> Budget brands normally buy older DRAM fabrication equipment from mega-producers like Samsung when Samsung upgrades their DRAM lines to the latest and greatest equipment. This allows the DRAM market to expand more than it would otherwise because it makes any upgrading of the fanciest production lines to still be additive change to the market. However, Korean memory firms have been terrified that reselling old equipment to China-adjacent OEMs might trigger U.S. retaliation…and so those machines have been sitting idle in warehouses since early spring.
This seems to almost be mentioned off-hand, but isn't this a really bad and un-free market, and a much bigger issue? Korean companies are afraid of doing business with Chinese companies because of the US, because of retaliation? This was not the "free and global market" I thought we were supposed to have at this point.
If production lines of DRAM are hindered by the politics of a unrelated 3rd party, then this seems to be a stronger cause of the current shortage than "a very large customer buying a lot in a short period of time".
> This was not the "free and global market" I thought we were supposed to have at this point.
Perhaps you haven't noticed but the pendulum has been swinging the other way for a while already and has a lot of momentum behind it. It's mentioned off-hand because the ongoing return to a multi-polar global order is covered elsewhere already, across dozens of articles every day.
Far from it. Knowledge is power, but too much data can easily obfuscate whatever small signals of knowledge have been captured. Right now, data mostly requires power.
You are correct. But what I mean is that having the capacity for storing unlimited amounts of data allows you to go back in time to reveal something you didn't know that you didn't know or at help you creating complex models able to deal with the noise problem.
When that happens, it becomes knowledge.
An example of that, is that many car insurance companies are known to buy driving data from car dealers and mechanic shops to reveal driving habits of previous drivers.
Admittedly, Europe was being slow to reconfigure until finally forced to do so this year but they've jumped on board now too.
And in typical historical fashion, everybody with less little influence/independence to project their own sphere are now cautiously but attentively jockeying to accumulate the best deals they can gather among those they do.
The world is far from united, even if many do happen to share opinion about the administration.
You're not wrong, globalism and nationalism go hand in hand. All those Twitter bluechecks glazing Trump and trying to drum up the right-wing controversy of the day were exposed as people from third-world countries shitting up the American infosphere for cash. When British far-right nutters say "We can't import the third world", keep in mind that third-world dictators are saying the exact same shit about other parts of the third-world. Nationalists work in lock-step.
Left-wingers are only globalist in the most literal sense of "well, I'd like it if we got rid of these migration barriers". "Internationalist" would be a better term for them.
Globalist Nationalism is an ideology of contradictions. It purports that, no really, you're the real master race and everyone else's just a stooge that'll get taken out the moment we can get rid of these pesky liberals with their freedoms. They need shittons of spatial partitioning to make that work.
You can't get rid of migration barriers without getting rid of the welfare state. Do left-wingers want to get rid of the welfare state? Do they even want to get rid of the migration barriers that can be done away with freely, namely barriers to movement for highly skilled labor and services?
This wasn't what I was talking about, but for what it's worth I don't think welfare and migration are exclusive. If your concern is free-riding, then you can require a certain number of working years before a migrant is eligible for certain social services; or have a scheme in which free migration is required to be reciprocal and bundled with a minimum standard for welfare programs. I believe the EU does both.
The phrase "left-winger" is also kind of vague; lefties in America are very specifically left-libertarians and it was specifically that strain of leftism I was referring to. A lot of, say, European lefties are far less internationalist and far more authoritarian for my tastes. There's an argument in here that American lefties haven't ironed out the contradictions in their political ideologies yet. A lot of the support for free migration from American lefties and liberals only started because Trump took the opposite position and did so extremely. But I'd rather that than, say, mainstream American politics' obsession with endianness[0].
Additionally, the welfare state is currently being crushed under the weight of a bunch of retiring Baby Boomers. Which isn't a uniquely American problem. Every rich country with a welfare state has too many old people in the system. That's due to demographic collapse, which is itself a consequence of children being too expensive to raise in a world where effective birth control is easy to access. Immigration can actually make this less of a problem, by adding to the tax base. They don't even have to be highly-skilled immigrants, they just have to be generating taxable income that goes into the welfare system.
The Globalist Nationalist argument is that this is bad because Those People™ are criminal or uncivilized in some way. But that begs the question: where do we find more taxpayers to keep our welfare state solvent? The answer is to take away birth control and hope enough horny teenagers get pregnant to fix the shortfall eventually. Except this doesn't solve the affordability problem[1]. So all those kids are going to be in broken homes or no home at all. This creates a crime wave. In fact, this is exactly why Those People™ from Those Countries™ are so damned scary!
tl;dr The integrity of the welfare state is only tangentially related to the openness of its migration programs, because migrants can be both beneficiaries of and contributors to the system.
[0] In the original sense of endianness being a culture war over what side to break eggs on.
[1] The affordability problem is itself a cluster of related cost diseases endemic to urban centers. More people want to live in the city than there is housing available for them, so prices have to rise until somebody loses. And increasing the housing supply creates further demand for land, construction labor, and materials. That raises cost, meaning that housing has to get more expensive to make new units economic. And as housing gets more expensive, there's less children being born, which means less people working to build that housing, which makes the housing more expensive, so there's less children being born...
The solution to the affordability problem is trivial: go back to the land. Sure, it would be nice if we could also increase density in urban areas to Hong-Kong-like levels and leverage all the benefits in productivity that would be thereby unlocked, but that's a bit of a separate issue.
In a world of international tensions, governments tend to favor their mega-corps and monopoly. It is a way of weakening your adversaries. It is commercial war.
China did not let US companies establish completely foreigner subsidiaries, yet the US granted it MFN status.
However, the US had protective tariffs since ever too.
Not to mention the controls on migration and remote work (which is a very significant drag on economic growth, as it prevents more efficient allocation of labor).
Err… is there any question that the US is trying to slow down China’s high-tech computer development? I thought that was our open goal.
Countries decided the extent to which they’d like to engage in free trade together. It is a knob that we’d hope our leaders would turn strategically. (Regardless of whether or not we think our leaders actually are doing a good job of it…).
On paper it can sound rational. In reality you look at stuff like cars, for only so long people will tolerate buying a car for $60k when other countries, whom you are also competing with, get buy similar cars for $10-20k from China. Those same vehicles are used to boost productivity in your own domestic industries.
There is always a ton of risk involved with protectionism. Primarily whether your taxpayer-subsidized domestic jobs and hypothetical national security risk significantly outweighs all the very real economic costs.
> buying a car for $60k when other countries, whom you are also competing with, get buy similar cars for $10-20k from China
I'd love to hear your examples of this happening. For $22K you can get a BYD Dolphin Surf in Europe. And that's a pretty small car. What are you paying $60K for in the US that's the same size?
Maybe let's try a different match up. The BYD Atto 3 seems to start around 40K in Europe. It's smaller than a Model Y, and people say it is slightly lower in market position, but close enough. The Model Y starts at around 40K as well.
Are the comparisons between expensive US cars (remember the average is just above 50K, and plenty of perfectly good cars like a Honda Civic can be had for half that) and Chinese cars in China?
I read that's not really the case - there's a bunch of equipment on EU-spec (and some other market) BYDs that comes from EU vendors such as Bosch. It additionally has a completely different AC unit as the kind of refrigerant BYD uses in China is illegal in the EU.
I'm not saying it justifies the price difference, but there are changes between the cars.
Right, so we are never going to see it for 20 grand in the US. Maybe because of tariffs and taxes, as you say, or maybe just because BYD isn't going to set the price at 20K in a market with 10x the average income.
That is a fair point. But then it just reveals that the comparison was contrived from the outset and there was no point to be made. It has never been the case that products in different markets were priced in coordination. The price is always whatever the market will bear, it has zero relationship to the cost to produce unless the market has a lot of competition.
What better way to hurt the designated enemy and make others bare the cost?
Trump's America First in practice relies on a near-sided and overly simplistic understanding of the world (Win-lose, whatever is benefitting others must be a hinderance to the USA). Hence fighting the tariff wars against allies (Canada, Eu). Hence destroying Nato' credibility that was carefully built for 70 years. Hence ceasing to be Ukraine's ally (but continuing to be a trade partner, that sells weapons as long as Europe is paying). Hence helping Putin. Hence instigating problems with Taiwan if that means that TSMC will move some manufacturing to the USA.
It's a really miopic view, but at least on their part the behavior is intentional (consequences, on the other hand, are surprise for them).
Thank you. I was also wondering which one was correct, but I didn't have the opportunity to check it, so I went with my gut feeling. Thank you again for your help.
That’s more or less how a trade war works, yes. Obviously, it’s not just Korean dram manufacturers that have been or will be the only collateral damage.
Is it collateral damage if they get sky-high prices for their products instead of a smaller amount for the production equipment? The US thing seems like a convenient excuse of sorts for cartel behavior.
I don't think you can have a "free and global market" when countries participate in large-scale state industrial policy. Given those constraints, you have to either enforce a zero-subsidy environment (the US has no power to do this) or you have to accept that trade control is one arm of your foreign policy goals and surrendering it entirely is unlikely to help your aims.
For the most part, free and open trade is beneficial to the Western world order. But I think it's quite straightforward to imagine conditions under which it is not, many of which are currently in effect.
US control of EUV technology is probably the most obvious present one, but limitations on nuclear proliferation are an obvious case where there is no free market. Even selling civilian nuclear technology is controlled.
You may think of it analogously to Free Speech. The dream is complete and total expression. The reality is that if you allow convincing enough liars, your society starts to falter. Consequently, certain kinds of expression are not permitted - notably defamation. Think of it as more a North Star navigation ideal constrained by the trade winds (I suppose the Westerlies would be more relevant, but I couldn't resist the pun).
If you want a couple of reads, I enjoyed A Splendid Exchange about the history of trade, which I followed by the resurgent-though-once-dismissed Zeihan's Disunited Nations (which is more a hypothesis book than a history book).
> I don't think you can have a "free and global market" when countries participate in large-scale state industrial policy.
All industrialized countries participate in large-scale state industrial policy. It's a pre-condition of industrialization. A nation cannot industrialize without large scale state policy. And once industrialized, all nations maintain large-scale state industrial policy. Are you saying there never has or can be a "free and global" market? Or just when china does it?
> You may think of it analogously to Free Speech.
It's nothing like free speech as free speech is a constitutional right granted within a nation.
> The reality is that if you allow convincing enough liars, your society starts to falter.
That's rich coming from someone peddling zeihan. I've always wondered what kind of morons actually believe his nonsense. Now I know.
The US industrialized without much in the way of large scale state industrial policy. The federal government was quite weak in the 19th century and, excepting tariffs on British goods, I can't think of any explicit policies it established that were intended to foster industrial capacity. And I think it's debatable how much tariffs actually helped the US develop its manufacturing capacity
> The US industrialized without much in the way of large scale state industrial policy.
What? From funding the Lewis&Clark missions, to forcing japan open, to clearing out the natives for railroad companies, to helping found colleges ( check out many engineering/tech focused colleges like MIT was founded in the 1800s ). You can even argue that american independence and the civil wars were about expanding state industrial policy.
> The federal government was quite weak in the 19th century
So "weak" that we went from 13 small states on the east coast and expanded 3000 miles all the way to the pacific? What the hell are you talking about?
> I can't think of any explicit policies it established that were intended to foster industrial capacity.
The US became the dominant industrial power in the 1800s and you can't think of any policies that helped? You think all the territories in the ohio valley, texas, oklahoma, california, etc chock full of oil were just given to americans by overly generous natives, brits or mexicans? Are you a moron?
If the US didn't have state industrial policy, the US would have never become and industrial power. We'd have just gone down the jeffersonian agrarian paradise road.
Having access to large tracts of land is not a necessary precondition to industrialization (see South Korea). Did the capital accumulation from the exploitation of resources in the American West make it easier to industrialize? Probably. But America would have industrialized if it never expanded beyond the Ohio River valley (access to coal probably was necessary).
Also, as an aside, yes, most of the American West was largely lucked into. America was lucky that France and Spain were dirt broke, that Britain was distracted by continental conflicts with France and Russia, and that native societies had been decimated by disease and a subsequent collapse in governance. That's not to say that there wasn't smart, farsighted leadership in American government, but it was a weak power.
> Having access to large tracts of land is not a necessary precondition to industrialization
But having access to large tracts of land with resources ( like oil ) is. I guess you missed the "chock full of oil" part.
> (see South Korea).
Perfect example. How did Korea industrialize? By being annexed by the japanese ( who went about acquiring tons of land with resources ) and then being annexed by the US with our global network of resource links.
> Also, as an aside, yes, most of the American West was largely lucked into.
It wasn't "lucked into". We won wars against the british and spanish and mexico. And we intimidated the french. And we fought wars against the natives.
> but it was a weak power.
Yes. Because major world powers cede a continent sized piece of territory with an infinite amount of resources to a "weak power".
If the US was a "weak power", then the french, british and russian empires in the 1800s must have been a joke.
We left behind any pretense of a free global market once we entered a post-tariff world. You can't have large universal tariffs or even the threat of them and expect the market to act freely, the two are fundamentally incompatible.
If the free market requires both that companies both ignore that they exist in a world with consequences and that they manage to perfectly predict future demand, that sounds more like an issue with the idea that the free market will solve everything than an issue with the market not being free enough. Otherwise, if you're not happy with the way a company acts, and you don't seem to trust that another company will come and rest their lunch for their perceived poor decisions, your only remaining remedy is to pressure them by non-economic methods to increase production, at which point you don't really believe in the free market either.
> This seems to almost be mentioned off-hand, but isn't this a really bad and un-free market, and a much bigger issue? Korean companies are afraid of doing business with Chinese companies because of the US, because of retaliation? This was not the "free and global market" I thought we were supposed to have at this point.
The globalized free market wasn't all it was cracked up to be, so 90s-era understandings of how things are "supposed" to be will need to be revised.
It's becoming increasingly clear that OpenAI is going to get lapped by Google on technical merits. So this is the "code red" solution? Supply shenanigans?
They are getting beat in the developer market by Anthropic. And getting beat on fundamental tech by Google. This is a company whose ostensible mission is to "benefit all of humanity" ...
This is a desperate move by a company that is in huge trouble.
I paid almost every month since gpt4 came out but mine lapsed when Gemini was released and I haven't even thought of logging in.
The subscribers are exactly the users who would migrate to Gemini. Then your left with the prospect of this giant free chatgpt user base setting money fire.
Wouldn't be shocking at all looking back 10 years from now that maybe the path that Altman stays fired would have been the better path.
Should be, as in, new legislation should criminalize it? What's the generalized principle? Or should be, as in existing law should cover it? And if so, what law / how?
It wouldn't shock me if this is actually just market manipulation. OpenAI in the past year seems to be operating more and more like a pump and dump machine. Their recent AMD deal seems to have been AMD giving them a bunch of stock for free in exchange for them announcing that they would use AMD GPUs for training, and OpenAI doesn't have any fab equipment so the only thing they can do with 40% of the global dram supply is sell it to someone else.
It's market manipulation. You get a chase going (panic buying) by those that are short (need to buy memory in the future). Run up the price on those shorts and squeeze them out as they chase price higher, meanwhile you bought in low and can distribute out your supply when you see fit, or run it up higher until everyone is wrecked. If I were the memory makers I wouldn't want to cede control to openai, you'd rather have a healthy, steady ecosystem than a rigged market that people don't want to be involved in.
Should be investigated as anticompetitive behavior under the FTC Act. Of course that's unlikely to happen. Maybe also market manipulation under the Commodity Exchange Act.
What are the chances the deal doesn’t go through because OpenAI fails to find enough money?
Between Google, other labs and China the risk of commoditization is climbing and so why would investors continue to throw money at them? Kind of the same problem people are starting to bring up regarding Nvidia order book no? Do SoftBank and Oracle have $500B in cash to go through, or does it count on new investors coming in to not implode?
Edit: From the Stargate page on Wikipedia it seems indeed there is a big uncertainty regarding financing:
> On August 7, 2025, Bloomberg reported that the project had not started and no funds were raised to meet the project's initial $500 billion budget. Market uncertainty, American trade policy, and AI hardware valuations caused the delay according to a Bloomberg News report
It's possible this could be construed as price fixing. If the DOJ cared it could open an investigation, leading to a suit, leading to discovery of communications between Altman and all the other relevant players. If price manipulation is an apparent factor in their decision making it may be a very easy case.
Market manipulation is a crime under the Securities Exchange Act of 1934. You can't buy things to influence the price or the market, only to use or resell.
It doesn't really matter, because the first question is: can the government suspend the contract (injunction?) while this is sorted out.
There's also the question of if OpenAI operated In good Faith (from a search: "Another sign of bad faith is withholding crucial information..."), and- of course- the South Korean government can step in as well. In fact- as a worldwide issue- any sufficiently large State(or group of States) can take issue with it.
OpenAI will have issues if they find themselves unable to buy power equipment (Schnider, Eaton). Or, perhaps anybody associated with OpenAI management or funding is arrested the second they step foot in Europe. This is already a nightmare of an International Incident.
It may be a commodity, but there is no established commodity market for RAM in the US, as there is for energy and agricultural products. The laws relate to the manipulation of a commodity market. Commodity markets are usually established where the products are produced, not where they are consumed.
It's interesting that this isn't actually illegal to do except in the specific context of an exchange market. I did a very cursory search of the US Code and the Securities Exchange Act of 1934 and yeah, unless there's some additional legal precedent or other applicable law I didn't find, then this might just be a gap in the law.
It even seems to skirt around notions of illegal vertical integration. For example in this address from 1998, a former FTC commissioner describes several types of illegal "vertical alliances", all of which rely on both the upstream supplier and the downstream consumer being aligned in anticompetitive intent, which (if the article is to be believed) they couldn't have been here because there are two suppliers who were unaware of each other's deals.
Is it really not illegal to just buy up a huge chunk of a critical input for an industry and stockpile it for the purpose of locking out competitors? Seems hard to imagine that some robber baron of the 19th century didn't already do this.
Let's not forget that if it's not illegal now, it could be illegal in a matter of days. Add 12 if a president decides to sit on their thumbs, it's happened before.
Not exactly the 19th century, but in 1980 Softsoap bought up a year's worth of manufacturing capacity of hand pump mechanisms in order to block competitors from entering the consumer market for liquid hand soap.
> Seems hard to imagine that some robber baron of the 19th century didn't already do this.
They tried it and lost their shirts. When you hoard a critical input, that input doesn't just disappear. You spend huge amounts of money pushing the price up, and then the market price turns against you as people finally figure out that the hoarding was just a stunt, that you don't really have a higher-valued use for the stuff you hoarded, and all of it is going to be right back on the market sooner or later.
It's hard to tell from the news but it sounds like they've negotiated contracts for a recurring stream of DRAM rather than just trying to buy a ton of it on an open market. Wouldn't we be stuck with high prices until manufacturing capacity increases and/or the deal expires?
That'd probably make more sense if there wasn't also 50 other tech companies buying up RAM for the same reason (a sudden huge spike in demand due to AI taking off).
They mean to resell them in a different form: as part of their PaaS or SaaS. Per the article, OpenAI is just hoarding the wafers, not purchasing the final product.
It's about volume, not a naive count of consumers. Article claims that OpenAI holds contracts for 40% of world DRAM production. That's just really obviously manipulation if they can't actually power those chips, come on.
So the prosecution will gamble that OpenAI won't in fact use the RAM in a relevant timeframe and they only bought them to exclude the other swath of AI companies from competing?
From the article
> OpenAI isn’t even bothering to buy finished memory modules! No, their deals are unprecedentedly only for raw wafers — uncut, unfinished, and not even allocated to a specific DRAM standard yet. It’s not even clear if they have decided yet on how or when they will finish them into RAM sticks or HBM! Right now it seems like these wafers will just be stockpiled in warehouses – like a kid who hides the toybox because they’re afraid nobody wants to play with them, and thus selfishly feels nobody but them should get the toys!
I guess we'll have to see if they in fact just keep "unfinished" RAM in warehouses like the article says and not roll them out into datacenters for a legitimate use as they are finished.
Last time I used OpenAI computer-use agent (Atlas iirc), I asked it to go on vast.ai and configure an instance template for me. It gave me back control to input my username and password, but alas, since I had used my google account to sign in, I had to go through Google login portal, which put an end to my experiment since it deemed the agent's browser wasn't trustworthy.
As in producers not over-producing RAM should be illegal? A presumably short-term price spike in RAM of all things is a non-issue. It is a luxury good that only a very small number of people care about and there is no reason to think this blip is going to last. Apple did stuff like this all the time at their high point in the late 2000s and early 2010s, and it would happen often in other markets. The world is not static and sometimes the situation changes and lots of supply is soaked up.
> It is a luxury good that only a very small number of people care about
This is an incorrect and incredibly out of touch comment fragment. Computer part derivatives are an essential item to economic activity in most countries.
> It is a luxury good that only a very small number of people care about
The world runs on computers. It is as essential as oil for the functioning of societies. Increase in silicon costs is going to increase costs unilaterally across the board. It happened during the pandemic and something similar will happen now. If anything it should be a wake up call to countries to start thinking about securing their own supply chains.
> Apple did stuff like this all the time at their high point in the late 2000s and early 2010s, and it would happen often in other markets.
Interesting in that I thought about their purchase of $1B of solid state memory at the height of their iPod run. The difference is that Apple had a hit product that was selling as quickly as they could be produced and there was a legitimate need if they wanted to meet the demand.
FTFA:
> No, their deals are unprecedentedly only for raw wafers — uncut, unfinished, and not even allocated to a specific DRAM standard yet. It’s not even clear if they have decided yet on how or when they will finish them into RAM sticks or HBM!
I don't consider this legitimate. It's not illegal, but it sure seems unethical and scummy and it pissed me off. OpenAI throwing its weight around is harming ordinary people who aren't competing with them.
> The difference is that Apple had a hit product that was selling as quickly as they could be produced and there was a legitimate need if they wanted to meet the demand.
What if OpenAI expects to be in the same boat? Their "hit product" is just R&D and training for new very large models. Of course if they're wrong, they've just set a huge pile of their own cash on fire.
If there was a law against buying the supplies of materials and letting them rot in a storehouse just to deprive competitors of them, your argument would be what OpenAI would try to make in court...
Who in the developed world doesn't have a few luxuries? Pretty much all of history people have had to make do with RAM being a lot less accessible than it is now. It isn't essential and people can still buy RAM in the rare situations where they actually need it.
There is nothing here worth invoking the legal system over. OpenAI can buy huge amounts of RAM if they want. Good luck to them, hope it works out, looks like an expensive and risky manoeuvre. And we're probably going to have a RAM glut in a few years looking at these prices.
DRAM is one of the categories of advanced semiconductors that the US considers important enough to national security that exporting it to China is forbidden. It's a fundamental industrial product.
Yeah. Companies like OpenAI need a lot of RAM. That is why they just bought up what is apparently a material chunk of the market.
There is a certain level of crazy that crowds can find when people identify something as a fundamental industrial product critical to national security and simultaneously someone is calling for companies buying a lot of it [0] to be made illegal. If something is critically important to industry then companies should be encouraged to dump as much money as they like in the sector. Otherwise industry will suffer.
[0] And OpenAI is probably going to turn out to be closely associated with US national security too.
They didn't actually buy up the finsished product they actually require. Arguably they raised the price of an input they cannot immediately use hurting themselves by raising the price for what they do actually use in order not to serve a need but to hurt others including the 99.9% of households that use devices with RAM.
It is the malicious purpose and the clear harm to most of America that ought to provide motivation to enforce any law this is at odds with.
Pretending computing is a luxury in 2025 is nonsense as is ignoring the obviously manipulative purpose that is so clear.
> It isn't essential and people can still buy RAM in the rare situations where they actually need it.
There are about five billion smartphone users worldwide. An increase in the price of RAM will, for a start, increase costs for those 5 billion, as smartphones do not last forever.
There's nothing dirty about this deal. When making a large deal with one vendor he didn't disclose to them that he was making a deal with another vendor. That's pretty normal when you're trying to buy a lot of stuff. Otherwise, they can collude to shake you down.
I'm not thrilled about this genre of "guy I don't like does totally normal thing so it's bad". It's too engagement baity.
EDIT: Though even that may be wrong. TechCrunch reports that it was a joint meeting between the South Korean President, the heads of the two companies, and Sam Altman. I won't claim that TC is the bible but there's lots of stuff being reported that makes no sense, and this is a good deal for both these companies so it's more believable than news from someone that OpenAI is going to buy a bunch of wafers and stick it in a warehouse.
It's bad for consumers period. A deal that hampers 40% of global supply shouldn't be a thing, it's predatory. I know DRAM is not a necessity, but considering that PCs are going to be affected means this affects real things like schools and hospitals. There's being smart while making a deal and there's knee-capping the market with your leveraged to the tits business
The line imo is the amount of DRAM OpenAI actually needs/can use. If they end up piling some of it in a warehouse just so nobody else can use it, lock em up.
There's no need to make it illegal: it doesn't happen because the companies with the RAM to sell are motivated by profit. The problem is that OpenAI has plenty of money to buy RAM. They raised $40 billion in April.
> OpenAI isn’t even bothering to buy finished memory modules! No, their deals are unprecedentedly only for raw wafers — uncut, unfinished, and not even allocated to a specific DRAM standard yet. It’s not even clear if they have decided yet on how or when they will finish them into RAM sticks or HBM! Right now it seems like these wafers will just be stockpiled in warehouses
It's not really that different from Apple reserving wafer starts on TSMC's next node and so on. It's just that this kind of capacity requirement has rarely shown up in DRAM before. Vendors prefer this kind of capacity reservation over a more variable finished product requirement. It allows them to know that they can build at the bottleneck rather than having to start up more capacity and then having that lie idle while everything downstream in DRAM packaging and DIMM production can't actually consume anything.
Apple maintains efficiency and single-core performance dominance with the M-series processors by reserving capacity for cutting edge fabrication processes.
They're also reserving capacity on something that doesn't exist yet, and finalize their purchases into finished products. OpenAI is reserving wafers so others can't buy them. They are not the same
I don't buy it. It's easier to make this argument for companies that are building their own hardware, since they know it can be immediately used. OpenAI's move is tantamount to hoarding for the sake of strangling competition. There was plenty of supply to allow for their plans without this move (especially since they will probably go bankrupt at this rate).
If stored in proper conditions, the shelf life of undiced wafers is pretty much limitless. But if moisture or dust get in, it’ll start to have corrosion and other damage. Diced wafers on the other hand are placed on UV tape and I recently learnt that the shelf life of that is like a few months at best.
Nevertheless, unless you’re not storing them in a clean room of appropriate class (or in a vacuum pack), assume 18 months for undiced wafers.
In terms of size, I would think that stored in cassettes or carriers a vacuum sealed, it shouldn’t be too much. A 25 wafer pack is about half meter cube? A 40ft shipping container is about 68m^3. So one container can store about 140 packs of 25 wafers each. That’s 3500 wafers. They’re talking about 900k wafers. That’s about 260 containers. Call it 300 with extra space and stuff. Not exactly hard to provision.
Granting this premise is true (I have no idea), that makes it even worse. They would deliberately be hoarding 40% of the global supply not to lock in for future growth but simply to make sure no one else gets to have it. It’s figuratively setting chips on fire.
It seems not normal (in the sense that it is obviously quite weird to but like half of the world’s RAM supply). But I wonder if they are also just not ready to announce what they are doing with it?
I mean with that many wafers, I guess it is possible that they’d be doing something pretty custom with the things…
> To be clear - the shock wasn’t that OpenAI made a big deal, no, it was that they made two massive deals this big, at the same time, with Samsung and SK Hynix simultaneously
That's not "dirty." That's hiding your intentions from suppliers so they don't crank prices before you walk through their front door.
If you want to buy a cake, never let the baker know it's for a wedding.
What they mean is that they bought 40% of all RAM production, they managed to do that by simultaneously making two big deals at the same time. It's buying up 40% of all RAM production with the intention to have most of it idle in warehouses that is "dirty". And in order to be able to do that, they needed to be secretive and time two big deals at the same time.
The incentive suggested in the article is to block other competitors from scaling training, which is immensely RAM hungry. Amongst other things. Even Nvidia could feel the pressure, since their GPUs need RAM. It could be a good bargaining chip for them, who knows.
I'm not saying it's true, but it is suspicious at the very least. The RAM is unusable as it stands, it's just raw wafer, they'd need a semiconductor fab + PCB assembly to turn them into usable RAM modules. Why does OpenAI want to become a RAM manufacturer, but of only the process post-wafer.
> The RAM is unusable as it stands, it's just raw wafer, they'd need a semiconductor fab
The wafers are processed. That means Samsung/Hynix have taken the raw ("blank") wafers, then run them through their DRAM lithography process, etching hundreds of DRAM dies ("chips") onto the wafer.
You could attach test probes to individual chips on the wafer and you'd have a working DRAM chip. In fact, that's how testing is performed: you connect to each die one at a time with a "probe card" which supplies power & ground, plus an electrical interface for functional testing.
If OpenAI takes possession of the processed wafers and wants finished RAM modules, they need to do a few things: test each die (expensive), saw the wafer into individual chips (cheap), package them (moderately expensive), test them again (medium expense), and then assemble the final module (inexpensive). Modern semiconductor test facilities cost billions of dollars and take years to build, so they'd need to immediately outsource that work (typically done in Southeast Asia)
OpenAI likely doesn't want to do any of this. They probably just want to make sure they're in control of their own destiny with regard to DRAM, then decided the best place to accomplish that was by cutting deals directly with the DRAM semiconductor producers. This will allow them to take the wafers to the existing supply chain, then contract them to turn the wafers into finished modules.
This wasn't buying a cake from a baker, this was a bakery buying 40% of the flour in the world so nobody else can sell wedding cakes, but now there's gonna be no bread for a year
That's not the dirty part. This is the dirty part:
> OpenAI isn’t even bothering to buy finished memory modules! No, their deals are unprecedentedly only for raw wafers — uncut, unfinished, and not even allocated to a specific DRAM standard yet. It’s not even clear if they have decided yet on how or when they will finish them into RAM sticks or HBM! Right now it seems like these wafers will just be stockpiled in warehouses
> OpenAI isn’t even bothering to buy finished memory modules
And? Why should they be obligated to pay for all the middleman steps from fab down to module? That includes: wafer-level test, module-level test (DC, AC, parametric), packaging, post-packaging test, and module fabrication. There's nothing illegal or sketchy about saying, "give me the wafers, I'll take care of everything else myself."
> not even allocated to a specific DRAM standard yet
DRAM manufacturers design and fabricate chips to sell into a standardized, commodity market. There's no secret evolutionary step which occurs after the wafers are etched which turns chips into something which adheres to DDR4,5,6,7,8,9
> It’s not even clear if they have decided yet on how or when they will finish them into RAM sticks or HBM
The implication here is that the primary goal is to corner the market, not to use the supply. If you aren't going to use them anyways then of course it is silly to pay for them to be finished.
Do you think that's fine, or do you think that implication is wrong and OpenAI does actually plan to deploy 40% of the world's DRAM supply?
> The implication here is that the primary goal is to corner the market
You have no evidence of that. Even at face value, the idea of "cornering the market" on a depreciating asset with no long-term value isn't a war strategy, it's flushing money down the toilet. Moreover, there's a credible argument OpenAI wanted to secure capacity in an essential part of their upstream supply chain to ensure stable prices for themselves. That's not "cornering the market," either, it's securing stability for their own growth.
Apple used to buy-up almost all leading-edge semiconductor process capacity from TSMC. It wasn't to resell capacity to everyone else, it was to secure capacity for themselves (particularly for new product launches). Nvidia has been doing the same since the CUDA bubble took off (they have, in effect, two entire fabs worth of leading-edge production just for their GPUs/accelerators). Have they been "cornering" the deep sub-micron foundry market?
> the idea of "cornering the market" on a depreciating asset with no long-term value isn't a war strategy, it's flushing money down the toilet
OpenAI's entire business strategy thus far can be summarized as "flushing money down the toilet", so that isn't actually as unlikely as you're making it sound.
Yes, they've made insane scaling bets before and they have paid off.
If what we've heard about no acceptable pre-training runs from them in the last two years trying to increase the memory for training by two orders of magnitude is just a rehash of what got them from gpt2 to gpt3.
I like how one of the reference links betrays how the article itself was researched, possibly written; HN hides the end of the url, which is "utm_source=chatgpt.com":
The real disappointment is none of the sources are linked in the text. Instead, it's just random underlined words, the classic chatgpt over formatting with lots of extra underlining and bolding.
I appreciate that a 10-15 minute long article summarizes a 25 minute video, but it's hard to hide the real author.
Guys, these are silicon wafers not bars of steel. You can't just stockpile 40% of annual capacity in a warehouse long-term. I would be incredibly surprised to find that any such large scale storage facility exists. Any storage of undiced wafers is temporary while the manufacturing pipeline proceeds. You've seen the pictures of clean rooms and stuff. There's no way you spend all that effort to make the wafer and then just stick it in a warehouse. Who even is going to make such a large cleanroom facility? And for what exactly? It doesn't even pass the basic sniff test.
Much more likely this is just a detail of the contract so that OpenAI can guarantee allocation. I would be surprised if the actual wafers entered OpenAI hands before being fully packaged.
Correct. It's rather routine for wafers manufactured for "last call" orders on ASICs exiting production to be stored as wafers due to not knowing how they need to be Packaged.
> It would require half of one distribution center of a major retailer.
That also meets the specifications of a clean room, and is actively maintained as one?
If OpenAI bought 40% of the annual capacity of finished memory, with the goal of using it in their server farms ASAP, that's one thing.
But unfinished wafers that still need to be protected to finish the manufacturing process, that OpenAI itself does not have any capability to do?
That to me looks like a preemptive strike against competitors, which also affects any other industry that requires RAM, in an attempt to develop a monopolistic position.
I can't see how this isn't a massive national security issue for any country that needs devices requiring RAM for new systems and maintenance of existing ones (pretty much all of them...) to manage critical infrastructure, national defence, public and social services, and so on.
Secondary RAM Manufacturing Had Stalled. Budget brands normally buy older DRAM fabrication equipment from mega-producers like Samsung when Samsung upgrades their DRAM lines to the latest and greatest equipment. This allows the DRAM market to expand more than it would otherwise because it makes any upgrading of the fanciest production lines to still be additive change to the market. However, Korean memory firms have been terrified that reselling old equipment to China-adjacent OEMs might trigger U.S. retaliation…and so those machines have been sitting idle in warehouses since early spring.
My takeaway, this sounds like an comparably easy fix for the consumer market, if prices are somewhat guarenteed to stay mid term significantly above this years spring floor for someone to sweep up the margins and negotiate a somewhat reliable way to get the last gen production lines up and running again.
Will take at least half a year to pick up, but this is not a longterm RAM doomsday scenario in any sense.
I'm more worried about the low to mid-end embedded systems, that a have a dollar budget for memory components, that could get unbearably slow for the current/next gen if manufactures just use the bare minimum of RAM the bloated TV or tablet OS can run on, if the 1GB raspberry move is any indication of that. And consumers stuck with no way to upgrade them to a reasonably usable state.
One of the big problems here is that all of the hardware companies have been burned by hype before (e.g. crypto). No one actually believes that these AI companies will still be around in 5 years so spending billions to build factories for them doesn't make sense.
I don't know what companies will be around 5 years from now, but I would bet there will be more demand for RAM and the price per GB will be at least what it was before this price shock.
RAM is a very cyclical market, historically. You can look at $MU historical charts and kind of see that it trades like a cyclical (compare it to $RIO, for example).
Cyclical companies are easily burned by investing in infrastructure right at the peak. It happens all the time with little mining companies, and I think DRAM manufacturers are sort of the mining companies of tech.
Cyclical markets are the sort of thing 'National Strategic Reserves' should address...
Am I crazy for wanting this to be in Full ECC RAM modules suitable for composition into many device factors with hope that we'll finally go to reliable memory for all markets as a result?
I'm curious how OpenAI has the funds to pay for 40% of the worlds ram production? Sure they are big and have a few billions but I kind of assumed that 40% for a year or whatever they are buying is easily double digit billions? That has to hurt even them, especially because they cant buy anything else?
Also what are these contracts? Surely Samsung could decide to cancel the contract by paying a large fee but is that fee truly so large that getting their ram back when prices are now 4x of what they used to be is not worth it?
I assume this includes more than just the raw price of modules but Openai only has 60 billion in funding altogether and was aiming for 20 billion ARR this year. This sounds like they are spending maybe half their money on RAM they never use? That just doesn't add up.
I think the event is big enough to stop them and send them behind bars.
> Samsung
I think that Samsung -and other manufacturers- have been intentionally limiting their production capacities so as not to devalue the prices of their chips (for SSD at least) so may be they are an interested part. This, combined with the madness we are seeing, is abuse^2 . I think they should also end up behind bars.
I remember it didn’t work out well for Randolph and Mortimer. Sam may pull it out, though, if he just sells the DRAM now while the market is still hot.
This is missing a key part of the picture - Nvidia just announced that partners will need to source RAM themselves.
OpenAI is basically ensuring that they can actually get the chips they need for the DCs they are building.
I can’t guess as to what move came first (Nvidia policy change or these DRAM deals) but I would bet this is a large if not larger factor here than “bloc my competitors.
Yes, I have an older gaming PC from ~2018 that I keep putting off upgrading (first GPU prices skyrocketed, now this...) and was hoping to replace it with a Steam Machine next year. Will be endlessly bummed if that doesn't happen.
Will also be interesting if Sony/Microsoft was planning on releasing a next-gen system anytime soon, and I wonder if this will affect Apple's hardware at all.
> Lenovo has begun notifying clients of coming price hikes, with adjustments set to take effect in early 2026.. Dell is expected to raise prices by at least 15-20%, with the increase potentially taking effect as soon as mid-December.. Dell COO Jeff Clarke warned that he’s “never seen memory-chip costs rise this fast,” .. Lenovo [cited] two key factors: an intensifying memory shortage and the rapid integration of AI technologies.. TrendForce has downgraded its 2026 notebook shipment forecast from an initial 1.7% YoY growth to a 2.4% YoY decline.
> your business model might end up being sort of a … startup incubator or private equity firm; you’d spend your time starting or acquiring companies on which the robot could work its magic. Your business model would be “general business, but with AI.” .. OpenAI has a $500 billion valuation largely as a bet that a lot of the value of AI will accrue to its builders, but it could hedge that bet by owning the users too. Either it will sell AI at high margins to lots of businesses, or it will sell AI at lower margins to lucrative businesses that it owns.
They definitely can, as anybody could, for a short enough period of time. The only ones betting they'll be able to sustain it for a long period are the ones paying the currently very inflated price for ram.
but it's not a short term deal. funny enough, there's a cohort of bonds that never even made one payment. They are called first payment default bonds. There is a cheekier nickname for them, but it escapes me at the moment.
With the amount of hatred for anything OpenAI it’s not surprising the author chose a clickbaity title. HNs quality of posts are going down and instead of objective analysis I often see very polarized and flamy articles, titles etc.
It to the point that bloggers know they can say things like "OpenAI did 9/11!", provide no sources, and it still will make the front page and have a large increase in traffic to there site.
Strange how LLM vendors are flooding the market with reasons not to do business with them. Every paid agentic interaction contributes to all the bad behavior we are seeing. From out of control web scraping to buying up available hardware LLMs are turning out to be highly efficient misery manufacturing mechanisms.
What happens to enterprise apps like SAP and S4 on the cloud that swallow huge amounts of RAM for their in-memory database? AWS, GCP, and Azure must be running around? With their predatory pricing customers were already hesitant about upgrades, it now gets more difficult. Extend EOL for ECC?
Is it possible that GPT-6 could possibly be highly reliant on RAM? For example running smaller long context models on the CPU? Just a thought, I could be far off.
RAM being super expensive right now is a hot topic, and I imagine people are curious about why, so anything that purports to explain that is somewhat likely to at least get a cursory look from people. Without taking a stance on either whether this is happening here, if enough people look at something and happen to think it makes sense, a lack of evidence might not be enough to prevent the narrative from taking over; otherwise, we'd never have issues with people believing false things in the first place, and that happens all the time.
I mean, say you have something in stock that others sell for twice as much as last week and it will cost you twice as much to restock, why would you sell it for last weeks prices?
Why would it matter if you have anything in stock or not? What does it matter what you've paid for it?
If it will be hard to restock and ill most likely will have to sell "no" ill be even more motivated to ask more for it.
almost feel like OpenAI's recent "fall" is a decoy setup by them intentionally.. something's cooking.. maybe they wanted to buy back their own shares at a lower price?
And all these data centers they want to build around the country. When consumers can’t get devices they want maybe they’ll fight even harder against these data centers being built in their back yard. He’s not making any fans with this move that’s for sure
OP is also marginally underestimating the impact this move would have on Google's competitiveness - they are making huge gains prototyping at light speed; this will halt their AI hardware acceleration plans pushing them back into slower software development on ever aging hardware.
It also shows why Nvidia is not afraid of competitors coming out with new desgings that obsolete their hardware: what good are superior designs with no fabs to produce them?
every one of these things that make the deal "good" for OpenAI is a direct result of negative externalities for everyone else: competitors, consumers, and people who wouldn't care otherwise.
The article even says that they don't have an obvious plan for how to use the wafers they bought, and very clearly suggests that this is purely an anticompetitive tactic to force everyone else to eat a price increase that OpenAI doesn't need to face. It's clever though because if any regulatory agency starts asking questions (not that they would do that in the current USA political climate) then OpenAI can just say it's a strategic reserve, we have plans to do something with it, etc. etc. What are you going to do? Take them to court and force them to auction off some % of the stock? Set an industry-wide limit on wafer inventory? Fine them? You'd need to find some evidence that it was done maliciously, and good luck with that.
There are some negative elements of captialism that we might simply have no reasonable regulatory apparatus to deal with. Preventing indivduals and companies from having so much market power in the first place seems to be the only thing that can work consistently.
As an AI researcher, I thought it was relatively well established (at least among my colleagues) that being pro-AI actually meant you were anti-Sam as well. He's the worst actor in the industry and has done an incredible amount of damage to its brand.
> Budget brands normally buy older DRAM fabrication equipment from mega-producers like Samsung when Samsung upgrades their DRAM lines to the latest and greatest equipment. This allows the DRAM market to expand more than it would otherwise because it makes any upgrading of the fanciest production lines to still be additive change to the market. However, Korean memory firms have been terrified that reselling old equipment to China-adjacent OEMs might trigger U.S. retaliation…and so those machines have been sitting idle in warehouses since early spring.
This seems to almost be mentioned off-hand, but isn't this a really bad and un-free market, and a much bigger issue? Korean companies are afraid of doing business with Chinese companies because of the US, because of retaliation? This was not the "free and global market" I thought we were supposed to have at this point.
If production lines of DRAM are hindered by the politics of a unrelated 3rd party, then this seems to be a stronger cause of the current shortage than "a very large customer buying a lot in a short period of time".
> This was not the "free and global market" I thought we were supposed to have at this point.
Perhaps you haven't noticed but the pendulum has been swinging the other way for a while already and has a lot of momentum behind it. It's mentioned off-hand because the ongoing return to a multi-polar global order is covered elsewhere already, across dozens of articles every day.
Although it is unknown what will happen with current generation of AI system, data is power.
We have been slowly moving towards centralized computing again since 2010s
With governments backing this trend of AI with military funds it is only natural for the others to follow and making computing even more centralized.
By 2020s, Nvidia announce the plan to become more b2b only and then at some point a cloud only company, not even selling the assets.
I don't mean you won't have your smartphone, but it is a dumb terminal compared to the capacities you'll have and maybe need daily.
data is power
Far from it. Knowledge is power, but too much data can easily obfuscate whatever small signals of knowledge have been captured. Right now, data mostly requires power.
You are correct. But what I mean is that having the capacity for storing unlimited amounts of data allows you to go back in time to reveal something you didn't know that you didn't know or at help you creating complex models able to deal with the noise problem.
When that happens, it becomes knowledge.
An example of that, is that many car insurance companies are known to buy driving data from car dealers and mechanic shops to reveal driving habits of previous drivers.
[dead]
> has a lot of momentum behind it
Does it? It's mostly the US pushing against globalism right now. Trump managed to unite almost the entire world.
Britain? Russia? China?
Admittedly, Europe was being slow to reconfigure until finally forced to do so this year but they've jumped on board now too.
And in typical historical fashion, everybody with less little influence/independence to project their own sphere are now cautiously but attentively jockeying to accumulate the best deals they can gather among those they do.
The world is far from united, even if many do happen to share opinion about the administration.
So... How are China and the UK working against globalism exactly?
Is it? I'm fairly certain many countries have had a very strong showing from nationalist parties in recent years, or am I way off base?
You're not wrong, globalism and nationalism go hand in hand. All those Twitter bluechecks glazing Trump and trying to drum up the right-wing controversy of the day were exposed as people from third-world countries shitting up the American infosphere for cash. When British far-right nutters say "We can't import the third world", keep in mind that third-world dictators are saying the exact same shit about other parts of the third-world. Nationalists work in lock-step.
Left-wingers are only globalist in the most literal sense of "well, I'd like it if we got rid of these migration barriers". "Internationalist" would be a better term for them.
Globalist Nationalism is an ideology of contradictions. It purports that, no really, you're the real master race and everyone else's just a stooge that'll get taken out the moment we can get rid of these pesky liberals with their freedoms. They need shittons of spatial partitioning to make that work.
You can't get rid of migration barriers without getting rid of the welfare state. Do left-wingers want to get rid of the welfare state? Do they even want to get rid of the migration barriers that can be done away with freely, namely barriers to movement for highly skilled labor and services?
This wasn't what I was talking about, but for what it's worth I don't think welfare and migration are exclusive. If your concern is free-riding, then you can require a certain number of working years before a migrant is eligible for certain social services; or have a scheme in which free migration is required to be reciprocal and bundled with a minimum standard for welfare programs. I believe the EU does both.
The phrase "left-winger" is also kind of vague; lefties in America are very specifically left-libertarians and it was specifically that strain of leftism I was referring to. A lot of, say, European lefties are far less internationalist and far more authoritarian for my tastes. There's an argument in here that American lefties haven't ironed out the contradictions in their political ideologies yet. A lot of the support for free migration from American lefties and liberals only started because Trump took the opposite position and did so extremely. But I'd rather that than, say, mainstream American politics' obsession with endianness[0].
Additionally, the welfare state is currently being crushed under the weight of a bunch of retiring Baby Boomers. Which isn't a uniquely American problem. Every rich country with a welfare state has too many old people in the system. That's due to demographic collapse, which is itself a consequence of children being too expensive to raise in a world where effective birth control is easy to access. Immigration can actually make this less of a problem, by adding to the tax base. They don't even have to be highly-skilled immigrants, they just have to be generating taxable income that goes into the welfare system.
The Globalist Nationalist argument is that this is bad because Those People™ are criminal or uncivilized in some way. But that begs the question: where do we find more taxpayers to keep our welfare state solvent? The answer is to take away birth control and hope enough horny teenagers get pregnant to fix the shortfall eventually. Except this doesn't solve the affordability problem[1]. So all those kids are going to be in broken homes or no home at all. This creates a crime wave. In fact, this is exactly why Those People™ from Those Countries™ are so damned scary!
tl;dr The integrity of the welfare state is only tangentially related to the openness of its migration programs, because migrants can be both beneficiaries of and contributors to the system.
[0] In the original sense of endianness being a culture war over what side to break eggs on.
[1] The affordability problem is itself a cluster of related cost diseases endemic to urban centers. More people want to live in the city than there is housing available for them, so prices have to rise until somebody loses. And increasing the housing supply creates further demand for land, construction labor, and materials. That raises cost, meaning that housing has to get more expensive to make new units economic. And as housing gets more expensive, there's less children being born, which means less people working to build that housing, which makes the housing more expensive, so there's less children being born...
The solution to the affordability problem is trivial: go back to the land. Sure, it would be nice if we could also increase density in urban areas to Hong-Kong-like levels and leverage all the benefits in productivity that would be thereby unlocked, but that's a bit of a separate issue.
In a world of international tensions, governments tend to favor their mega-corps and monopoly. It is a way of weakening your adversaries. It is commercial war.
"It's mostly the US pushing against globalism right now."
This is completely wrong. China and Russia are very much working against it.
I'm not sure any realistic definition of free market would mean your actions are free from consequence.
The global market is anarchy in the literal sense and no one is bound by a higher authority. Coercion and cartels are part of a free market.
Economic efficiency actually requires a lot of rules and regulations to achieve the free market playground we like to imagine.
David Graeber in three lines
Cartels aren't part of a free market, wtf
Nah, it is simply just not free market. USA is actively and systematically acting against free market .
Well, when was it last free? A few millena ago?
China did not let US companies establish completely foreigner subsidiaries, yet the US granted it MFN status.
However, the US had protective tariffs since ever too.
Not to mention the controls on migration and remote work (which is a very significant drag on economic growth, as it prevents more efficient allocation of labor).
Err… is there any question that the US is trying to slow down China’s high-tech computer development? I thought that was our open goal.
Countries decided the extent to which they’d like to engage in free trade together. It is a knob that we’d hope our leaders would turn strategically. (Regardless of whether or not we think our leaders actually are doing a good job of it…).
> I thought that was our open goal.
Is the goal also to hurt South Korean businesses and all businesses in the world, just to "pwn China" basically?
We’re probably also spurring China to develop more independently. I don’t think it is a good plan, just an unconfusing one.
On paper it can sound rational. In reality you look at stuff like cars, for only so long people will tolerate buying a car for $60k when other countries, whom you are also competing with, get buy similar cars for $10-20k from China. Those same vehicles are used to boost productivity in your own domestic industries.
There is always a ton of risk involved with protectionism. Primarily whether your taxpayer-subsidized domestic jobs and hypothetical national security risk significantly outweighs all the very real economic costs.
> buying a car for $60k when other countries, whom you are also competing with, get buy similar cars for $10-20k from China
I'd love to hear your examples of this happening. For $22K you can get a BYD Dolphin Surf in Europe. And that's a pretty small car. What are you paying $60K for in the US that's the same size?
Maybe let's try a different match up. The BYD Atto 3 seems to start around 40K in Europe. It's smaller than a Model Y, and people say it is slightly lower in market position, but close enough. The Model Y starts at around 40K as well.
Are the comparisons between expensive US cars (remember the average is just above 50K, and plenty of perfectly good cars like a Honda Civic can be had for half that) and Chinese cars in China?
Atto 3 is $19.3k USD in Thailand.
So it's really just tariffs and taxes making it that expensive elsewhere.
I read that's not really the case - there's a bunch of equipment on EU-spec (and some other market) BYDs that comes from EU vendors such as Bosch. It additionally has a completely different AC unit as the kind of refrigerant BYD uses in China is illegal in the EU.
I'm not saying it justifies the price difference, but there are changes between the cars.
The BYDs exported to Europe are made in Thailand.
So no, they are the Euro spec car.
Right, so we are never going to see it for 20 grand in the US. Maybe because of tariffs and taxes, as you say, or maybe just because BYD isn't going to set the price at 20K in a market with 10x the average income.
> Right, so we are never going to see it for 20 grand in the US
To be fair, @dmix explicitly mentioned the $20k price was for other countries
That is a fair point. But then it just reveals that the comparison was contrived from the outset and there was no point to be made. It has never been the case that products in different markets were priced in coordination. The price is always whatever the market will bear, it has zero relationship to the cost to produce unless the market has a lot of competition.
What better way to hurt the designated enemy and make others bare the cost?
Trump's America First in practice relies on a near-sided and overly simplistic understanding of the world (Win-lose, whatever is benefitting others must be a hinderance to the USA). Hence fighting the tariff wars against allies (Canada, Eu). Hence destroying Nato' credibility that was carefully built for 70 years. Hence ceasing to be Ukraine's ally (but continuing to be a trade partner, that sells weapons as long as Europe is paying). Hence helping Putin. Hence instigating problems with Taiwan if that means that TSMC will move some manufacturing to the USA.
It's a really miopic view, but at least on their part the behavior is intentional (consequences, on the other hand, are surprise for them).
« to bear the cost », not « to bare the cost »
Thank you. I was also wondering which one was correct, but I didn't have the opportunity to check it, so I went with my gut feeling. Thank you again for your help.
That’s more or less how a trade war works, yes. Obviously, it’s not just Korean dram manufacturers that have been or will be the only collateral damage.
Is it collateral damage if they get sky-high prices for their products instead of a smaller amount for the production equipment? The US thing seems like a convenient excuse of sorts for cartel behavior.
"America First" as an ideology means that question is never considered
I don't think you can have a "free and global market" when countries participate in large-scale state industrial policy. Given those constraints, you have to either enforce a zero-subsidy environment (the US has no power to do this) or you have to accept that trade control is one arm of your foreign policy goals and surrendering it entirely is unlikely to help your aims.
For the most part, free and open trade is beneficial to the Western world order. But I think it's quite straightforward to imagine conditions under which it is not, many of which are currently in effect.
US control of EUV technology is probably the most obvious present one, but limitations on nuclear proliferation are an obvious case where there is no free market. Even selling civilian nuclear technology is controlled.
You may think of it analogously to Free Speech. The dream is complete and total expression. The reality is that if you allow convincing enough liars, your society starts to falter. Consequently, certain kinds of expression are not permitted - notably defamation. Think of it as more a North Star navigation ideal constrained by the trade winds (I suppose the Westerlies would be more relevant, but I couldn't resist the pun).
If you want a couple of reads, I enjoyed A Splendid Exchange about the history of trade, which I followed by the resurgent-though-once-dismissed Zeihan's Disunited Nations (which is more a hypothesis book than a history book).
> I don't think you can have a "free and global market" when countries participate in large-scale state industrial policy.
All industrialized countries participate in large-scale state industrial policy. It's a pre-condition of industrialization. A nation cannot industrialize without large scale state policy. And once industrialized, all nations maintain large-scale state industrial policy. Are you saying there never has or can be a "free and global" market? Or just when china does it?
> You may think of it analogously to Free Speech.
It's nothing like free speech as free speech is a constitutional right granted within a nation.
> The reality is that if you allow convincing enough liars, your society starts to falter.
That's rich coming from someone peddling zeihan. I've always wondered what kind of morons actually believe his nonsense. Now I know.
The US industrialized without much in the way of large scale state industrial policy. The federal government was quite weak in the 19th century and, excepting tariffs on British goods, I can't think of any explicit policies it established that were intended to foster industrial capacity. And I think it's debatable how much tariffs actually helped the US develop its manufacturing capacity
> The US industrialized without much in the way of large scale state industrial policy.
What? From funding the Lewis&Clark missions, to forcing japan open, to clearing out the natives for railroad companies, to helping found colleges ( check out many engineering/tech focused colleges like MIT was founded in the 1800s ). You can even argue that american independence and the civil wars were about expanding state industrial policy.
> The federal government was quite weak in the 19th century
So "weak" that we went from 13 small states on the east coast and expanded 3000 miles all the way to the pacific? What the hell are you talking about?
> I can't think of any explicit policies it established that were intended to foster industrial capacity.
The US became the dominant industrial power in the 1800s and you can't think of any policies that helped? You think all the territories in the ohio valley, texas, oklahoma, california, etc chock full of oil were just given to americans by overly generous natives, brits or mexicans? Are you a moron?
If the US didn't have state industrial policy, the US would have never become and industrial power. We'd have just gone down the jeffersonian agrarian paradise road.
Having access to large tracts of land is not a necessary precondition to industrialization (see South Korea). Did the capital accumulation from the exploitation of resources in the American West make it easier to industrialize? Probably. But America would have industrialized if it never expanded beyond the Ohio River valley (access to coal probably was necessary).
Also, as an aside, yes, most of the American West was largely lucked into. America was lucky that France and Spain were dirt broke, that Britain was distracted by continental conflicts with France and Russia, and that native societies had been decimated by disease and a subsequent collapse in governance. That's not to say that there wasn't smart, farsighted leadership in American government, but it was a weak power.
> Having access to large tracts of land is not a necessary precondition to industrialization
But having access to large tracts of land with resources ( like oil ) is. I guess you missed the "chock full of oil" part.
> (see South Korea).
Perfect example. How did Korea industrialize? By being annexed by the japanese ( who went about acquiring tons of land with resources ) and then being annexed by the US with our global network of resource links.
> Also, as an aside, yes, most of the American West was largely lucked into.
It wasn't "lucked into". We won wars against the british and spanish and mexico. And we intimidated the french. And we fought wars against the natives.
> but it was a weak power.
Yes. Because major world powers cede a continent sized piece of territory with an infinite amount of resources to a "weak power".
If the US was a "weak power", then the french, british and russian empires in the 1800s must have been a joke.
Zeihan is wonderful if read as a post-modern, historical fiction novelist.
The characters and plots are obviously all exaggerations and fabrications of reality but there are many interesting themes in his novels.
Of course, reading him like he is Edward Gibbon would be pretty foolish but The End of the World is just the Beginning is a great book.
> Think of it as
did AI write this?
Boy, am I glad I wrote a killfile for Hacker News.
We left behind any pretense of a free global market once we entered a post-tariff world. You can't have large universal tariffs or even the threat of them and expect the market to act freely, the two are fundamentally incompatible.
Oh, the market will find a way around this too. The more US uses this particular button the less effective it becomes.
If the free market requires both that companies both ignore that they exist in a world with consequences and that they manage to perfectly predict future demand, that sounds more like an issue with the idea that the free market will solve everything than an issue with the market not being free enough. Otherwise, if you're not happy with the way a company acts, and you don't seem to trust that another company will come and rest their lunch for their perceived poor decisions, your only remaining remedy is to pressure them by non-economic methods to increase production, at which point you don't really believe in the free market either.
What free market?
> This seems to almost be mentioned off-hand, but isn't this a really bad and un-free market, and a much bigger issue? Korean companies are afraid of doing business with Chinese companies because of the US, because of retaliation? This was not the "free and global market" I thought we were supposed to have at this point.
The globalized free market wasn't all it was cracked up to be, so 90s-era understandings of how things are "supposed" to be will need to be revised.
Strategically, US+EU should buy them. Will they? Probably too stupid/clumsy to figure it out.
Moves like this should be illegal.
It's becoming increasingly clear that OpenAI is going to get lapped by Google on technical merits. So this is the "code red" solution? Supply shenanigans?
They are getting beat in the developer market by Anthropic. And getting beat on fundamental tech by Google. This is a company whose ostensible mission is to "benefit all of humanity" ...
This is a desperate move by a company that is in huge trouble.
I paid almost every month since gpt4 came out but mine lapsed when Gemini was released and I haven't even thought of logging in.
The subscribers are exactly the users who would migrate to Gemini. Then your left with the prospect of this giant free chatgpt user base setting money fire.
Wouldn't be shocking at all looking back 10 years from now that maybe the path that Altman stays fired would have been the better path.
> Moves like this should be illegal.
Should be, as in, new legislation should criminalize it? What's the generalized principle? Or should be, as in existing law should cover it? And if so, what law / how?
It wouldn't shock me if this is actually just market manipulation. OpenAI in the past year seems to be operating more and more like a pump and dump machine. Their recent AMD deal seems to have been AMD giving them a bunch of stock for free in exchange for them announcing that they would use AMD GPUs for training, and OpenAI doesn't have any fab equipment so the only thing they can do with 40% of the global dram supply is sell it to someone else.
It's market manipulation. You get a chase going (panic buying) by those that are short (need to buy memory in the future). Run up the price on those shorts and squeeze them out as they chase price higher, meanwhile you bought in low and can distribute out your supply when you see fit, or run it up higher until everyone is wrecked. If I were the memory makers I wouldn't want to cede control to openai, you'd rather have a healthy, steady ecosystem than a rigged market that people don't want to be involved in.
the only thing they can do with 40% of the global dram supply is sell it to someone else.
The way it works is that OpenAI will have the DRAM delivered to Nvidia/AMD/Broadcom to be assembled into the racks that OpenAI buys.
Should be investigated as anticompetitive behavior under the FTC Act. Of course that's unlikely to happen. Maybe also market manipulation under the Commodity Exchange Act.
> Should be, as in, new legislation should criminalize it? What's the generalized principle?
RAM manufacturers are a de facto duopoly manipulating the market to gouge prices. It's not that difficult to see.
What are the chances the deal doesn’t go through because OpenAI fails to find enough money?
Between Google, other labs and China the risk of commoditization is climbing and so why would investors continue to throw money at them? Kind of the same problem people are starting to bring up regarding Nvidia order book no? Do SoftBank and Oracle have $500B in cash to go through, or does it count on new investors coming in to not implode?
Edit: From the Stargate page on Wikipedia it seems indeed there is a big uncertainty regarding financing:
> On August 7, 2025, Bloomberg reported that the project had not started and no funds were raised to meet the project's initial $500 billion budget. Market uncertainty, American trade policy, and AI hardware valuations caused the delay according to a Bloomberg News report
Seems like it is, but the question is whether the current Justice Department will do anything about it.
I read "US Justice Department" the same way I read "Britain's Ministry of Truth".
When I hear about this US justice department, I hear the mafia enforcement.
It has been since forever. There was a reason J. Edgar Hoover denied the existence of organized crime for decades.
> Seems like it is
Do you have a citation for what law is being violated? Or just vibes?
It's possible this could be construed as price fixing. If the DOJ cared it could open an investigation, leading to a suit, leading to discovery of communications between Altman and all the other relevant players. If price manipulation is an apparent factor in their decision making it may be a very easy case.
https://fortune.com/2025/11/23/ai-rivals-like-openai-nvidia-...
Obviously kind of a moot point because whether it violates antitrust law or not, what is guaranteed is the US Government is not going to do anything
Whether the Clayton and Sherman acts apply to the Stargate initiative is not relevant to this RAM-hoarding activity.
Market manipulation is a crime under the Securities Exchange Act of 1934. You can't buy things to influence the price or the market, only to use or resell.
https://en.wikipedia.org/wiki/Market_manipulation
The law you refer to applies only to markets for securities. RAM is very clearly not a security, it fails the Howey test.
There are similar laws prohibiting the manipulation of commodity markets but I do not believe a US court would find RAM to be a commodity.
How is RAM not a commodity?
It doesn't really matter, because the first question is: can the government suspend the contract (injunction?) while this is sorted out.
There's also the question of if OpenAI operated In good Faith (from a search: "Another sign of bad faith is withholding crucial information..."), and- of course- the South Korean government can step in as well. In fact- as a worldwide issue- any sufficiently large State(or group of States) can take issue with it.
OpenAI will have issues if they find themselves unable to buy power equipment (Schnider, Eaton). Or, perhaps anybody associated with OpenAI management or funding is arrested the second they step foot in Europe. This is already a nightmare of an International Incident.
It may be a commodity, but there is no established commodity market for RAM in the US, as there is for energy and agricultural products. The laws relate to the manipulation of a commodity market. Commodity markets are usually established where the products are produced, not where they are consumed.
Do you think OpenAI plans to trade the semiconductor market? This would only apply in that scenario.
It's interesting that this isn't actually illegal to do except in the specific context of an exchange market. I did a very cursory search of the US Code and the Securities Exchange Act of 1934 and yeah, unless there's some additional legal precedent or other applicable law I didn't find, then this might just be a gap in the law.
It even seems to skirt around notions of illegal vertical integration. For example in this address from 1998, a former FTC commissioner describes several types of illegal "vertical alliances", all of which rely on both the upstream supplier and the downstream consumer being aligned in anticompetitive intent, which (if the article is to be believed) they couldn't have been here because there are two suppliers who were unaware of each other's deals.
Is it really not illegal to just buy up a huge chunk of a critical input for an industry and stockpile it for the purpose of locking out competitors? Seems hard to imagine that some robber baron of the 19th century didn't already do this.
Let's not forget that if it's not illegal now, it could be illegal in a matter of days. Add 12 if a president decides to sit on their thumbs, it's happened before.
Oops I forgot the link to the FTC document https://www.ftc.gov/news-events/news/speeches/vertical-issue...
and here's a followup from the same commissioner https://www.ftc.gov/news-events/news/speeches/vertical-issue...
Not exactly the 19th century, but in 1980 Softsoap bought up a year's worth of manufacturing capacity of hand pump mechanisms in order to block competitors from entering the consumer market for liquid hand soap.
I didn't know about this history. Looks like it actually worked out great for them.
> Seems hard to imagine that some robber baron of the 19th century didn't already do this.
They tried it and lost their shirts. When you hoard a critical input, that input doesn't just disappear. You spend huge amounts of money pushing the price up, and then the market price turns against you as people finally figure out that the hoarding was just a stunt, that you don't really have a higher-valued use for the stuff you hoarded, and all of it is going to be right back on the market sooner or later.
It's hard to tell from the news but it sounds like they've negotiated contracts for a recurring stream of DRAM rather than just trying to buy a ton of it on an open market. Wouldn't we be stuck with high prices until manufacturing capacity increases and/or the deal expires?
No, they want DRAM to be expensive to give them a competitive advantage over their competitors.
That'd probably make more sense if there wasn't also 50 other tech companies buying up RAM for the same reason (a sudden huge spike in demand due to AI taking off).
They mean to resell them in a different form: as part of their PaaS or SaaS. Per the article, OpenAI is just hoarding the wafers, not purchasing the final product.
It's about volume, not a naive count of consumers. Article claims that OpenAI holds contracts for 40% of world DRAM production. That's just really obviously manipulation if they can't actually power those chips, come on.
So the prosecution will gamble that OpenAI won't in fact use the RAM in a relevant timeframe and they only bought them to exclude the other swath of AI companies from competing?
From the article
> OpenAI isn’t even bothering to buy finished memory modules! No, their deals are unprecedentedly only for raw wafers — uncut, unfinished, and not even allocated to a specific DRAM standard yet. It’s not even clear if they have decided yet on how or when they will finish them into RAM sticks or HBM! Right now it seems like these wafers will just be stockpiled in warehouses – like a kid who hides the toybox because they’re afraid nobody wants to play with them, and thus selfishly feels nobody but them should get the toys!
I guess we'll have to see if they in fact just keep "unfinished" RAM in warehouses like the article says and not roll them out into datacenters for a legitimate use as they are finished.
[dead]
[dead]
The current Justice Department? You're kidding, right?
Google just need to give the US regime a solid gold award
for what? unimportant
Last time I used OpenAI computer-use agent (Atlas iirc), I asked it to go on vast.ai and configure an instance template for me. It gave me back control to input my username and password, but alas, since I had used my google account to sign in, I had to go through Google login portal, which put an end to my experiment since it deemed the agent's browser wasn't trustworthy.
Mmmh.
As in producers not over-producing RAM should be illegal? A presumably short-term price spike in RAM of all things is a non-issue. It is a luxury good that only a very small number of people care about and there is no reason to think this blip is going to last. Apple did stuff like this all the time at their high point in the late 2000s and early 2010s, and it would happen often in other markets. The world is not static and sometimes the situation changes and lots of supply is soaked up.
> It is a luxury good that only a very small number of people care about
This is an incorrect and incredibly out of touch comment fragment. Computer part derivatives are an essential item to economic activity in most countries.
> It is a luxury good that only a very small number of people care about
The world runs on computers. It is as essential as oil for the functioning of societies. Increase in silicon costs is going to increase costs unilaterally across the board. It happened during the pandemic and something similar will happen now. If anything it should be a wake up call to countries to start thinking about securing their own supply chains.
> Apple did stuff like this all the time at their high point in the late 2000s and early 2010s, and it would happen often in other markets.
Interesting in that I thought about their purchase of $1B of solid state memory at the height of their iPod run. The difference is that Apple had a hit product that was selling as quickly as they could be produced and there was a legitimate need if they wanted to meet the demand.
FTFA:
> No, their deals are unprecedentedly only for raw wafers — uncut, unfinished, and not even allocated to a specific DRAM standard yet. It’s not even clear if they have decided yet on how or when they will finish them into RAM sticks or HBM!
I don't consider this legitimate. It's not illegal, but it sure seems unethical and scummy and it pissed me off. OpenAI throwing its weight around is harming ordinary people who aren't competing with them.
> The difference is that Apple had a hit product that was selling as quickly as they could be produced and there was a legitimate need if they wanted to meet the demand.
What if OpenAI expects to be in the same boat? Their "hit product" is just R&D and training for new very large models. Of course if they're wrong, they've just set a huge pile of their own cash on fire.
If there was a law against buying the supplies of materials and letting them rot in a storehouse just to deprive competitors of them, your argument would be what OpenAI would try to make in court...
> A presumably short-term price spike in RAM of all things is a non-issue. It is a luxury good that only a very small number of people care about
Um... What?
Pretty much every adult owns one or more items with DRAM chips in them and depends on businesses that use even more.
The supply crunch will effect a surprising spread of the economy given how ubiquitous computers are now.
Looking at delivery dates, the dram price blip could last over a year and the price blips further down could last even longer.
To add to your message.
Memory is everywhere. In computer, phones, fridges, TVs, cameras, toys, watches, all kinds of home and industrial appliances.
> The supply crunch will effect a surprising spread of the economy given how ubiquitous computers are now.
If the OpenAI-induced supply crunch causes the AI bubble to burst, I may drop dead from irony-poisoning.
> It is a luxury good that only a very small number of people care about
It's raised the price of my in-progress workstation build by several thousand $, and now I'll likely not be able to build it. :(
I _really_ hope it's a "short term" price spike, but I kinda doubt it. :( :( :(
Who in developed countries doesn't buy computers and by extension ram
Who in the developed world doesn't have a few luxuries? Pretty much all of history people have had to make do with RAM being a lot less accessible than it is now. It isn't essential and people can still buy RAM in the rare situations where they actually need it.
There is nothing here worth invoking the legal system over. OpenAI can buy huge amounts of RAM if they want. Good luck to them, hope it works out, looks like an expensive and risky manoeuvre. And we're probably going to have a RAM glut in a few years looking at these prices.
DRAM is one of the categories of advanced semiconductors that the US considers important enough to national security that exporting it to China is forbidden. It's a fundamental industrial product.
Yeah. Companies like OpenAI need a lot of RAM. That is why they just bought up what is apparently a material chunk of the market.
There is a certain level of crazy that crowds can find when people identify something as a fundamental industrial product critical to national security and simultaneously someone is calling for companies buying a lot of it [0] to be made illegal. If something is critically important to industry then companies should be encouraged to dump as much money as they like in the sector. Otherwise industry will suffer.
[0] And OpenAI is probably going to turn out to be closely associated with US national security too.
They didn't actually buy up the finsished product they actually require. Arguably they raised the price of an input they cannot immediately use hurting themselves by raising the price for what they do actually use in order not to serve a need but to hurt others including the 99.9% of households that use devices with RAM.
It is the malicious purpose and the clear harm to most of America that ought to provide motivation to enforce any law this is at odds with.
Pretending computing is a luxury in 2025 is nonsense as is ignoring the obviously manipulative purpose that is so clear.
> It isn't essential and people can still buy RAM in the rare situations where they actually need it.
There are about five billion smartphone users worldwide. An increase in the price of RAM will, for a start, increase costs for those 5 billion, as smartphones do not last forever.
I don't follow how computers are not essential.
[flagged]
[flagged]
There's nothing dirty about this deal. When making a large deal with one vendor he didn't disclose to them that he was making a deal with another vendor. That's pretty normal when you're trying to buy a lot of stuff. Otherwise, they can collude to shake you down.
I'm not thrilled about this genre of "guy I don't like does totally normal thing so it's bad". It's too engagement baity.
EDIT: Though even that may be wrong. TechCrunch reports that it was a joint meeting between the South Korean President, the heads of the two companies, and Sam Altman. I won't claim that TC is the bible but there's lots of stuff being reported that makes no sense, and this is a good deal for both these companies so it's more believable than news from someone that OpenAI is going to buy a bunch of wafers and stick it in a warehouse.
https://techcrunch.com/2025/10/01/openai-ropes-in-samsung-sk...
It's bad for consumers period. A deal that hampers 40% of global supply shouldn't be a thing, it's predatory. I know DRAM is not a necessity, but considering that PCs are going to be affected means this affects real things like schools and hospitals. There's being smart while making a deal and there's knee-capping the market with your leveraged to the tits business
but where do we draw the line? Would 39% be okay? and who's gonna draw the line and enforce it?
I hope this is not a serious argument. This deal is OOM larger than other deals
The line imo is the amount of DRAM OpenAI actually needs/can use. If they end up piling some of it in a warehouse just so nobody else can use it, lock em up.
Do they actually have the cash to buy all of this RAM? If not then it should be very illegal.
There's no need to make it illegal: it doesn't happen because the companies with the RAM to sell are motivated by profit. The problem is that OpenAI has plenty of money to buy RAM. They raised $40 billion in April.
You’re just talking about putting limits on capitalism. Be careful, that’s blasphemy.
> OpenAI isn’t even bothering to buy finished memory modules! No, their deals are unprecedentedly only for raw wafers — uncut, unfinished, and not even allocated to a specific DRAM standard yet. It’s not even clear if they have decided yet on how or when they will finish them into RAM sticks or HBM! Right now it seems like these wafers will just be stockpiled in warehouses
That is not "totally normal".
It's not really that different from Apple reserving wafer starts on TSMC's next node and so on. It's just that this kind of capacity requirement has rarely shown up in DRAM before. Vendors prefer this kind of capacity reservation over a more variable finished product requirement. It allows them to know that they can build at the bottleneck rather than having to start up more capacity and then having that lie idle while everything downstream in DRAM packaging and DIMM production can't actually consume anything.
Did Apple reserve 40% of the supply and cause a massive shortage. If not, then I don't see how its the same.
Not only do they buy 40%, they bought 100% of TSMC's 3nm capacity for a year, locking everyone else out of 3nm chips.
Apple maintains efficiency and single-core performance dominance with the M-series processors by reserving capacity for cutting edge fabrication processes.
They're also reserving capacity on something that doesn't exist yet, and finalize their purchases into finished products. OpenAI is reserving wafers so others can't buy them. They are not the same
I don't buy it. It's easier to make this argument for companies that are building their own hardware, since they know it can be immediately used. OpenAI's move is tantamount to hoarding for the sake of strangling competition. There was plenty of supply to allow for their plans without this move (especially since they will probably go bankrupt at this rate).
How exactly could they 'hoard' this? There's no place in the world to store that much undiced wafer. It will all go bad.
If stored in proper conditions, the shelf life of undiced wafers is pretty much limitless. But if moisture or dust get in, it’ll start to have corrosion and other damage. Diced wafers on the other hand are placed on UV tape and I recently learnt that the shelf life of that is like a few months at best.
Nevertheless, unless you’re not storing them in a clean room of appropriate class (or in a vacuum pack), assume 18 months for undiced wafers.
In terms of size, I would think that stored in cassettes or carriers a vacuum sealed, it shouldn’t be too much. A 25 wafer pack is about half meter cube? A 40ft shipping container is about 68m^3. So one container can store about 140 packs of 25 wafers each. That’s 3500 wafers. They’re talking about 900k wafers. That’s about 260 containers. Call it 300 with extra space and stuff. Not exactly hard to provision.
In what time will packaged wafers go bad? What is this based on?
Wonder if they'd rather have it go bad than let their competitors get it?
Granting this premise is true (I have no idea), that makes it even worse. They would deliberately be hoarding 40% of the global supply not to lock in for future growth but simply to make sure no one else gets to have it. It’s figuratively setting chips on fire.
It seems not normal (in the sense that it is obviously quite weird to but like half of the world’s RAM supply). But I wonder if they are also just not ready to announce what they are doing with it?
I mean with that many wafers, I guess it is possible that they’d be doing something pretty custom with the things…
> But I wonder if they are also just not ready to announce what they are doing with it?
Didn't they announce some kind of AI-silicon recently? Wouldn't it be for that?
> To be clear - the shock wasn’t that OpenAI made a big deal, no, it was that they made two massive deals this big, at the same time, with Samsung and SK Hynix simultaneously
That's not "dirty." That's hiding your intentions from suppliers so they don't crank prices before you walk through their front door.
If you want to buy a cake, never let the baker know it's for a wedding.
What they mean is that they bought 40% of all RAM production, they managed to do that by simultaneously making two big deals at the same time. It's buying up 40% of all RAM production with the intention to have most of it idle in warehouses that is "dirty". And in order to be able to do that, they needed to be secretive and time two big deals at the same time.
> It's buying up 40% of all RAM production with the intention to have most of it idle in warehouses
They have no incentive to purchase a rapidly-depreciating asset and then immediately shelve it, none
They might have to warehouse inventory until they can spin-up module-manufacturing capacity, but that's just getting their ducks in a row
The incentive suggested in the article is to block other competitors from scaling training, which is immensely RAM hungry. Amongst other things. Even Nvidia could feel the pressure, since their GPUs need RAM. It could be a good bargaining chip for them, who knows.
I'm not saying it's true, but it is suspicious at the very least. The RAM is unusable as it stands, it's just raw wafer, they'd need a semiconductor fab + PCB assembly to turn them into usable RAM modules. Why does OpenAI want to become a RAM manufacturer, but of only the process post-wafer.
> The RAM is unusable as it stands, it's just raw wafer, they'd need a semiconductor fab
The wafers are processed. That means Samsung/Hynix have taken the raw ("blank") wafers, then run them through their DRAM lithography process, etching hundreds of DRAM dies ("chips") onto the wafer.
You could attach test probes to individual chips on the wafer and you'd have a working DRAM chip. In fact, that's how testing is performed: you connect to each die one at a time with a "probe card" which supplies power & ground, plus an electrical interface for functional testing.
If OpenAI takes possession of the processed wafers and wants finished RAM modules, they need to do a few things: test each die (expensive), saw the wafer into individual chips (cheap), package them (moderately expensive), test them again (medium expense), and then assemble the final module (inexpensive). Modern semiconductor test facilities cost billions of dollars and take years to build, so they'd need to immediately outsource that work (typically done in Southeast Asia)
OpenAI likely doesn't want to do any of this. They probably just want to make sure they're in control of their own destiny with regard to DRAM, then decided the best place to accomplish that was by cutting deals directly with the DRAM semiconductor producers. This will allow them to take the wafers to the existing supply chain, then contract them to turn the wafers into finished modules.
You need 40% of the world's ram for testing?
Each die gets tested. Not that they’re testing their stuff with these dies.
> They have no incentive to purchase a rapidly-depreciating asset and then immediately shelve it, none
It screws up the price for their competitors. That's an incentive. Particularly with so many "AI datacenter" buildouts on the horizon.
This wasn't buying a cake from a baker, this was a bakery buying 40% of the flour in the world so nobody else can sell wedding cakes, but now there's gonna be no bread for a year
That's not the dirty part. This is the dirty part:
> OpenAI isn’t even bothering to buy finished memory modules! No, their deals are unprecedentedly only for raw wafers — uncut, unfinished, and not even allocated to a specific DRAM standard yet. It’s not even clear if they have decided yet on how or when they will finish them into RAM sticks or HBM! Right now it seems like these wafers will just be stockpiled in warehouses
> OpenAI isn’t even bothering to buy finished memory modules
And? Why should they be obligated to pay for all the middleman steps from fab down to module? That includes: wafer-level test, module-level test (DC, AC, parametric), packaging, post-packaging test, and module fabrication. There's nothing illegal or sketchy about saying, "give me the wafers, I'll take care of everything else myself."
> not even allocated to a specific DRAM standard yet
DRAM manufacturers design and fabricate chips to sell into a standardized, commodity market. There's no secret evolutionary step which occurs after the wafers are etched which turns chips into something which adheres to DDR4,5,6,7,8,9
> It’s not even clear if they have decided yet on how or when they will finish them into RAM sticks or HBM
Who cares?
The implication here is that the primary goal is to corner the market, not to use the supply. If you aren't going to use them anyways then of course it is silly to pay for them to be finished.
Do you think that's fine, or do you think that implication is wrong and OpenAI does actually plan to deploy 40% of the world's DRAM supply?
> The implication here is that the primary goal is to corner the market
You have no evidence of that. Even at face value, the idea of "cornering the market" on a depreciating asset with no long-term value isn't a war strategy, it's flushing money down the toilet. Moreover, there's a credible argument OpenAI wanted to secure capacity in an essential part of their upstream supply chain to ensure stable prices for themselves. That's not "cornering the market," either, it's securing stability for their own growth.
Apple used to buy-up almost all leading-edge semiconductor process capacity from TSMC. It wasn't to resell capacity to everyone else, it was to secure capacity for themselves (particularly for new product launches). Nvidia has been doing the same since the CUDA bubble took off (they have, in effect, two entire fabs worth of leading-edge production just for their GPUs/accelerators). Have they been "cornering" the deep sub-micron foundry market?
> the idea of "cornering the market" on a depreciating asset with no long-term value isn't a war strategy, it's flushing money down the toilet
OpenAI's entire business strategy thus far can be summarized as "flushing money down the toilet", so that isn't actually as unlikely as you're making it sound.
Yes, they've made insane scaling bets before and they have paid off.
If what we've heard about no acceptable pre-training runs from them in the last two years trying to increase the memory for training by two orders of magnitude is just a rehash of what got them from gpt2 to gpt3.
I like how one of the reference links betrays how the article itself was researched, possibly written; HN hides the end of the url, which is "utm_source=chatgpt.com":
> https://www.economist.com/business/2025/11/19/cracks-are-app...
You have to appreciate the irony :)
Quality of the writing and thought also reveals the construction method.
The real disappointment is none of the sources are linked in the text. Instead, it's just random underlined words, the classic chatgpt over formatting with lots of extra underlining and bolding. I appreciate that a 10-15 minute long article summarizes a 25 minute video, but it's hard to hide the real author.
Quite lazily done and just not pleasant to read.
Would it be as ironic if one wrote an article about Google and used Google search to research their sources?
If the article were about Google ruining everything, yes.
No because google can both ruin things and be a tool that is helpful in certain circumstances
If OpenAI were actually using the RAM that’s one thing - but stockpiling raw wafers in warehouses is egregious.
Guys, these are silicon wafers not bars of steel. You can't just stockpile 40% of annual capacity in a warehouse long-term. I would be incredibly surprised to find that any such large scale storage facility exists. Any storage of undiced wafers is temporary while the manufacturing pipeline proceeds. You've seen the pictures of clean rooms and stuff. There's no way you spend all that effort to make the wafer and then just stick it in a warehouse. Who even is going to make such a large cleanroom facility? And for what exactly? It doesn't even pass the basic sniff test.
Much more likely this is just a detail of the contract so that OpenAI can guarantee allocation. I would be surprised if the actual wafers entered OpenAI hands before being fully packaged.
Why would you need a clean room if it is only for storage? Pack the wafer in a clean room and store the packages in an ordinary storage.
Correct. It's rather routine for wafers manufactured for "last call" orders on ASICs exiting production to be stored as wafers due to not knowing how they need to be Packaged.
Really, you don’t think there is enough storage room in the world?
It would require half of one distribution center of a major retailer.
> It would require half of one distribution center of a major retailer.
That also meets the specifications of a clean room, and is actively maintained as one?
If OpenAI bought 40% of the annual capacity of finished memory, with the goal of using it in their server farms ASAP, that's one thing.
But unfinished wafers that still need to be protected to finish the manufacturing process, that OpenAI itself does not have any capability to do?
That to me looks like a preemptive strike against competitors, which also affects any other industry that requires RAM, in an attempt to develop a monopolistic position.
I can't see how this isn't a massive national security issue for any country that needs devices requiring RAM for new systems and maintenance of existing ones (pretty much all of them...) to manage critical infrastructure, national defence, public and social services, and so on.
You do not need clean room specifications for storing wafers.
Maybe they’ll build nice little forts with the wafers
If they do, it'll be a house made of glass.
It would be fitting for a company, or rather, a former nonprofit organization, that builds on empty promises.
Yup. Not exactly a move "for the good of the world".
Also consider:
Warehouses of small, high-value items that are fungible and untraceable.
That will create multiple huge targets for a big heist. And they'd best have good eyes on their security people too.
Sounds like something big enough for organized crime to target.
Secondary RAM Manufacturing Had Stalled. Budget brands normally buy older DRAM fabrication equipment from mega-producers like Samsung when Samsung upgrades their DRAM lines to the latest and greatest equipment. This allows the DRAM market to expand more than it would otherwise because it makes any upgrading of the fanciest production lines to still be additive change to the market. However, Korean memory firms have been terrified that reselling old equipment to China-adjacent OEMs might trigger U.S. retaliation…and so those machines have been sitting idle in warehouses since early spring.
My takeaway, this sounds like an comparably easy fix for the consumer market, if prices are somewhat guarenteed to stay mid term significantly above this years spring floor for someone to sweep up the margins and negotiate a somewhat reliable way to get the last gen production lines up and running again. Will take at least half a year to pick up, but this is not a longterm RAM doomsday scenario in any sense.
I'm more worried about the low to mid-end embedded systems, that a have a dollar budget for memory components, that could get unbearably slow for the current/next gen if manufactures just use the bare minimum of RAM the bloated TV or tablet OS can run on, if the 1GB raspberry move is any indication of that. And consumers stuck with no way to upgrade them to a reasonably usable state.
One of the big problems here is that all of the hardware companies have been burned by hype before (e.g. crypto). No one actually believes that these AI companies will still be around in 5 years so spending billions to build factories for them doesn't make sense.
I don't know what companies will be around 5 years from now, but I would bet there will be more demand for RAM and the price per GB will be at least what it was before this price shock.
RAM is a very cyclical market, historically. You can look at $MU historical charts and kind of see that it trades like a cyclical (compare it to $RIO, for example).
Cyclical companies are easily burned by investing in infrastructure right at the peak. It happens all the time with little mining companies, and I think DRAM manufacturers are sort of the mining companies of tech.
Cyclical markets are the sort of thing 'National Strategic Reserves' should address...
Am I crazy for wanting this to be in Full ECC RAM modules suitable for composition into many device factors with hope that we'll finally go to reliable memory for all markets as a result?
The problem is that the cycle time is about the same as the generation time. By the time the current cycle completes, we'll be on DDR6/HBM4.
If prices are guaranteed to stay high in the medium term, then I'm not sure that's acutally a fix.
I have a 32 GiB DDR5 set, happy to exchange for $500K in cash or a nice little house in Spain.
I have a nice little house in Spain, I'm willing to trade it for 128GB of RDIMM DDR5.
No issue pal, I have that too. I can survive on my 32 GiB set for a while.
4800MHz single rank ok?
I'm curious how OpenAI has the funds to pay for 40% of the worlds ram production? Sure they are big and have a few billions but I kind of assumed that 40% for a year or whatever they are buying is easily double digit billions? That has to hurt even them, especially because they cant buy anything else?
Also what are these contracts? Surely Samsung could decide to cancel the contract by paying a large fee but is that fee truly so large that getting their ram back when prices are now 4x of what they used to be is not worth it?
I found this which claims ram market in 2024 was almost 100 billion: https://www.grandviewresearch.com/industry-analysis/random-a...
I assume this includes more than just the raw price of modules but Openai only has 60 billion in funding altogether and was aiming for 20 billion ARR this year. This sounds like they are spending maybe half their money on RAM they never use? That just doesn't add up.
Ponzi scheme [1] , Anticompetitive hoarding [2] , Cornering the market, Raising rivals' costs (RRC), Consumer welfare harm, and so on
[1] https://en.wikipedia.org/wiki/Ponzi_scheme
[2] https://en.wikipedia.org/wiki/Hoarding_(economics)
I think the event is big enough to stop them and send them behind bars.
> Samsung
I think that Samsung -and other manufacturers- have been intentionally limiting their production capacities so as not to devalue the prices of their chips (for SSD at least) so may be they are an interested part. This, combined with the madness we are seeing, is abuse^2 . I think they should also end up behind bars.
I'm reminded of the 1983 deal to corner the market on Frozen Concentrated Orange Juice.
Or the Hunt brothers and silver which was just a few years before that.
How'd that turn out? https://en.wikipedia.org/wiki/Silver_Thursday#:~:text=On%20J...
That's a great comparison. The consequences are pretty universal too. History implies this won't end well for OpenAI.
I remember it didn’t work out well for Randolph and Mortimer. Sam may pull it out, though, if he just sells the DRAM now while the market is still hot.
"Mortimer ... we're back!"
Sell! Sell! Get back in there and sell!
"...their deals are unprecedentedly only for raw wafers — uncut, unfinished, and not even allocated to a specific DRAM standard..."
wtf. life sucks.
That $7 trillion number makes more sense now.
This is missing a key part of the picture - Nvidia just announced that partners will need to source RAM themselves.
OpenAI is basically ensuring that they can actually get the chips they need for the DCs they are building.
I can’t guess as to what move came first (Nvidia policy change or these DRAM deals) but I would bet this is a large if not larger factor here than “bloc my competitors.
Guess OpenAI finally found a business model that works: memory futures.
I wonder if this kills Valve's Steam Machine and Steam Frame
Yes, I have an older gaming PC from ~2018 that I keep putting off upgrading (first GPU prices skyrocketed, now this...) and was hoping to replace it with a Steam Machine next year. Will be endlessly bummed if that doesn't happen.
Will also be interesting if Sony/Microsoft was planning on releasing a next-gen system anytime soon, and I wonder if this will affect Apple's hardware at all.
I’m in a similar place, tried bazzite recently and was pleasantly surprised that it plays all my games well
I've been selfishly wondering the same thing. The Frame is on my shortlist (as long as the price wasn't too crazy).
As is mentioned in the article, depends on when they bought their DRAM contracts. If they were in before this then they'll be fine for a while.
16 gigs announced in the Steam Frame, hope they bought the memory already.
https://www.trendforce.com/news/2025/12/05/exclusive-memory-...
> Lenovo has begun notifying clients of coming price hikes, with adjustments set to take effect in early 2026.. Dell is expected to raise prices by at least 15-20%, with the increase potentially taking effect as soon as mid-December.. Dell COO Jeff Clarke warned that he’s “never seen memory-chip costs rise this fast,” .. Lenovo [cited] two key factors: an intensifying memory shortage and the rapid integration of AI technologies.. TrendForce has downgraded its 2026 notebook shipment forecast from an initial 1.7% YoY growth to a 2.4% YoY decline.
Matt Levine, https://www.bloomberg.com/opinion/newsletters/2025-12-01/ope... | https://archive.is/S3MPq
> your business model might end up being sort of a … startup incubator or private equity firm; you’d spend your time starting or acquiring companies on which the robot could work its magic. Your business model would be “general business, but with AI.” .. OpenAI has a $500 billion valuation largely as a bet that a lot of the value of AI will accrue to its builders, but it could hedge that bet by owning the users too. Either it will sell AI at high margins to lots of businesses, or it will sell AI at lower margins to lucrative businesses that it owns.
An American company, combined with American tariffs, and fear of American retaliation.
Getting pretty tired of that place tbh.
[dead]
Can OpenAI hurry up and go bankrupt?
> On October 1st OpenAI signed two simultaneous deals with Samsung and SK Hynix for 40% of the worlds DRAM supply.
The market doesn't believe they can pay for the Oracle cloud deal. Why do these vendors believe OpenAI can pay for 40% of the world's DRAM?
They definitely can, as anybody could, for a short enough period of time. The only ones betting they'll be able to sustain it for a long period are the ones paying the currently very inflated price for ram.
> for a short enough period of time
but it's not a short term deal. funny enough, there's a cohort of bonds that never even made one payment. They are called first payment default bonds. There is a cheekier nickname for them, but it escapes me at the moment.
Suckers Buckers seems safe enough for work to say on a text only Internet thread. Just don't read it aloud.
Ironically this deal has increased DRAM prices so much that if OpenAI doesn't pay, Samsung/Hynix can make more money selling on the open market.
unless everyone else walks when they sense the "big buyer" has reneged.
With the amount of hatred for anything OpenAI it’s not surprising the author chose a clickbaity title. HNs quality of posts are going down and instead of objective analysis I often see very polarized and flamy articles, titles etc.
That’s funny, I found the post high quality and interesting despite agreeing with your point about the title.
It to the point that bloggers know they can say things like "OpenAI did 9/11!", provide no sources, and it still will make the front page and have a large increase in traffic to there site.
I’ve been on this site a while, and it has always been like this.
If it's confirmed that Altman's played the Koreans against each other, it's going to cause a furore in Korea.
Can OpenAI use their stockpile as leverage? e.g. threaten to sell their stockpile to a market that's about to stabilize to crash prices.
And are Samsung, etc. happy for this state of affairs to keep their prices elevated but not seem like the bad guy.
I bet RAM is a great carrot to dangle when negotiating first-party GPU allocations with Nvidia and AMD
The biggest question is, can they even pay for half of the deals they have been making?
nice 5d chess move, saltman paobably got this idea from GPT 4.5 high thinking
Strange how LLM vendors are flooding the market with reasons not to do business with them. Every paid agentic interaction contributes to all the bad behavior we are seeing. From out of control web scraping to buying up available hardware LLMs are turning out to be highly efficient misery manufacturing mechanisms.
What happens to enterprise apps like SAP and S4 on the cloud that swallow huge amounts of RAM for their in-memory database? AWS, GCP, and Azure must be running around? With their predatory pricing customers were already hesitant about upgrades, it now gets more difficult. Extend EOL for ECC?
Force a divestiture of Microsoft.
I can’t help but wonder if their product orchestrated this deal.
Daniel Suarez's 2011 Daemon.
Is it possible that GPT-6 could possibly be highly reliant on RAM? For example running smaller long context models on the CPU? Just a thought, I could be far off.
I didn't keep up with the news lately and had to pay 650 bucks for two 32GB DRAM5 sticks for my new build. Yay.
I see no real evidence of any of the claims in this article. Why is it garnering so much attention?
RAM being super expensive right now is a hot topic, and I imagine people are curious about why, so anything that purports to explain that is somewhat likely to at least get a cursory look from people. Without taking a stance on either whether this is happening here, if enough people look at something and happen to think it makes sense, a lack of evidence might not be enough to prevent the narrative from taking over; otherwise, we'd never have issues with people believing false things in the first place, and that happens all the time.
Does this violate anti trust law?
its criminal to underline non links in ur site come on
Why would one sell something cheaper than the current market value? I wouldnt care if i had stock or not, prices should be what things cost.
When someone buys up all the supply, the price will rise because of supply and demand. It's the nature of markets.
TFA mentions that if Samsung and SK Hynix had known what shenanigans were underway, they would have pursued better pricing terms.
I mean, say you have something in stock that others sell for twice as much as last week and it will cost you twice as much to restock, why would you sell it for last weeks prices?
Why would it matter if you have anything in stock or not? What does it matter what you've paid for it?
If it will be hard to restock and ill most likely will have to sell "no" ill be even more motivated to ask more for it.
almost feel like OpenAI's recent "fall" is a decoy setup by them intentionally.. something's cooking.. maybe they wanted to buy back their own shares at a lower price?
Imagine the outrage if OpenAI built their own fab or memory factory. Like back when Henry Ford built his own steel foundry.
I fail to see why that would be outrageous.
That would make sense _except_ the amount of time and specialized knowledge it takes to build one. Easier by _far_ to go deal with TSMC and Samsung
Altman was already unpopular. After this will he be able to show his face in Silicon Valley?
I don't think Ellison is popular, but he seems to be doing fine with all his billions.
And all these data centers they want to build around the country. When consumers can’t get devices they want maybe they’ll fight even harder against these data centers being built in their back yard. He’s not making any fans with this move that’s for sure
[dead]
This will make AI even more palatable for the general population /s
[dead]
[flagged]
Now this... was a really good move.
OP is also marginally underestimating the impact this move would have on Google's competitiveness - they are making huge gains prototyping at light speed; this will halt their AI hardware acceleration plans pushing them back into slower software development on ever aging hardware.
It also shows why Nvidia is not afraid of competitors coming out with new desgings that obsolete their hardware: what good are superior designs with no fabs to produce them?
every one of these things that make the deal "good" for OpenAI is a direct result of negative externalities for everyone else: competitors, consumers, and people who wouldn't care otherwise.
The article even says that they don't have an obvious plan for how to use the wafers they bought, and very clearly suggests that this is purely an anticompetitive tactic to force everyone else to eat a price increase that OpenAI doesn't need to face. It's clever though because if any regulatory agency starts asking questions (not that they would do that in the current USA political climate) then OpenAI can just say it's a strategic reserve, we have plans to do something with it, etc. etc. What are you going to do? Take them to court and force them to auction off some % of the stock? Set an industry-wide limit on wafer inventory? Fine them? You'd need to find some evidence that it was done maliciously, and good luck with that.
There are some negative elements of captialism that we might simply have no reasonable regulatory apparatus to deal with. Preventing indivduals and companies from having so much market power in the first place seems to be the only thing that can work consistently.
More anti sam anti AI propaganda, nothing dirty about this deal
As an AI researcher, I thought it was relatively well established (at least among my colleagues) that being pro-AI actually meant you were anti-Sam as well. He's the worst actor in the industry and has done an incredible amount of damage to its brand.
Care to elaborate on this?