last week IBM announced they had made an experimental chip that contains about 6000 transistors and it used molybdenum disulfide (MoS₂) instead of silicon. It formed a two-dimensional semiconductor only a single molecule thick, so they can be stacked on top of each other. IBM said the molybdenum disulfide chip was 50% faster than a silicon chip while using 70% less electrical energy, but most important of all it can operate at much higher temperatures. A silicon chip needs to be below 80°C ( 353 K), a molybdenum disulfide chip needs to be below 500°C (773 K).  

The area a space radiator needs to be to get rid of a given amount of heat energy is inversely proportional to the temperature raised to the fourth power. (353/773)^4= 0.044;  so if an orbital data center used molybdenum disulfide instead of silicon then its heat radiation panels would only need to have 4.4% of the area that a data center using silicon would need. And molybdenum disulfide is less vulnerable to cosmic radiation than silicon is. I think all this makes the idea of orbital data centers much more practical. Step one, an orbital data center. Step two, a Dyson swarm. 

Silicon Is Over. Here's What's Next

John K Clark    See what's on my list at  Extropolis

evt