UK company sends factory with 1,000C furnace into space | Hacker News Reader

UK company sends factory with 1,000C furnace into space

bbc.co.uk

150 pointsvekerdyb87 comments

I'm only YouTube-level informed on how silicon manufacturing works, but something that is, perhaps intentionally, not made clear to someone unfamiliar with the field is that this is not manufacturing chips in space. This is to grow the crystals only, the very first step in silicon chip manufacturing. This is how you get the ingot, then you slice it to get the wafers upon which the chips are built. The reason you would even consider doing it in space in the first place is because, on Earth, gravity and other forces are stronger and result in lower-purity crystals. Basically, what I'm getting at, is that I believe this is pretty much a glorified oven. Moving the entire manufacturing process in space wouldn't make sense, as I don't think the benefits to other steps of the process like CVD would outweigh the insane costs of sending things into orbit.

In a way that's good because they don't need logistics for the entire supply and production chain up there, they can just drop the (small and presumably valuable) silicon crystals back to Earth.

Silicon ingots used in chip manufacturing aren’t that small: like a ~400mm diameter cylinder a metre or so long, and from memory several hundred kilos.

Looking up their website, it sounds like the company are making only the seed crystal in space, not the entire silicon boule:

"Space Forge is using space-derived crystal seeds to grow ultra-high quality semiconductor substrates on Earth."

https://www.spaceforge.com/

Note that growing ingots is an incredible feat at that purity and size that they achieve on earth. It’s already a very very hard step in a crazy process for entire chip manufacturing.

Leaving aside the small issue of getting there, wouldn't it be easier to achieve in space given there is fewer impurities like air floating around?

Even given that the sending finished products back to earth in same clean room conditions for next step seems challenging to make profitable. If it’s a proof of concept, okay, but to take it further the lithography step takes RV sized machines.

>this is not manufacturing chips in space

*crisps

It's from the UK.

Chips are definitely a thing in the UK. Like French fries but usually chunkier: https://en.wikipedia.org/wiki/Fish_and_chips

Which is called Fish and Chips in Canada, even though it's served with fries.

I heard some restaurant getting sued for selling "fish and chips" without fish, but I don't remember how it ended

Same in the US

In the Netherlands we have two words for fries and you know if someone is from the north or the south based on their use: Patat, north en Friet, south, particularly in the South people are sensitive to using the wrong, northern word. (And chips are just crisps here.)

What? Every time I see kids on the train they’re talking about going to the appie to buy a redbull and “zakje chips”. I live in Eindhoven though so idk if that plays a roll.

"Zakje chips" is a small bag of crisps (like Lays)... If they go for fries they'd say they go for a "frietje". Eindhoven is distinctly in the South :)

That damn Red Bull though, somehow the kids love it, part of it is probably that their parents keep them away from it. Sugar and Caffeine. Diabetes and poor sleep, great stuff.

In the UK a fish and chips shop is sometimes called a "chip shop". The New York Times helpfully translated this in a recent article:

> “I’ve seen lots of students my age struggling, trying to get work and even the basic necessities,” Agastya Dhar, 17, said. Mr. Dhar has a part-time job in a French fry restaurant, but said even getting that job was tough.

French fry restaurant is now my preferred term for the local chippy. For those outside the UK chip shops normally have no seating, or maybe a couple of uncomfortable, uninviting, flourescent lit plastic benches and tables, normally bolted down, maybe sprayed clean at the end of the night.

https://www.nytimes.com/2025/11/26/world/europe/uk-budget-yo...

The chunkier fries are akin to what the US calls steak fries and are very common in the US as well.

That's gastro pub chips though, not chippy chips.

If you are lucky, you get triple-fried chips. Which are just as good/bad as they sound.

> just as good/bad as they sound.

I.e. extremely good and not at all bad.

healthwise i guess for the bad

Triple-cooked chips are usually parboiled and twice fried; if anything it's probably healthier than a single fry, because the first fry is brief and hot to crisp up the exterior, so the oil absorption's less the second time, and perhaps less overall than a single (necessarily cooler) fry.

The chips in question are deeply brown and shiny, but your reasonable-sounding explanation suits my need for self-delusion so well that I will accept it whether it’s true or not.

What Americans call chips (potato or corn) the UK typically calls crisps.

There is a Douglas Adams inspired The Great British Bake Off subplot out there somewhere.

Do gravity-based defects outweigh displacement defects from fast particles?

Also, there are large headwinds from having to ship up a large quantity of raw material and have to deorbit a payload so fragile that any amount of shock is unacceptable. Maybe a high purity silicon boule pays for these headwinds with room for profit on top. I'm skeptical, but time will tell.

It would be pretty easy for them to have done the math in advance and I’m sure they’re aware that space travel is expensive.

I'm sure they've done the maths and determined that it's not economically feasible at all.

But if it works as a proof of concept, in three or four generations time perhaps they'll have a scalable process which pays for itself.

This exactly. At scale there are a lot of things that are profitable that are idiotic when done Ad Hoc. To get to scale though you need a particular kind of bullshitter who can hold on to the idea that what they say is true and also on some level understand that they have to idea how to do the bullshit they're promoting.

If you've managed to find more details about what process exactly they're implementing I'd be glad to see it - I assumed plasma-based growth, since the BBC article mentions that it's a plasma that is at 1000C here (making heat dissipation less of a problem too), but if they're growing ingots that would usually be done from liquid silicon, which sounds like a mess in space. So are they doing plasma-growth of ingots (which I haven't heard of, but I haven't heard of many things), or are they bringing wafers up and growing ultra-pure layers on top... The website is not super clear on this from what I've seen.

I'm curious what the thermal management system on this looks like. On one hand, vacuum being in essence a perfect insulator works in favor of keeping the silicon hot for the very long time it takes to pull a boule while requiring very little energy. On the other hand, you have to make sure the control electronics don't also heat up to 1000C. I'm also curious how you keep the molten silion separate from the crystal without gravity keeping it in the crucible. I bet a lot of interesting engineering going on here.

Something like a vacuum flask, I imagine. Vacuum is a very good insulator already and you minimise radiative heat transfer (infrared glow) by making a surface shiny and metallic usually (low emissivity)

If you have to bring your own vacuum flask, don't you lose half the benefit of doing it in space?

Isn't the primary benefit the lack of gravity?

You can reduce metals through vacuum pyrolysis at much lower temperatures without a reducing agent if you have a vacuum. This could make industrial scale processing of steel relatively easy on the moon.

Reducing ferric oxide to magnetite, perhaps, but I think if you tried that with ferrous oxide you'd get iron vapor coming off along with the oxygen.

An issue with any high temperature process is things start evaporating. This is part of why carbothermal reduction of aluminum oxide doesn't work: at the required temperature aluminum oxide is volatile.

(There are thermochemical water splitting technologies that exploit partially reducing transition or rare earth oxides at high temperature, then reacting them with steam at a bit lower temperature to make hydrogen. I believe cerium oxides are the current best approach there, although still not competitive.)

I imagine you just need the flask part, the vacuum is rather easy in space.

Good electrical conductors are also good thermal conductors. It's a fun system challenge to minimize what needs to be hot, but some things will have to get hot. It could be reduced to a photodiode, transistor, and a relay.

But how do you get the power to the heater in a compact way?

Diamond is my favorite exception to this, one of the best thermal conductor and insulators.

Boron nitride too. I guess the thermal vibrations transmit well through a stiff microstructure.

One notable exception to this is superconductors. One might naively think that because superconductors have zero electrical resistance, they also have zero thermal resistance. But this is wrong (sorry, Larry Niven)! The superconducting charge carriers (Cooper Pairs) have zero entropy, so they can't carry heat. Thermal conductivity of a superconducting material drops when it becomes superconductive.

I believe high Tc superconductors have been used (or at least proposed to be used) as current leads for carrying current into low Tc superconductors from somewhat higher temperature normal conductors.

Vacuum is only a nearly perfect insulator until a few hundred °C. After that, radiation dominates over every other form of heat exchange, and it stops making any difference.

A couple of meters long steel rod with a dissipator on the end can easily keep electronics at Earth surface temperatures even if you heat the other end to 1000°C.

The air on earth is taking the heat from the dissipators more than blackbody radiation.

4000 times purer seems a questionable claim to me. 4000 times purer then what? Current Earthbound state of the art? A guy in shed with a vacuum pump and a heater? On what axis: crystal defects or contamination?

High vacuums aren't at all impossible on Earth and silicon boules are already single crystals.

What exactly about their process permits such a huge quality improvement?

The best silicon single crystals still contain impurities at the 1E-11 level. This project is about doing crystal growing in low gravity (the ultra-high purity is only achieved due to the growth process). The vacuum of space is a lot worse than what can be achieved in the lab, especially in low orbits.

What I'm struggling with is the hard radiation, which causes defects and even impurities even if you start with isotopically pure feedstock

I had a similar thought regarding radiation - but also curious about cost to continuously launch materials and the overall reliability of recovery. Just seems like a lot of "what ifs" for the sake of finding funding.

There have been experiments to create higher vacuums in space, which have been used to grow semiconductor crystals:

https://scfh.ru/en/papers/vacuum-in-the-wake/

https://en.wikipedia.org/wiki/Wake_Shield_Facility

I'm not saying it's practical, but it's pretty cool.

What is the current defect rate preventing from happening? If you can lower it three orders of magnitude, what then becomes possible?

Earth's surface isn't a microgravity environment. The principle that you can get purer crystalline structures with less interference from the earth's gravitational pull has explored on space stations since the 1970s; that aspect and potential applications in higher performing semiconductors and drugs are fairly well understood

What hasn't been solved yet and Space Forge (and Varda et al for biotech) are hoping to solve are the unit economics of reentry vehicles to actually make it viable for manufacturing.

UK - the only country to achieve to-orbit capability only to shelve it after the first flight.

As an ESA member state, the UK has had launch capabilities via Arianespace for a couple of decades.

France essentially did that too, in favour of the ESA (I think the French launcher may have had a couple launches, but certainly no more than ten).

France has its own agency, the CNES (though most research go through ESA nowaday), and had it for a long time.

Its launchers are still the best when it comes to reliability I believe, though not competitive on cost anymore since the advent of spaceX (Ariane6's first flight was in 2024 and its price per kilogram is just an order of magnitude worse than spaceX). Definitely missed a step.

Still, France has an active and ongoing space program since about 1970.

The Ariane family is, at least formally, ESA, not CNES, tho. The UK also has its own agency, but launches via ESA (or private). I think it would be probably fair to say that Ariane is more French than anything else, but it’s not strictly a French project.

Confusingly, the EU also has its own agency, though it doesn’t, as far as I can see, do much outside of operating Galileo. ESA, though obviously very EU aligned, isn’t an EU agency, and has non-EU/EEA/former-EU members (notably, _Canada_).

One thing that is unclear to me from the article:

Is the idea that it will manufacture all of these chips and then both the 'factory' and the resulting materials will return from space, or that the factory would stay in orbit and send materials back?

Return.

Just a tiny part of the operation will happen in space. The result of that will return, then will be made into chips on earth.

The idea is that the furnace will melt the silicon into a 'boule' (cylinder). <https://en.wikipedia.org/wiki/Boule_(crystal)> The solid boule will be returned to Earth for lab analysis.

The idea is that they get government funding from credulous civil servants. There is no actual idea here, there is no business. The idea that a country that is unable to supply basic infrastructure is suddenly going to build CPUs in space is obvious bullshit.

"This sort of semiconductor would go on to be in the 5G tower in which you get your mobile phone signal, it's going to be in the car charger you plug an EV into, it's going to be in the latest planes."

LOL! Talk about an anticlimax. Either this is a lack of imagination on the CEO's part, or he's dumbing it down for us, or that level of silicon purity/regularity, is one of those nice-sounding but impractical Platonic ideals that ends up being kind of a waste. But it, and/or solving the attendant problems, might be an important precursor for some future innovation.

I think that this is far from the first experiment in growing crystals in space, so they must have good evidence for what properties to expect from single crystal silicon grown in space, my guess is that the purity is not so much which elements are or are not included, but more the perfection of the physical crystaline structure, which may introduce the posibility of the crystal having different characteristics in different planes that can be exploited for various purposes.

The crystal planes indeed do have different properties which are exploited already. Some planes have better flatness and defect rates then others, and etching rates and undercutting varies according to the surface orientation and the orientation of long thin details to the bulk lattice.

Silicon boules are grown with defined orientations. There's a system of flat edges ground into the wafers which indicate the orientation.

> "The work that we're doing now is allowing us to create semiconductors up to 4,000 times purer in space than we can currently make here today," says Josh Western, CEO of Space Forge.

> "This sort of semiconductor would go on to be in the 5G tower in which you get your mobile phone signal, it's going to be in the car charger you plug an EV into, it's going to be in the latest planes."

Okay, but, we have 5G towers, car chargers, and planes right now?

I understand that purer material is better, but to what extent are the impurities of current wafer production methods limiting us? Why is shooting the furnace into space the best option? Why is making wafers 4+ orders of magnitude more expensive the solution we should go for?

There is no reason to do it with current processes. It could possibly reduce defect rate/increase yields but I'm not sure impurities are even leading cause now.

But we might discover other uses, that's what science is.

No way. This is being done because there’s a predictable path to profitability. It’s not just random shot in the dark science you can sometimes see in academia. It’s just this path isn’t clear to us laymen… I know because launching into space isn’t something that will be done just for science

Hubble Space Telescope. James Webb Space Telescope.

I should’ve specified corporate vs. academia. Those telescopes clearly aren’t corporate. More akin to academia.

Corporations won’t go to space just for “science” and that is the case here.

That doesn't mean they're going to space for traditional profitability, though. The advertising budget is a classic example: none of those activities are profitable, but they alter market behaviour such that other parts of the company can make money.

Yeah going to space is not an "advertising budget".

Which doesn't mean it's not another example of the same phenomenon. Many companies do things that are not, and will never be, profitable (in the traditional sense), because they have ulterior motives.

False. The objective a company is to make profit, anything else is a side quest and is less relevant.

There isn’t an “ulterior” motive here. The motive is clear: profit. We just can’t pinpoint the exact pathway to it in terms of details.

Not all companies have "make profit" as their primary goal. But what I meant by "ulterior motive" was that their space factory may never provide enough output for that alone to justify the cost. Their plan may be for the factory in space to bring money in other ways (e.g. making them eligible for grants they would otherwise not have received).

I came across space forge due to some algorithmic discovery on YT a few months ago.

If I am recalling it correctly - They’re focusing on goods that can be made, that have a high $ value vs volume and density. High purity silicon is what they identified.

Come on, it’s the BBC - which has a much wider spectrum of audience than you and me.

Think with it a little - the statement “in the 5G tower” is intended to bring the context of this event closer to those who are not knowledgeable about this technology, but would nevertheless read the article. You and I may understand that the economies of scale don’t make sense yet - but they could, some day, if this technology succeeds, be relevant to the local neighborhood.

To many, the 5G tower is the most mystical, mysterious technology in their neighborhood - and indeed, the silicon ingots being manufactured this way would, eventually, find their way to the local neighborhood if this technology is successful.

It would probably have been more appropriate to say “some day these ingots will power the supercomputers in your pocket”, which would be an accurate statement - but that is a whole order of magnitude of different economic scale than in the industrialization of cell networks. Maybe it’d be more appropriate for the BBC writer to have said that satellites might one day benefit from space-grown silicon wafers - but that is still to distant to the Mom and Pop readership they’re targeting in these articles…

This is a microwave oven. I wonder how feasible is to have a solar concentrator, if you can build it out of foil, mostly, like a light sail.

You would have to 1) keep turning it toward the sun and b) reduce time in earth's shadow, which means a polar orbit?

> a heat shield named Pridwen after the legendary shield of King Arthur will be deployed to protect the spacecraft from the intense temperatures it will experience as it re-enters the Earth's atmosphere.

Why bring back the entire spacecraft and not just the finished product?

I'm also curious how they handle cooling the silicon, since dissipating heat in space is kind of difficult.

Harder engineering to separate the components? Making the oven + crystal formation work is probably maxing out their novelty budget without trying to make the craft partially disassemble.

Plus, they can study the oven after the process which is likely to be helpful if the entire experiment poops the bed.

This is really, really exciting. The moment this spark ignites, it will herald a new reality for industrialization of space, and I for one cannot wait to see it succeed.

One of the things I truly believe will elevate our species is space industrialization. If, in 20 years or so, we send a fleet of space furnaces to 16 Psyche [1], there is a very real possibility that we will be able to move a lot of Earths heavy industrial processes to space. Can you imagine - 3D printed Starship hulls being made from the immense resources of 16 Psyche?

Literal pallets of iPhones being landed from near earth orbit.

It sounds like whacky science fiction now, and for now it really is just that, but the launch and successful mission of Space Forge and other companies like it bring us all a single step closer to seeing that reality play out.

I truly hope we can survive long enough to move heavy metal industry to space, and use that event to return the Earth to a garden state. It’s a long shot, but oh what a beautiful world it would be in 100 years time if this dream can be kept alive, and actually achieved.

[1] - https://en.wikipedia.org/wiki/16_Psyche

Research has gone this direction several times. Every time it did, ground factories discover some way to improve their quality so that space manufacturing became irrelevant again.

That's not to say it will happen again. But it's not a certain thing.

Wins at both ends, whats to not like?

Um... i see some red flags. This test to heat gasses in a furnace looks a little sus. It is basically a micowave oven ... in space!! My kitchen microwave can hit 1000c. Let styropyro at it and it could probably do 10,000c. Then i came across this gem on thier website.

>> Radiators facing cold space can freely produce temperatures near absolute zero for ultra-fast curing without the need for cryogenics.

https://www.spaceforge.com/

I dont see a path to a product. I see a company farming investment and government programs with overhyped "experiments" that have been already done many times. They even talk about testing a heat shield for reentry as if that tech is somehow new. Want to bring samples back? Send your microwave to the space station and bring them back like everyone else.

This is incredible, the huge temperature shifts often create either metal damage or explosions because of instant stretching.

Reply on news.ycombinator.com