No robot-built data centers on Earth, let alone the Moon • The Register


Opinion Calling on the quality of a new idea in technology can be difficult. But if you ask me, not in the case of Lonestar Data Holdings, whose plan to build data centers on the Moon is literal madness.

Every detail of the roadmap, from tiny tentative proofs of concept to huge underground server farms built and maintained by Moon robots, is priceless nonsense. Since Apollo, every spacecraft has data storage and network access. We retrieved the data contained in New Horizon’s 16GB filing system from the Kuiper Belt object Arrokoth, 16,000 times further away than the Moon. Damn well proven concept.

As for building bit lairs in lava pipes by robot, no one has built a robot data center on Earth yet. And nobody seems to want to try.

Since the current rule of thumb is that landing a kilo of stuff on the Moon costs around $1 million, with an HP ProLiant DL380 G5 storage server clocking in at around 30kg, even FedEx suddenly seems like a bargain. Everything about space is very expensive, very difficult, and absolutely reserved for things you can’t do otherwise.

The biggest mystery about LoneStar is how on Earth it got $5 million in seed funding, just enough to land a laptop, but tackling that mystery is surprisingly helpful.

The proposal has resonance because it seems like an exciting sci-fi answer to problems we instinctively know exist. What LoneStar is good at is inducing a thought experiment. His ultimate justification is that our home world is too dangerous a place to keep our most valuable data. Bombs! Earthquake ! Hackers ! It’s worth adding an insanely expensive extra tier to Amazon’s storage offerings to make things really, really secure.

You might think you’d be better off with a bunch of high-reliability data centers underground all over Earth, and you’d be right, but that’s beside the point. Everyone – individuals, companies, nations, and cultures – has their data crown jewels, and everyone is instinctively ready to protect them more than the mundane. Identifying what these jewels are, how much they are worth and how best to protect them is a very useful exercise.

The real Crown Jewels of the United Kingdom, a set of royal ceremonial objects kept in the Tower of London, wonderfully illustrate the problems. Their monetary value is unknown, but given that the British Crown is the symbol of national authority, they are more than significant. While easy to identify, it turns out sticking them in a large castle marked “Crown Jewels Here” isn’t entirely safe. They’ve been stolen, melted down and sold, with a nearly full reboot needed after controversy 17th century leader Oliver Cromwell disposed of the lot. (The Scots cleverly hid theirs during the Interregnum*, but bits continued to be stolen afterwards.) Putting your data equivalents somewhere super safe, even if it’s on the Moon, will attract Warning.


Cost is the main obstacle to extra-special guarantees for extra-special data. It is expensive to classify and refresh categories of data worthy of attention. Conversely, because bulk data storage pricing is inversely proportional to latency, very slow storage tiers used for disaster recovery backups are the cheapest per byte, so you can treat everything the same. And since there are legal limits on how long things can be kept, that’s often the event horizon of the whole case.

This attitude is short-sighted. It’s perfectly fine, if you can’t identify your crown jewels now, to keep them wherever they might be for later. But what? Some data may seem useless after a few milliseconds, such as market conditions that you obtained slightly ahead of your competitors and allow you to make an advantageous trade.

Ditto the grocery bill from last August or the stats from an app you just pulled. Yet data is not just data, it is also context, and as any historian will tell you, as data ages and its primary utility fades, the picture it paints of this that was happening at the time is often increasingly valuable. That August grocery bill is much more interesting now than it was in January.

Terabytes of daily routine encode signals about what a company is doing right and wrong, not just for today, or the quarter, or even the past five years, but signals that can be read decades from now. coming.

Countries jealously guard their national archives; we now have the tools, from the individual to the SME via the company, to think in the same way. Data that is so expensive to obtain and process does not lose its value lightly, but it transforms over time. Your future AIs will love it.

Even if you don’t change anything about data storage, thinking this way takes us out of short-termism and sharpens the appreciation of long-term strategy and lasting value, and there are very few organizations of any size who don’t need more. But the tangible benefits of having a substantial archive of data for future AIs to sample, while hard to define, are just as hard to deny.

Businesses thrive or die based on the quality of their perception of the market and the decisions they make accordingly. The best record of your outfit’s evolution is there for as long as you choose to keep it.

But beware. The built-in context may not be suitable for you. The finest and most conspicuous jewels of the British Crown Jewels, the Koh-I-Noor Diamond and the Cullinan Diamond, are there by dint of historically recent conquest, a problematic symbolism that will ultimately cause controversy over conservation by the British sculptures of the Greek Parthenon. the British Museum (where they are known as Elgin Marbles according to the British Ambassador to the Ottoman Empire who took them) look like a game of cribbage. Perhaps today’s data emperors, the Googles and the Metas, would like to ponder this.

The ethics of where your most valuable data comes from is up to you, but thinking about it is also a natural by-product of asking yourself what the true value of your data is over time and what might be worth be preserved. .

There will be data centers on the moon one day, but only when they are the best answer to the questions, not just the dumbest ones. For now, the questions themselves, of risk versus value, long term versus short term, evolution versus mere survival, are worth every moment spent on them.

This is as true of a harassed data manager in a 10-year-old company of one hundred people as it is of a 96-year-old leader of a 966-year-old dynasty. The long term has its merits. ®

* The period in English history when the throne was vacant between the execution of Charles I in 1649 and the Restoration of Charles II in 1660.


About Author

Comments are closed.