When I first read about the proposal to build a supercomputer on the moon, I thought it must be some sort of joke. After all, didn't Kubrick teach us about the perils of space computers in 2001: A Space Odyssey? However, the more I learned about the plan, the more I began to wonder if perhaps it had some merit.
A University of Southern California graduate student, Ouilang Chang, presented his idea of building a lunar supercomputer at a recent space conference. His idea came as a possible solution to the coming deep space network congestion.
The problem with the current deep space network has been a concern for NASA scientists for some time. Currently, the Deep Space Network (DSN) consists of huge satellite dishes in Australia, Spain and the United States. The DSN is used to communicate with spacecraft, including the International Space Station. It also gathers and stores important spaceflight data to send back to Earth. Already, this network is showing signs of being overloaded and there is competition over bandwidth. For example, currently the NASA space probe, Voyager 1, which was launched 35 years ago still uses the DSN to communicate with Earth and send data back to the scientists who are analyzing the first data from interstellar flight.
Chang's solution is to build a large supercomputer deep inside a lunar crater. By using a crater on the side of the moon that faces away from Earth, this nuclear-powered computer would be able to process data and serve as a hub for future space travel. The plan also calls for radio dishes that would be used in unison with ones on Earth through laser technology.
The idea is interesting, but it still has several problems. For example, how would a giant system like this be cooled? Chang proposed building the structure near one of the moon's poles, allowing the frigid temperatures to help dissipate the heat and he has plans for an additional water cooled system. Although it's unclear how much water there is to be had at the lunar poles, this type of structure could not only help ease congestion on the DSN, but also the logistics of future moon settlements.
Sound insane? Stephen Hawking, a world renowned cosmologist, has been calling for a huge investment in developing the means to create a colony on the moon. According to him, a settlement on the moon would serve as base for deeper space travel. Although it's not likely that the United States will be returning to the moon any time in the near future, there are many private companies that are prepared to travel to the moon. In fact, Elon Musk's SpaceX is preparing to launch a resupply mission to the International Space Station. As commercial flight to the moon grows, it may become more realistic to establish a base on the moon, and a computer hub would certainly make creating the technology easier. But how much would it cost?
First, we have to consider shipping costs. Unfortunately, Amazon Prime doesn't provide free shipping to the moon. Currently, it costs about $50,000 per pound to ship materials to the moon. Added to that, there is the cost of excavating the selected crater and building the supercomputer center and the nuclear power generator. The project could easily run into the $20 billion range, making it the most expensive supercomputer in history.
These costs don't include labor cost for the actual construction and maintenance of the system. Can you imagine how difficult it would be to house a construction crew during the build? Then, what about system admin? Although it's possible to do a great deal remotely over the network, there are still times when you have to physically be in front of the machine. Can you imagine the commute? Your job site would be cold, dark and extremely lonely but your office would have one heck of a view.
Image courtesy of NASA
Become a member to take advantage of more features, like commenting and voting.
Register or sign in today!