News Startup Plans Nuclear-Powered Data Centers on the Moon

[Admin]

The moon can be an ideal place for datacenters in the distance future: easy cooling, loads of power, but still very expensive.

Startup Plans Nuclear-Powered Data Centers on the Moon : Read more

 

[-Fran-]

It's not THAT bad of an idea, TBH... For workloads that don't depend on latency and they're "fire and forget" types, this can be a good thing. I'm sure it'll be expensive to maintain though... Can you imagine upgrading those machines? LOL. "Honey, I need to upgrade some hardware on the Moon... No, I'm not trying to make excuses to not see your mother... Yes, it'll take 6 months... I don't sound happy, no...".

The solar array they'll need would be massive as well. This wouldn't be cheap, but at least it makes a sliver of sense to me.

Regards.

 

[gg83]

How do you protect the servers from debris that hits the moon? Mini-meteors and such? Maybe the servers will be buried?

 

[jp7189]

Wouldn't floating a datacenter in orbit have all these benefits without the latency and intermittent connectivity?

 

[gg83]

Wouldn't floating a datacenter in orbit have all these benefits without the latency and intermittent connectivity?

Maybe the access to helium-3 for the reactors? I'd like to learn more about it.

 

[USAFRet]

What benefit is there to having this on the Moon?
Cooling? Antarctica or sink in the ocean
Solar power? There are multiple desert areas. Or wave energy from the ocean. Or temp diff in the ocean.

There are a zillion drawbacks.
Transmission time, radiation, UV, micrometiorites...


I see this proposal as a pump and dump. "Gimme money!"

50 years from now, maybe.

 

[ex_bubblehead]

I'd like to know where they get this "easy cooling" from. The moon has no atmosphere (to speak of), thus radiative cooling is virtually non existant. This is a very real problem that has to be dealt with on everything we've ever sent into space.

 

[digitalgriffin]

Why am I reminded of this?

 

[digitalgriffin]

I'd like to know where they get this "easy cooling" from. The moon has no atmosphere (to speak of), thus radiative cooling is virtually non existant. This is a very real problem that has to be dealt with on everything we've ever sent into space.

Yep. Sun side gets cooked, while the dark side freezes and no natural convection. Thus all heat has to be radiated in IR. It's a real challenge for some satellites.

 

[watzupken]

This to me is a case of trying to create a cool idea to make money, but without thinking of potential impacts. Just because the moon is not part of earth does not mean that if we lose the moon, we will be fine. Just because we have destroyed the earth, and now its time to destroy something else. Everything in creation serves its purpose.

 

[ex_bubblehead]

Yep. Sun side gets cooked, while the dark side freezes and no natural convection. Thus all heat has to be radiated in IR. It's a real challenge for some satellites.

And on satellites and planetary probes what direction, and how much in that/those directions, as good old Newton and his "equal and opposite reaction" comes into play (the Pioneer anomaly)

 

[-Fran-]

You can create micro-atmosphere for heat transfer within the box and route it to ground and just dig a hole for not burning to a crisp for direct exposure. The Moon has plenty of those already xD

Regards.

 

[USAFRet]

I'd like to know where they get this "easy cooling" from.

Investorspeak for "Its very cold in space, so...easy!"

 

[pixelpusher220]

I'd like to know where they get this "easy cooling" from. The moon has no atmosphere (to speak of), thus radiative cooling is virtually non existant. This is a very real problem that has to be dealt with on everything we've ever sent into space.

On the moon itself this is fairly trivial. It's basically a lunar-thermal 'water' loop. There's no radiative in 'space' since no atmosphere. But running piping through the regolith is nothing different than a geothermal heat pump system.

Do the human habitats need to be heated? I'm assuming so, so there's a heat sink you could use as well.

 

[Eximo]

Only thing that makes sense to me would be where you:

A) Need gravity
B) Want a cost-free Vacuum
C) Need ZERO electrical/radio interference (ie dark side of the moon) You would still have to shield against natural radio emissions, but there would be no man-made interference
Actually one good reason to put a radio telescope on the darkside.

Otherwise getting stuff assembled on the Moon is a huge undertaking.

 

[ex_bubblehead]

Just a small thing here "dark side of the moon" as people think of it doesn't exist. Every part of the moon (except a couple of craters at the poles) sees the sun at least once a month. Proper terminology is "Near side" and "Far Side" or "Front Side" and "Back Side".

 

[Denniswingo]

The moon can be an ideal place for datacenters in the distance future: easy cooling, loads of power, but still very expensive.

Startup Plans Nuclear-Powered Data Centers on the Moon : Read more

Tom, just read your article. Somewhat mystified regarding the meteor danger. Here are the numbers..

(I am the CEO of Skycorp by the way, the ones building the payload for Lonestar).

Surface area of the Moon. = 38 million square kilometers.

Area of the lander 25 square meters....

Assume (which is not correct, but it is a simplification) that the distribution of material is uniform..

1.4 tons (I assume here imperial tons is what was recorded, but lets use metric tons to make it easier) = 1400 kilograms or 1,400,000 grams

1,400,000 grams over 38,000,000 square km is about 0.037 grams per square km.

So, lets look in square meters...

1,400,000 grams over 3.8 x 10^ 13 square meters = 368 micrograms per square meter per year.....

now obviously there is a distribution curve for sizes, but it can easily be seen that on human time scales in tech (decades) that this is not an issue....

As for the rest of the comments, there is a lot of misunderstanding about the environment of the Moon. Lets deal with them (and yes I am a subject matter expert in this area for NASA and DoD...

1. power....

In the lunar polar regions there are the areas of almost eternal sunlight as the Moon's tilt relative to the sun is only ~3 degrees and with the topography of the Moon, there are many areas that have sunlight almost continuously, and with a solar array on even a modest (10 meters) it is about 98%. Thus a solar array on the Moon with a fuel cell for the shadowed periods would work, up to the lower megawatt levels easily. For higher power, you would want nuclear, but that is a ways off.

2. Cooling...

The majority of cooling on the Moon is radiative (the person above mistook convection for radiation) toward the 4 degrees kelvin deep space heat sink. Then it is just surface area ratio. The same heat pipe technology used to cool hot CPU's works great on the Moon. On the Earth it relies on convection but just convert that to surface area and its good.

3. Other "events"....

One person above talked about burying the data center. You would certainly do that long term and one meter of lunar regolith equals the radiation protection of the earth's atmosphere. The hardware that we are flying for Lonestar is currently on the International Space Station and has been operating for well over a month, with a six-9 month mission life for its qualification. We are running Yocto Linux on a quad RISC-V cpu and all is well.

Events that you have here on the Earth but not on the Moon.

1. Earthquakes...

Vantage data centers are within there blocks of our offices here in silicon valley. A nice mag 7.3 earthquake would take those data centers out, and we are overdue for one.

2. Carrington Events....

This, if it happened today, would instantly transform much of our civilization back to the 18th century. At least the data would still be there on the Moon as a Carrington event does its electrical thing due to the coupling between the ionosphere and the Earth with the atmosphere as the dielectric. Not fun. In 1989 one took out the Canadian power grid, and if the one in 1859 happened today, buy buy Tomes Hardware page (and most everything else not shielded).

3. War

Ever hear of EMP? Much like a Carrington event, but on a smaller scale. The Norks are just itching to do this and not only would it take out all satellites in Low Earth Orbit (This happened in 1964 accidentally), it would take out large sections of our national grid, and all those data centers underneath. Never discount what Vlad the bad might do either....

4. Hacks

Much easier to deal with, but that is between us and our customers.....

Think Class V data centers. Yes, I know they don't exist today, but they will...

All in all this is a proof of concept mission that will be followed by more ambitious efforts.

Just remember friends and neighbors in a few years Elon Musk will be able to put 100 tons on the Moon for less than $100m dollars...

that will change the world.

I will answer any reasonable questions on this mission for those interested.

 

[-Fran-]

Tom, just read your article. Somewhat mystified regarding the meteor danger. Here are the numbers..

(I am the CEO of Skycorp by the way, the ones building the payload for Lonestar).

Surface area of the Moon. = 38 million square kilometers.

Area of the lander 25 square meters....

Assume (which is not correct, but it is a simplification) that the distribution of material is uniform..

1.4 tons (I assume here imperial tons is what was recorded, but lets use metric tons to make it easier) = 1400 kilograms or 1,400,000 grams

1,400,000 grams over 38,000,000 square km is about 0.037 grams per square km.

So, lets look in square meters...

1,400,000 grams over 3.8 x 10^ 13 square meters = 368 micrograms per square meter per year.....

now obviously there is a distribution curve for sizes, but it can easily be seen that on human time scales in tech (decades) that this is not an issue....

As for the rest of the comments, there is a lot of misunderstanding about the environment of the Moon. Lets deal with them (and yes I am a subject matter expert in this area for NASA and DoD...

1. power....

In the lunar polar regions there are the areas of almost eternal sunlight as the Moon's tilt relative to the sun is only ~3 degrees and with the topography of the Moon, there are many areas that have sunlight almost continuously, and with a solar array on even a modest (10 meters) it is about 98%. Thus a solar array on the Moon with a fuel cell for the shadowed periods would work, up to the lower megawatt levels easily. For higher power, you would want nuclear, but that is a ways off.

2. Cooling...

The majority of cooling on the Moon is radiative (the person above mistook convection for radiation) toward the 4 degrees kelvin deep space heat sink. Then it is just surface area ratio. The same heat pipe technology used to cool hot CPU's works great on the Moon. On the Earth it relies on convection but just convert that to surface area and its good.

3. Other "events"....

One person above talked about burying the data center. You would certainly do that long term and one meter of lunar regolith equals the radiation protection of the earth's atmosphere. The hardware that we are flying for Lonestar is currently on the International Space Station and has been operating for well over a month, with a six-9 month mission life for its qualification. We are running Yocto Linux on a quad RISC-V cpu and all is well.

Events that you have here on the Earth but not on the Moon.

1. Earthquakes...

Vantage data centers are within there blocks of our offices here in silicon valley. A nice mag 7.3 earthquake would take those data centers out, and we are overdue for one.

2. Carrington Events....

This, if it happened today, would instantly transform much of our civilization back to the 18th century. At least the data would still be there on the Moon as a Carrington event does its electrical thing due to the coupling between the ionosphere and the Earth with the atmosphere as the dielectric. Not fun. In 1989 one took out the Canadian power grid, and if the one in 1859 happened today, buy buy Tomes Hardware page (and most everything else not shielded).

3. War

Ever hear of EMP? Much like a Carrington event, but on a smaller scale. The Norks are just itching to do this and not only would it take out all satellites in Low Earth Orbit (This happened in 1964 accidentally), it would take out large sections of our national grid, and all those data centers underneath. Never discount what Vlad the bad might do either....

4. Hacks

Much easier to deal with, but that is between us and our customers.....

Think Class V data centers. Yes, I know they don't exist today, but they will...

All in all this is a proof of concept mission that will be followed by more ambitious efforts.

Just remember friends and neighbors in a few years Elon Musk will be able to put 100 tons on the Moon for less than $100m dollars...

that will change the world.

I will answer any reasonable questions on this mission for those interested.

Interesting stuff.

Out of curiosity, will you be implementing "self healing" nodes for hardware redundancy and mirroring? Have you considered a microwave-type protocol for comms to Earth? Something like re-using the good 'ol VHF bands? I ask this, because the spectrum is overly abused by all the Satellites and Earth stuff already there, so ensuring 24/7 upkeep of comms for monitoring and instrumentation sound like an interesting challenge. Upgrade cycles? Preferred vendors and/or in-house custom platform (like outsourcing how the server would be to Dell or HP, for instance or making your own).

Also, will you be providing paid leave to tour the Moon for technicians making repairs? Haha.

Regards.

 

[froggx]

(Quoted from the article)
"While there are no fires, floods, earthquakes, tornadoes, and other natural disasters on the Moon, the planet has another major problem... "
(End quote)

The Moon is, well, a moon, not a planet...

3. War

Ever hear of EMP? Much like a Carrington event, but on a smaller scale. The Norks are just itching to do this and not only would it take out all satellites in Low Earth Orbit (This happened in 1964 accidentally), it would take out large sections of our national grid, and all those data centers underneath. Never discount what Vlad the bad might do either....

Couldn't have been 1964 with the partial nuclear test ban treaty of 1963 prohibiting atmospheric and outer space nuclear testing. The 1962 Starfish Prime test certainly did this, although detonating a >1 Megaton thermonuclear warhead in outer space isn't the kind of thing that happens "accidentally."

The hardware that we are flying for Lonestar is currently on the International Space Station and has been operating for well over a month, with a six-9 month mission life for its qualification.

I'm probably misunderstanding this, but it sounds like you're saying you're building stuff to send to the moon with a 6-9 month guaranteed lifespan. I don't see the point of this if you're having to replace everything almost twice a year.

 

[digitalgriffin]

Moon period and orbit match at 27 days. So the same side is always facing us. The question is why? It's not like a geosync satellite that stays stationary. The time it is in line of sight constantly shifts. I still don't understand the purpose.

While I get putting the centers near the poles to control heating and cooling, there's a distinct issue with radiative heat and reduction in solar power generation.

 

[USAFRet]

I still don't understand the purpose.

"We have this REALLY good idea! Give us money"
'OK'
<couple years later> "Oh, this isn't going to work like we thought. Sorry. Company is folding" (as he drives to his new beach house in a new Tesla)

Some decades from now, this may be a thing. Not anytime soon, though.

 

[Denniswingo]

Interesting stuff.

Out of curiosity, will you be implementing "self healing" nodes for hardware redundancy and mirroring? Have you considered a microwave-type protocol for comms to Earth? Something like re-using the good 'ol VHF bands? I ask this, because the spectrum is overly abused by all the Satellites and Earth stuff already there, so ensuring 24/7 upkeep of comms for monitoring and instrumentation sound like an interesting challenge. Upgrade cycles? Preferred vendors and/or in-house custom platform (like outsourcing how the server would be to Dell or HP, for instance or making your own).

Also, will you be providing paid leave to tour the Moon for technicians making repairs? Haha.

Regards.

We are building the payload for the customer, which is in turn a customer of Intuitive Machines, that is building the lander. Intuitive Machines has their FCC and International filings. Chris Stott, CEO of Lonestar, was the founder and CEO of ManSat, the world's premier licensing entity for spectrum coordination around the world. The bandwidth on this first mission is very limited, so the range of datacenter activities will be focused on data integrity testing in that environment. This is a key datapoint going forward in the future as Lonestar expects to build more and more capable systems as time moves on.

This is a proof of principle mission and thus the full spectrum of services will not be provided. The mission duration for the first one is only 14 days. In the future, as Elon Musk's Starship gets flying data centers on the Moon, and in deep space will become a reality.

 

[Denniswingo]

(Quoted from the article)
"While there are no fires, floods, earthquakes, tornadoes, and other natural disasters on the Moon, the planet has another major problem... "
(End quote)

The Moon is, well, a moon, not a planet...





Couldn't have been 1964 with the partial nuclear test ban treaty of 1963 prohibiting atmospheric and outer space nuclear testing. The 1962 Starfish Prime test certainly did this, although detonating a >1 Megaton thermonuclear warhead in outer space isn't the kind of thing that happens "accidentally."



I'm probably misunderstanding this, but it sounds like you're saying you're building stuff to send to the moon with a 6-9 month guaranteed lifespan. I don't see the point of this if you're having to replace everything almost twice a year.

It was Starship Prime. The "accident' was basically wiping out many of the operational satellites in low Earth orbit at the time . There is a report on this somewhere if you want to dig it up. Today, a nuke like that would take out basically everything outside of radiation hardened MILsats.

The point of this activity is a proof of principle. There will be a radiation experiment that goes along with the payload to characterize the radiation environment and its effects on semiconductors (and SSD drives) on the surface of the Moon. You have to crawl before you fly and this mission will gather valuable data for follow on longer duration missions in the future. Lonestar has customers on this mission, who see the panache of being first to do this, but later on there will be much more mundane customers and reasons for doing this.

 

[Denniswingo]

Moon period and orbit match at 27 days. So the same side is always facing us. The question is why? It's not like a geosync satellite that stays stationary. The time it is in line of sight constantly shifts. I still don't understand the purpose.

While I get putting the centers near the poles to control heating and cooling, there's a distinct issue with radiative heat and reduction in solar power generation.

Here is an article I wrote on this subject a few years ago. Solar power generation is BETTER in the polar regions, and radiative cooling is proportional to the area of the radiator.

https://denniswingo.wordpress.com/2...adical-idea-for-lunar-industrial-development/

 

[froggx]

It was Starship Prime. The "accident' was basically wiping out many of the operational satellites in low Earth orbit at the time . There is a report on this somewhere if you want to dig it up. Today, a nuke like that would take out basically everything outside of radiation hardened MILsats.

The point of this activity is a proof of principle. There will be a radiation experiment that goes along with the payload to characterize the radiation environment and its effects on semiconductors (and SSD drives) on the surface of the Moon. You have to crawl before you fly and this mission will gather valuable data for follow on longer duration missions in the future. Lonestar has customers on this mission, who see the panache of being first to do this, but later on there will be much more mundane customers and reasons for doing this.

Starfish Prime was awesome. I wish I could have been at the meeting where someone said, "Let's set off some nukes in space, it'll be awesome!"

Is the ISS a good analog to the radiation environment that you would expect on the moon? I could be wrong but isn't the ISS close enough to Earth for the magnetosphere to most of the solar wind and other assorted charged particles whereas the moon doesn't have that benefit? It's always been somewhat fascinating to me how much extra goes into the hardware for meant to leave Earth. Reading about the effects of radiation on the systems installed on craft like Galileo and Juno so I'm curious. Is there a reason you're looking at SSDs rather than HDDs or other storage options?

2. Carrington Events....

This, if it happened today, would instantly transform much of our civilization back to the 18th century. At least the data would still be there on the Moon as a Carrington event does its electrical thing due to the coupling between the ionosphere and the Earth with the atmosphere as the dielectric. Not fun. In 1989 one took out the Canadian power grid, and if the one in 1859 happened today, buy buy Tomes Hardware page (and most everything else not shielded).

It seems like accessing data on the moon would be somewhat difficult if this happens.