The Spruce Pine narrative is a classic case of the internet oversimplifying a complex supply chain into a single point of failure story.
Yes, the quartz there is unusually pure and cheap to refine, and it dominates supply today. But semiconductors aren’t going to vanish overnight if those mines shut down. Synthetic quartz exists, other deposits exist, and the fabs already buffer crucibles.
The problem isn’t “chips depend on one town,” it’s “alternatives cost more and yields suffer.”
What’s actually interesting is that crucibles themselves are a hidden bottleneck. They drive up to a third of the cost of ingot production, and they wear out fast.
If someone develops a longer-lasting crucible material it wouldn’t just de-risk supply, it would lower solar PV costs and boost semiconductor efficiency.
The Spruce Pine hype is fun apocalypse bait, but the real opportunity is crucible innovation.
It’s a mixture. But yeah, today I pushed the boundary further.
A few days ago I flagged a piece someone else had written with ai. It has a specific cadence and some typical patterns. But many people seemed to buy it before I commented.
So I’ve been running a bit of an experiment lately where I write posts and LLM-ify them in varying amounts to see if an LLM actually produce something more upvotable and when people start to notice.
I started out just saying “rephrase this so it sounds tighter” and moved recently towards just jotting rough notes and saying “make an HN comment out of this” and then editing.
Today was the most AI-influenced set of posts, and people clearly noticed, so that’s definitely the threshold. Fascinating.
Check my post history to see the evolution. Using gpt-5.
The part about the crucibles is somewhat interesting, if true, as it reminded me of a passage in the Manhattan Project engineering story that was posted the other day. They thought the melting point of plutonium would be a lot higher than it turned out to be, so they commissioned a lot of exotic crucibles [1] that turned out to be unneeded. They were apparently developed at MIT with cerium sulfide, a compound I'd never heard of before.
A correct summary without anything added is a very valuable thing. I personally don't have the time to read every article, and highly value the scientific paper format, with an abstract given upfront.
A correct summary like this is what any LLM will give you for free so it's not that valuable anymore, but that would be OK if it was advertised as a TL;DR;, and not presented as if it was OP's personal take like it is right now.
I see a few other posts by electric_muse that are LLM-suspicious, but it bugs me that the style is so much like my own writing. Only a matter of time until I'm the witch on trial, I suppose.
I don't really look at the writing style, as it's a very good way to fall for false positive. And I actually won't blame a non native speaker for editing their thoughts using an AI much more proficient in English than them.
Instead I look how much actual information (be it anecdotes, personal opinion, different perspective, etc.) a comment contains.
Here it's just a very straightforward TL;DR; of the original article, it contains no substance beyond that so I don't think anyone would bother write it (or at least they would likely advertise it as a TL;DR;).
See also all the hype about rare earths being needed for electric motors and generators, or cobalt for batteries. Neither of those are really strictly needed. Induction motors can be used instead which don't require any magnets (but those are a few percent less efficient); Lithium Iron Phosphate batteries don't need cobalt (they are slightly less powerful but cheaper and safer).
Most of the automotive industry uses Electrically Excited Synchronous Motors (EESMs) as an alternative to NdFeB magnet motors not induction motors, but they have's taken off yet because investment in mass scale production only began 1-2 years ago.
This will change over the next 2-3 years because BMW, JLR/Tata, Mercedes-Benz, Nissan, and Renault Group have begun moving to EESMs. EU OEMs who have a major foothold in the EESM market like Valeo are lobbying 80% EU indigenization for automotive parts in the EU [0], and India is also started a fairly large industrial subsidy to mass produce EESM powertrains [1] and a number of the players like Sterling Gtake have gotten tech transfers from European vendors like AEM that started off as NatSec funded applications [2]. That said, I also haven't seen a same push for a domestic EESM supply chain in the US unlike the EU and India - GM vendor Niron Magnetics [2] is the only American manufacturer I know trying to manufacture EESMs domestically.
That said, for a number of brownfield military applications, the migration away from NdFeB magnets cannot occur without essentially rendering entire product lines inoperable and unrepairable. This is why NiB magnet chains are viewed as extremely critical.
This topic has come up several times on HN [1][2] and no, the industry does not because there are other source, although more expensive. Other customers had to switch to Russian/Chinese sources due to the disruption but HPQ customers did not.
I figured at the time that other vendors would have a chance to take some HPQ market share but when Spruce Pine went down, Quartz Corp just shifted their refining operations to another plant in Drag, Norway and used existing feedstock and reserves to maintain supply to their customers. Sibelco restarted operations within a few weeks [3].
It was all a non-event in the semiconductor industry, especially compared to real disruptions like the 2011 Tōhoku earthquake that took out a fifth of 300mm wafer supply leading some fabs to shut down entire product lines.
This is similar to the "rare earths" myths floating around. Its a prepper story for sure.
"Rare Earths" are literally everywhere but require large open mines which are not environmentally acceptable and cost effective in the US (depending). The existing mines are still here but shut down decades ago.
Ask Apple how it feels about paying a pretty big chunk of money to reopen one.
That guys mine was worth more closed than open.
Like so many things we sent that task to China a long time ago.
> "Rare Earths" are literally everywhere but require large open mines which are not environmentally acceptable and cost effective in the US (depending). The existing mines are still here but shut down decades ago.
The problem is, "cost effective" has different meanings depending on if one includes externalities such as geopolitical risks into the cost.
Not all rare earths are equally distributed. Lithium sure, but not a number of other REEs like Neodymium, Dysprosium, Molybdenum, and others are only found in unequally distributed concentrated deposits.
It’s interesting that the North Carolina quartz deposits are mined by two foreign interests. That despite the size of the US there was no company that could do it as economically as the Belgians and Norwegians. Presumably due to lack of expertise?
Nope. The American entities used to be independent companies but faced financial troubles when the mining industry died in the US during the 2010s due to a mix of a commodity glut, lack of state support, and competitors like Norway and China infusing state originated capital into their players
The Spruce Pine narrative is a classic case of the internet oversimplifying a complex supply chain into a single point of failure story.
Yes, the quartz there is unusually pure and cheap to refine, and it dominates supply today. But semiconductors aren’t going to vanish overnight if those mines shut down. Synthetic quartz exists, other deposits exist, and the fabs already buffer crucibles.
The problem isn’t “chips depend on one town,” it’s “alternatives cost more and yields suffer.”
What’s actually interesting is that crucibles themselves are a hidden bottleneck. They drive up to a third of the cost of ingot production, and they wear out fast.
If someone develops a longer-lasting crucible material it wouldn’t just de-risk supply, it would lower solar PV costs and boost semiconductor efficiency.
The Spruce Pine hype is fun apocalypse bait, but the real opportunity is crucible innovation.
Did you write this or is it AI? Not hating, but it pings several of my "AI writing" heuristics and I'd like to improve my model if possible.
edit: Never mind, given your comment history this is definitely LLM output.
It’s a mixture. But yeah, today I pushed the boundary further.
A few days ago I flagged a piece someone else had written with ai. It has a specific cadence and some typical patterns. But many people seemed to buy it before I commented.
So I’ve been running a bit of an experiment lately where I write posts and LLM-ify them in varying amounts to see if an LLM actually produce something more upvotable and when people start to notice.
I started out just saying “rephrase this so it sounds tighter” and moved recently towards just jotting rough notes and saying “make an HN comment out of this” and then editing.
Today was the most AI-influenced set of posts, and people clearly noticed, so that’s definitely the threshold. Fascinating.
Check my post history to see the evolution. Using gpt-5.
Is it wrong?
The part about the crucibles is somewhat interesting, if true, as it reminded me of a passage in the Manhattan Project engineering story that was posted the other day. They thought the melting point of plutonium would be a lot higher than it turned out to be, so they commissioned a lot of exotic crucibles [1] that turned out to be unneeded. They were apparently developed at MIT with cerium sulfide, a compound I'd never heard of before.
1: https://www.construction-physics.com/p/an-engineering-histor...
> Is it wrong?
No, but it's a summary of the original article without anything added. I agree with GP that it's very likely LLM generated from the article.
A correct summary without anything added is a very valuable thing. I personally don't have the time to read every article, and highly value the scientific paper format, with an abstract given upfront.
A correct summary like this is what any LLM will give you for free so it's not that valuable anymore, but that would be OK if it was advertised as a TL;DR;, and not presented as if it was OP's personal take like it is right now.
I see a few other posts by electric_muse that are LLM-suspicious, but it bugs me that the style is so much like my own writing. Only a matter of time until I'm the witch on trial, I suppose.
I don't really look at the writing style, as it's a very good way to fall for false positive. And I actually won't blame a non native speaker for editing their thoughts using an AI much more proficient in English than them.
Instead I look how much actual information (be it anecdotes, personal opinion, different perspective, etc.) a comment contains.
Here it's just a very straightforward TL;DR; of the original article, it contains no substance beyond that so I don't think anyone would bother write it (or at least they would likely advertise it as a TL;DR;).
See also all the hype about rare earths being needed for electric motors and generators, or cobalt for batteries. Neither of those are really strictly needed. Induction motors can be used instead which don't require any magnets (but those are a few percent less efficient); Lithium Iron Phosphate batteries don't need cobalt (they are slightly less powerful but cheaper and safer).
> Induction motors can be used instead
Most of the automotive industry uses Electrically Excited Synchronous Motors (EESMs) as an alternative to NdFeB magnet motors not induction motors, but they have's taken off yet because investment in mass scale production only began 1-2 years ago.
This will change over the next 2-3 years because BMW, JLR/Tata, Mercedes-Benz, Nissan, and Renault Group have begun moving to EESMs. EU OEMs who have a major foothold in the EESM market like Valeo are lobbying 80% EU indigenization for automotive parts in the EU [0], and India is also started a fairly large industrial subsidy to mass produce EESM powertrains [1] and a number of the players like Sterling Gtake have gotten tech transfers from European vendors like AEM that started off as NatSec funded applications [2]. That said, I also haven't seen a same push for a domestic EESM supply chain in the US unlike the EU and India - GM vendor Niron Magnetics [2] is the only American manufacturer I know trying to manufacture EESMs domestically.
That said, for a number of brownfield military applications, the migration away from NdFeB magnets cannot occur without essentially rendering entire product lines inoperable and unrepairable. This is why NiB magnet chains are viewed as extremely critical.
[0] - https://www.reuters.com/business/autos-transportation/valeo-...
[1] - https://www.reuters.com/world/china/india-revs-up-alternate-...
[1] - https://www.ncl.ac.uk/business-and-partnerships/case-studies...
[2] - https://www.nironmagnetics.com/
This topic has come up several times on HN [1][2] and no, the industry does not because there are other source, although more expensive. Other customers had to switch to Russian/Chinese sources due to the disruption but HPQ customers did not.
I figured at the time that other vendors would have a chance to take some HPQ market share but when Spruce Pine went down, Quartz Corp just shifted their refining operations to another plant in Drag, Norway and used existing feedstock and reserves to maintain supply to their customers. Sibelco restarted operations within a few weeks [3].
It was all a non-event in the semiconductor industry, especially compared to real disruptions like the 2011 Tōhoku earthquake that took out a fifth of 300mm wafer supply leading some fabs to shut down entire product lines.
[1] https://news.ycombinator.com/item?id=41701862
[2] https://news.ycombinator.com/item?id=39818248
[3] https://www.theverge.com/2024/10/11/24267697/north-carolina-...
This is similar to the "rare earths" myths floating around. Its a prepper story for sure.
"Rare Earths" are literally everywhere but require large open mines which are not environmentally acceptable and cost effective in the US (depending). The existing mines are still here but shut down decades ago.
Ask Apple how it feels about paying a pretty big chunk of money to reopen one. That guys mine was worth more closed than open.
Like so many things we sent that task to China a long time ago.
> "Rare Earths" are literally everywhere but require large open mines which are not environmentally acceptable and cost effective in the US (depending). The existing mines are still here but shut down decades ago.
The problem is, "cost effective" has different meanings depending on if one includes externalities such as geopolitical risks into the cost.
AFAIK the mine is less of a problem than the refinery and its waste.
Not all rare earths are equally distributed. Lithium sure, but not a number of other REEs like Neodymium, Dysprosium, Molybdenum, and others are only found in unequally distributed concentrated deposits.
You might consider looking up what a “rare earth” is. Your list is rather confused.
Fair enough.
Neodymium and Dysprosium are REEs.
Molybdenum is not but it is a critical element that is also commonly brought up in the same conversations as REEs.
I think recentering the convo to "critical minerals" would solve the issue.
Hunterbrook did a great piece about this after Hurricane Helene last year.
https://hntrbrk.com/essential-node-in-global-semiconductor-s...
!
Spruce Pine quartz that doesn’t quite make the purity cut gets used for sandtraps in high-end golf courses.
It’s interesting that the North Carolina quartz deposits are mined by two foreign interests. That despite the size of the US there was no company that could do it as economically as the Belgians and Norwegians. Presumably due to lack of expertise?
Belgium and Norway are firmly in the US' orbit. They're not going anywhere.
... assuming the Trump administration doesn't fuck things up even more than it already did. Europe is fast divesting from the US.
> Presumably due to lack of expertise
Nope. The American entities used to be independent companies but faced financial troubles when the mining industry died in the US during the 2010s due to a mix of a commodity glut, lack of state support, and competitors like Norway and China infusing state originated capital into their players