Earthrise over the surface of the Moon

Graduate student Ouliang Chang and Professor Madhu Thangavelu of the University of Southern California’s Viterbi School of Engineering are suggesting that NASA should build a nuclear-powered mainframe supercomputer center on the Moon. For some reason, Wired Magazine is taking the idea seriously.

Chang wants to bury the supercomputer in a cold crater on the far side of the Moon, away from Earth-based radio chatter. The apparent justification for the center is to replace NASA’s Deep Space Network, which is rapidly becoming overloaded.

While it’s true the Deep Space Network is suffering capacity problems, the problem is due to a shortage of antennas and receivers, not computers. New antennas will still be necessary, and building such antennas on the Moon won’t be cheap – at least, not while shipping cargo to the Moon costs $50,000 per pound, as Chang and Thangavelu assume.

It’s even more odd that Chang and Thangavelu are calling for a giant mainframe computer center when the most powerful computer ever sent into space is an Apple iPhone. NASA technologists are now looking at ways to make space probes smarter. By processing the raw bits in deep space and sending only important data back to Earth, they hope to greatly reduce the demands being placed on the Deep Space Network. Distributed computing is the wave of the future, yet Chang and Thangavelu are proposing a monolithic centralized computer center straight out of 1960’s science fiction. (Even Wired comments on a similarity to Stanley Kubrick’s 2001: A Space Odyssey.)

Moreover, the cost of decentralized computer systems won’t be anywhere near the $10-20 billion Chang and Thangavelu are asking for their Moon-based supercomputer (moonbase not included).

It’s time for lunar-development advocates to put aside fantasies like this and acknowledge reality. Routine, low-cost access to space is a necessary prerequisite for any type of sustainable lunar development. No one is going back to the Moon, to stay for any length of time, while space transportation costs $50,000 per pound.

Written by Astro1 on October 2nd, 2012 , Space Policy and Management

Leave a Reply

Your email address will not be published. Required fields are marked *

COMMENTS
    Fred Bourgeois commented

    There are far greater areas of lunar development than super-computing centers, and many essential elements to long-term sustainable lunar business development. The real failure in the proposed approach is the propnents belief that this might happen under the auspices of NASA. Take a look at the annual budgets, and you realize that not only are the real (nominal) dollars allocated to NASA declining or remaining stagnant, but as a percentage of the federal budget, NASA now receives barely 0.50% (down from a peak of about 4.4% in 1966). While the nominal dollar amounts allocated to NASA under Obama have remained stagnant, as a percentage of the federal budget NASA spending has decreased from 0.60% in 2008 under G.W.Bush to an estimated 0.48% this year — the lowest percentage since 1960 — under President Obama.

    Space development and colonization must move out of the public sector and into the private arena, commercialization is the only option remaining. The opportunities available to forward-thinking entrepreneurs and investors are just waiting to be pursued, developed, and capitalized upon. Many of us in this industry are already working on these things.

    A free people demand a new frontier.

    Fred Bourgeois
    Creator, Open Source Space Exploration
    Found & CEO, Team FREDNET, The Open Space Society, Inc.

    Reply
    October 2, 2012 at 5:40 pm