When Elon Musk announced plans to launch AI data centers into orbit, many knew it was not true.
The world of data centers is ultra competitive on an industrial scale, with big players relying on a variety of advantages to deliver more cost effective solutions to clients.
Iceland relies on cold weather to save on server cooling costs. Google experimented with container based servers on ships cooled by ocean water in an effort to reduce cooling and real estate costs. Other big players receive large tax breaks or subsidies to setup operations in various regions in exchange for the number of jobs they create.
Bitcoin mining illustrates yet another major cost point of data centers: electrical consumption. Data centers have opted to setup operations in texas to leverage newly constructed wind turbines offering electricity at lower cost.
During the peak of the NASA era the space shuttle was able to launch 1 kilogram of cargo into orbit at a cost of more than $50,000. Space X improved considerably on this model, 1 kilogram of cargo can now be launched at around $2,000.
While the price of launching cargo into space diminished significantly, it still represents additional overhead which would make data centers launched into space less cost effective. There is also a question of how servers in space would be powered given their incredible energy consumption. With cooling being a not insignificant factor.
Given the competitive arena of data centers, and the disadvantages a space based format would suffer, space x will not follow through with announced plans to launch AI servers into orbit...
This should not be considered a fault of Elon or space x.
It only means Space X's IPO was a stepping stone to generate capital to fund planned mars missions. To achieve this, increasing the volume of rocket traffic from earth to space to reduce costs and normalize large capacity shipping becomes necessary. Elon will build businesses to establish this precedent. Starlink was the first step. Future steps are still in their planning stages.
The idea of putting AI data centers in orbit is merely a placeholder, to be thrown away, when Elon comes up with the real plan.
