Worried about your datacenter carbon footprint? Why not put it in orbit?
- Reference: 1668607515
- News link: https://www.theregister.co.uk/2022/11/16/datacenter_orbit/
- Source link:
[1]Thales Alenia Space , a joint venture between European aerospace and defence companies Thales and Leonardo, announced that it has been selected by the European Commission to lead the ASCEND feasibility study, which stands for Advanced Space Cloud for European Net zero emission and Data sovereignty.
The impetus behind this project is apparently environmental, according to Thales Alenia Space. The ever-expanding market for IT services means that datacenters in Europe and across the globe are growing at a rapid pace, which in turn has an impact on energy consumption and can harm the environment.
[2]
It isn't perhaps an obvious solution, but Thales Alenia Space will lead a consortium of companies examining whether datacenters can be put into orbit, powered by solar energy and use optical communications to provide a data link with facilities on the ground.
[3]
[4]
The company said the project could help meet Europe's Green Deal goal of achieving carbon neutrality by 2050, and would also be an "unprecedented development in the European space and digital ecosystem."
According to Thales Alenia Space, the energy to power such orbital datacenters could come from solar power plants capable of generating several hundred megawatts, while the link with the ground would use high-bandwidth internet connections based on optical communications, "for which Europe has mastered the underlying technologies."
[5]
The first objective of the study will be to assess if the CO 2 emissions from the building and launching of the space infrastructure would be significantly lower than the emissions generated by simply building more ground-based datacenters.
A second objective will be to prove that it is possible to develop the required launch solution – ie, that it is actually feasible to stick a datacenter on a rocket and put it into orbit in a working state – and to ensure the operability of the space-borne datacenters.
According to Thales Alenia Space, the latter could be accomplished using robotic assistance technologies currently being developed in Europe, such as the [6]EROSS IOD demonstrator.
[7]US Department of Energy has $42m to make datacenter cooling more efficient
[8]US warns cryptominers must cut power use to avoid busting US carbon goals
[9]Datacenter operator groups pledge to cut water consumption
[10]EU lawmakers vote to ban sales of combustion engine cars from 2035
EROSS IOD (European Robotic Orbital Support Services In Orbit Demonstrator) is a project to develop technology for in-orbit servicing of satellites in order to extend their usable life, and is also led by Thales Alenia Space.
The idea may sound wacky, but at least the European Commission is stumping up the cash to investigate how achievable it is. It seems likely that an orbital datacenter would be constrained by what fits atop the largest launch vehicle, and thus would be much smaller and have far less capacity than ground-based datacenters, unless they can perhaps be linked together in orbit.
[11]
Other concerns would be that it would be very difficult to send out an engineer to fix a failed storage array, and the lifecycle of servers means that the infrastructure would be out of date after about five years or so.
ASCEND was detailed in a [12]factsheet published in October by the European Commission and stems from the 2021 Horizon Europe calls for space-based proposals. It states that ASCEND will receive a grant of over €2 million ($2.08 million).
The consortium includes numerous companies with relevant areas of expertise, including the environment (Carbone 4, VITO), cloud computing (Orange, CloudFerro, HPE Belgium), launch vehicles (ArianeGroup) and orbital systems (German aerospace center (DLR), Airbus Defence and Space and Thales Alenia Space). ®
Get our [13]Tech Resources
[1] https://www.thalesgroup.com/en/global/activities/space
[2] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_onprem/front&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=2&c=2Y3UXLnnLRZz3acZ7i2sbzwAAAQA&t=ct%3Dns%26unitnum%3D2%26raptor%3Dcondor%26pos%3Dtop%26test%3D0
[3] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_onprem/front&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=4&c=44Y3UXLnnLRZz3acZ7i2sbzwAAAQA&t=ct%3Dns%26unitnum%3D4%26raptor%3Dfalcon%26pos%3Dmid%26test%3D0
[4] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_onprem/front&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=3&c=33Y3UXLnnLRZz3acZ7i2sbzwAAAQA&t=ct%3Dns%26unitnum%3D3%26raptor%3Deagle%26pos%3Dmid%26test%3D0
[5] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_onprem/front&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=4&c=44Y3UXLnnLRZz3acZ7i2sbzwAAAQA&t=ct%3Dns%26unitnum%3D4%26raptor%3Dfalcon%26pos%3Dmid%26test%3D0
[6] https://eross-h2020.eu/
[7] https://www.theregister.com/2022/09/26/coolerchips_program_doe/
[8] https://www.theregister.com/2022/09/09/crypto_mining_regulations_us/
[9] https://www.theregister.com/2022/07/28/climate_neutral_datacenter_pact/
[10] https://www.theregister.com/2022/06/10/european_union_votes_to_ban_gasoline_cars/
[11] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_onprem/front&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=3&c=33Y3UXLnnLRZz3acZ7i2sbzwAAAQA&t=ct%3Dns%26unitnum%3D3%26raptor%3Deagle%26pos%3Dmid%26test%3D0
[12] https://defence-industry-space.ec.europa.eu/results-horizon-europe-space-related-calls-2021-2022_en
[13] https://whitepapers.theregister.com/
I don't need an expensive feasibility study,....
... to know this is a really dumb idea. I mean really dumb. So totally dumb I'm struggling to see how people are even putting their name to this. So, a 1U server costs upwards of a few grand, depending on spec, but would cost 20 X that to get it into LEO, so three grand for the server, $60k+ just to get it into space, and that's before the R&D of the 'datacentre' that operates in the vacuum of space, and all the launch costs and energy requirements that go along with it, plus comms, and well, engineer visits.
Re: I don't need an expensive feasibility study,....
But telling your honourable colleagues that they are seriously hula-hula-bonka-bonka-ta-tu-ta-tu-mad is not a career improving move.
The correct handling is to profit from this stupid idea long enough to get settled for life. Then you get out, just before the real stupidity of the project hits the scandal and sensation newspaper(s).
Continuity?
" examining whether datacenters can be put into orbit [using] optical communications to provide a data link with facilities on the ground "
In my experience of medium haul free space optical communication the big problems were aiming precision and transient obstructions. These were tough enough to deal with when we did it over a couple of km with the occasional pigeon crossing the path, but from even low Earth orbit there are so many ways the signal can be interrupted that it seems highly unlikely to be reliable enough for business purposes.
Re: Continuity?
Besides, think of the cost of the fibre!
Mind you, you could use it as a space elevator... that would make it easier when you need to swap hardware
It's much more feasible that there will be ...
Indeed, downright stupid if you ask me.
There's a long list of things that can/will affect it before it starts to melt down, among them:
- man generated space rubbish
- small/tiny undetectable meteorites
- solar flares/radiation
- etc., etc.
... and last but not least, a strike from a weapon.
O.
Utter nonsense - even for the claimed goal
OK: they want to get to "net zero" carbon for data centres.
How many tons of fuel does it take to lift, say, one ton of server racks (plus batteries and solar panels plus radiation shielding) into orbit?
And how does that compare with the emissions for that same rack of servers on the ground, over 5 years, powered by any current representative power mix?
And what about the waste when this de-orbits?
Some ideas are just too stupid to spend real research money on. That doesn't stop people claiming it, but funding committees should have more sense.
Low carbon?
it needs a fucking rocket to get there.
Once there, there's zero cooling as there's no convective cooling in space.
Why?
High School Paper?
This is about as dumb as an idea as you can get. Some scared high school student (the world is going to end unless...) write this?
Net Zero is what you'll have in your bank account when your industry can't compete due to the high cost of "green" energy. It's already happening at scale in Europe.
Smart countries will ignore the climate scam and profit by importing those industries using reliable and low-cost hydrocarbons.
Re: High School Paper?
Not so smart countries will build Drax power station powered by 'wood waste' then import thousands of tons of wood pellets from virgin (or near virgin) forests around the globe
Re: High School Paper?
"Not so smart countries will build Drax power station powered by 'wood waste' then import thousands of tons of wood pellets from virgin (or near virgin) forests around the globe"
Actually they built Drax on top of a coal mine with direct access to "clean coal" ... Nasty miners cost money so they shut the mine and imported open cast, brown "dirty coal" from Poland ... Nasty brown coal is not very green so they then converted 1/6 to burning wood as an "experiment" which they import from "not at all" virginal forests and they are assured that the forests are "fully replanted" as plantation monocultures ... So that's cleared that up then.
Remember
This is a government study to figure out using green madness. The launch Co2 will probably be discounted by some clever guestimate of Co2 savings before it gets up there. Or just be ignored in the name of green. When it comes to green madness everything is flexible including the price
What are they smoking!
We've got countless useful things to be doing down here with the heat from DCs, all of which are a much smaller engineering challenge than lofting racks onto space.
Even low grade heat has it's used if the right planning, zoning and incentives can be put in place.
Plus when one of these renters the atmosphere what's happening to all the random element plasma that gets dumped in the upper atmosphere?
Re: What are they smoking!
random element plasma
Ah, that's the secure data destruction feature, sir!
out of sight out of mind
and GREEN!!! too, what's not to like? Until it decides to de-orbit rather faster than designed.
Data sovereignty?
I'll bet the lawyers are salivating at the thought of the arguments to come about exactly who has sovereignty of the data in an orbital DC.
A daft idea
There is no atmosphere in space, so no aircooling is feasible. The only way to get rid of heat is by radiating it away.
Out of date after 5 years ? That's the lucky scenario. It's much more feasible that there will be a heat meltdown before that.