I'm not a business guy, but I think I would be very worried if I created any opening (even on the very long term) for competition to justify itself being built up to meet demand (or national security requirements of other countries) on an industry I mostly control, because I prioritized extremely high prices for a few years that might have the effect of changing the arithmetic for potential competitors that wouldn't have done it before.
It makes me consider whether there would be retaliation in later years against the US intellectual property system in countries other than China. While RAM is a tangible product, the areas it serves aren't really.
If I understand this comment correctly, it's saying that Micron and the others would be better off selling RAM at below the price they could get, in order to keep competitors from deciding to enter the market.
Such collusion might well be a violation of anti-trust law, but in any case, it wouldn't work. If they decide to sell at below-market price, they must (by definition) be preventing some customers from buying by some non-price mechanism (eg, only selling to companies run by other members of the CEO's family). Potential competitors would then see an opportunity to sell to these excluded customers.
Not exactly, I think it's saying that Micron would be better off increasing supply themselves (and thus lowering prices) than waiting for a competitor to come and provide more supply to meet the higher demand.
If there was enough existing competition, that might be the right move, because if they don't increase capacity, a competitor probably would in order to gain more market share. But I think the barrier of entry for making RAM is high enough that Micron can probably get away with keeping supply low (and thus prices high) for at least a few years before a new competitor appears. Assuming demand stays high that long.
They have two massive DRAM fabs (56000 m^2 cleanroom size each) that they started construction on in 2022 and 2023 in Idaho, with production starting in 2027 and 2028.
The point they were stating was that if they kept up the high prices, that leaves room for a low priced competitor, likely from China, to occupy the lower end (and larger) share of the market. Like what BYD did to the EV market everywhere except the US. Or what Chinese mobiles did to the smartphone market. Eventually they become so powerful that they start going upmarket, trying to catch everyone in the higher end segments too. And by the time the Western companies think they can retaliate, too bad, China's got them by the bollocks on raw material.
But the reason that semiconductors are so vulnerable to boom/bust cycles (going back decades) is that takes a very long time to build the fab- you have nothing for years, and then the fab completes and enormous numbers of chips get made. So if you get demand wrong either direction you have problems. Either you flood the market or prices go way up.
I don't think it is possible to beat the fab construction delays. This isn't about things being slow in the US: TSMC took four years to go from site selection to some production at their Fab 22 plant in Taiwan (it is currently only 60% built out, is my understanding). And TSMC is best in the world at getting frontier (logic) semiconductor fabs built, it took them slightly longer (five years) for Fab 21 in Arizona. Logic fabs are not the same as memory fabs, but they are close enough to be a good comparison for timeline.
Basically, unless you started fab construction (specifically to make memory chips) in 2023 you cannot change production for 2027 and 2028. The only thing that will affect prices between now and 2028 is demand, production can't go up because the fabs to make those chips don't exist. If you started now you might get memory chips in 2031, and can you bet 40 billion dollars that the demand will still be there? (A current generation logic fab costs about 20 billion. By 2031 a new frontier logic fab will cost about 40 billion). I don't have prices for a current memory fab, but as I understand it they are comparable.)
Lowering their prices would result in supply running out sooner. Their competitors would still get the same bite at the apple but would have a longer window where they are the only option. Or resellers would just arbitrage the surplus left on the table.
Yeah there is no such thing as selling a commodity below its market price on the open market. Someone else will come and arbitrage that away, or it’s not an open market.
A well-known anticompetitive tactic is for a big player with a vast bank account to move into an area and eat losses while waiting for their competitors to starve. Well observed in the 90's movie rental market, so certainly real.
You don’t lower your prices on the off chance a competitor might enter the market in the future. You wait until they have invested capital but before they start making profits and are at their most vulnerable.
I don't think that would work. They wouldn't demolish the competitor's factory when it goes bankrupt. Instead the factory, being an asset, gets sold for cheap, and another company gets a shot at competing while having much lower debts. This process could repeat a few times until the debt is low enough to profitability compete.
So if that's the expected outcome, it might be better to prevent competitors from building factories in the first place.
No, because they're hardly enforced. China has been blatantly dumping a buttload of products for decades and the U.S. has been asleep at the wheel as it continues to hollow out our domestic manufacturing industries even further.
I mean, we can assume the people who make the memory love getting huge prices for it. But beyond that, I just dont see why someone who invest several multi billion dollars to squeeze into a potential window that may or may not last 3 years. Especially with how volatile the whole AI bubble is.
> Especially with how volatile the whole AI bubble is.
There's nothing volatile about the AI bubble, it's been going on for a while regardless of cost and that will continue because of politics - which is rather transparent.
> I just don't see why someone who invest several multi billion dollars to squeeze into a potential window that may or may not last 3 years.
Interesting, Chinese companies don't seem to know this, they're investing bigly in RAM and other silicon despite being sanctioned and tariffed up the wazoo. Should we tell them that it's a crunch for no more that 3 years so they don't waste themselves for nothing?
The more interesting question is why Chinese RAM is hugely tariffed, if not for preserving the crunch and its monopoly pricing?
> I mean, we can assume the people who make the memory love getting huge prices for it.
Your mistake is to think of the RAM makers instead of the systemic reasons for the problem, these would have to include tariffs, sanctions, subsidies, etc that drive market pricing, RAM manufacturers and their potential competitors.
> Interesting, Chinese companies don't seem to know this, they're investing bigly in RAM and other silicon despite being sanctioned and tariffed up the wazoo.
It's also possible they're preparing for the possibility that they get cut-off and so profit is not the primary motivation. As you mention, they're already sanctioned...
Also, CMXT started in 2016 so that gives some indication of the timelines involved. Even in Taiwan it's like a year and a half just to get the building made and these are experts operating in their own country working 24/7. That timelines doesn't include equipment into it, which is also in shortage, and getting lines qualified.
> Yes, one of those markets with infinite demand where supply has no impact on price. Those ones.
In economics, "demand" doesn't mean "I want this", it means paid, funded demand. 100s of millions of "free" LLM users aren't funded demand, they're subsidized. This leads to front-running and scalping the silicon market at an unreal scale, it also monopolizes access to silicon and compute. Nothing happens by chance here.
The argument I've seen about why someone else doesn't start selling RAM is it takes a year or two to ramp production. But here's a CEO saying he's going to make sure prices stay high enough, long enough that you can spin up a competitor
> But here's a CEO saying he's going to make sure prices stay high enough, long enough that you can spin up a competitor
2 years is not remotely enough time to spin up a competitor. It's not even enough time for a current leading competitor to spin up another fab.
If you are a newcomer you are talking a decade or more before mass production. If you are HK Hynix perhaps 5 years or so.
The majors all have large fabs under construction, but the majority of new production from those facilities won't be coming fully on-line until 2028 at the earliest. Even from the date of first wafer shipped most of these facilities take 6-12mo to reach mass production level output.
Chinese DRAM is interesting only because they started building capacity (expertise) a decade ago. Only now are the expected to have any material impact on the total volume of the market. And if they want to gain even more market share, they will be under the same ~5 years to spin up a brand new fabrication plant. Perhaps a bit less because they are China, but still multiple years.
And it's important to point out how long these fabs have been under construction.
Micron started building it's new fab in 2022. They are expecting their first chips to be out by 2027 (maybe). The projected cost, $50B.
This is a fundamental problem which the free market cannot handle. It takes multiple years and billions for competitors to spin up. No sane bank or business person would try and get into cutting edge memory fabrication. That means the players we have are the players we'll likely always have (With MAYBE china entering the arena).
The free market only works when the required upfront capital for a new player to enter a business is relatively low. In industries where the players are static or shrinking, you have a free market failure.
Imagine being the sucker to go and do a big capex build out for a hypothetical demand in two years. Have fun explaining that to the investors who supported you. Also you probably need 5 years to cover your expenses at whatever the current costs are. If the volatile prices collapse you are in a world of hurt.
Someone will certainly take this bet - but it is a high risk one.
Micron is a trillion dollar company. The CEO would never say something like that directly. He's talking in the "business dialect" of English and you have to read between the lines a bit.
The CEO is saying nothing that everyone who analyzes the situation and knows at least something about semiconductor production wouldn't say. Demand is way up, supply cannot keep up, it will take some time to increase supply.
Micron is working on increasing that supply. The started building a massive fab for DRAM in 2022 in Idaho. It should start production in 2027. They started a second massive fab for DRAM next to it in 2023 with production scheduled for 2028.
This is the info that people should be looking at. Building these facilities takes time. I'm guessing they are thinking that demand is not going to decline anytime soon to justify this buildout. EDIT both SKHynix and Samsung are also expanding their DRAM production lines and new plants. So they are all working on adding new capacity.
Not to be snarky, but the CEO is saying that the supply will be tight for years. I know Econ 101 is over simplified but I do recall that when you reduce supply and demand stays constant, or increases, the price goes up.
I also recall some Econ 101, but I seem to recall a little but more than you since I also recall that the price goes up when the supply stays constant or even increases but the demand increases faster than the supply increases.
The CEO is saying that that is the situation we are in.
Demand has increased massively over a short time. Supply cannot increase as rapidly because for big demand increases it involves building new factories and equipment which takes significant time.
The CEO says that their new stuff will start coming online in 2028 but they can't forecast when supply will catch up.
or less, as you seem to forget what this whole discussion is about.
If they dont make sure the prices dont go up, they go up.
Once theyre up, and expected to stay up for the foreseeable future, competitors see a gigantic opportunity. Suddenly, the large amount of investment necessary to the ball rolling is worth it.
Once this opportunity is taken advantage of, the original market position of the company being able to freely set whatever price they wanted is eroded forever -- and in this particular case: it opens them up to being out-competed by the new competitor(s).
In this case those will will likely be coming out of china and be state backed, so theyll likely operate at a loss for maximum damage to get as much of the market as they can get.
in 5 years, Micron is likely going to go cry for state support to socialize their upcoming losses.
and before you say "but there are other competitors already": you should be aware that theyve previously been convicted of running a cartel. And the way theyve been behaving is exactly how they were behaving back then. While unproven, it is more then just likely that theyre still price fixing (and have been for years)
the price didnt just go up for consumers, they themselves have deeply dipped into the opportunity to maximize profits.
its so hilariously outlandish that eg SK Hynix has to pay out >$450k to union employees (per employee - not in total!) because they have a contract that a percentage of profits get shared as a bonus across the company. this payout is usually not particularly high.
Would you prefer the money goes to middlemen? When supply is constrained and the manufacturer doesn't price to match, all that happens is scalping. Even if the memory die manufacturers gave away their products for free it wouldn't budge consumer RAM prices in this market environment.
youre taking a very consumer oriented look, while we're explicitly talking about the perspective of the producer in this thread.
the consumer price barely matters for the perspective of the company evaluating wherever they can enter the market, because thats just a tiny fraction of it. And the largest piece of the market buys directly from the producer, SK Hynex, Samsung and Micron.
For consumers wanting to buy RAM sticks the difference is academic, i give you that. But for the capital that decides wherever to invest millions if not billions into getting a factory up and running - the price they can set for the b2b sales is the more important factor.
If it wasnt already proven to be basically whatever number they feel like setting... the opportunity would be a lot less obvious.
Large tech companies like cloudflare are spending developer hours lowering memory usage because they feel the impact too.
The idea that companies should have priority price controlled access to memory puts us even deeper into a situation where we can never afford local compute ever again.
For stuff I wrote that I run on my server, I have been converting from Clojure to Rust. Sometimes mechanically, often by hand.
I actually like Clojure more but even with GraalVM AOT it still tends to take significantly more memory than even naive Rust code (e.g. using Clone everywhere). If you’re looking willing to spend time abusing the borrow checker you can often get memory usage down to like 5% of what you’d get out of Java stuff.
The price is up because demand is way above supply (and will be for a long time because it takes years to expand supply to match the current level of demand).
If the manufacturers decided to keep the price constant that would NOT stop the price that most people pay from going up. It would just mean that the people who managed to buy from the manufacturer before the supply ran out would resell on the secondary market for a giant profit.
"would resell on the secondary market for a giant profit."
But you've no evidence that that "giant profit" would be more than the increase that the suppliers would apply.
Arbitrage allows the market to find an agreeable price; a middleman arbiter with a warehouse of RAM has an incentive to agree a price with a buyer because he's stuck with that inventory if he doesn't shift it, and he has costs related to financing it.
A supplier has no such incentive, they can just jack the price up and if sales drop off they run the machines at a lower rate, or sack employees. But sales aren't going to drop off, because the baseline for demand has been lifted
Once this "RAM Crisis" is over, what reason would manufacturers have to reduce prices?
They have to cover their fixed costs somehow. Collusion and cartels only get you so far. OPEC and Mexican drug cartels are still exposed to market pressures.
Exactly. Demand won't be going down for quite some time, so prices will stay high unless supply goes up. So where is the new competition entering the market?
If there's no competition entering the market, wouldn't that be a clear sign that this is not a free market but effectively a cartel?
The Big 3 (Samsung, Hynix, Micron) have been playing this game for 30+ years.
There used to be many more DRAM companies but most of them went bankrupt when DRAM was cheap because the big 3 will intentionally drive the price down to bankrupt the competition.
Now that they have a tri-opoly, they're milking the profits.
New competition will emerge but they don't have the capital of the big 3 so they likely won't survive the next DRAM winter.
You seem to be adding a lot of intentionality to price swings that seem expected when supply is relatively fixed while demand fluctuates. (Maybe you’re right, but it’s not the simpler explanation.)
The supply isn't fixed. A new plant takes 3-5 years to build and cost tens of billions. DRAM companies don't always know where the demand will be when the next plant completes. 2 years ago there was too much supply so DRAM price hit rock bottom. So DRAM companies held back on investing in new plants so here we are now.
They are not prioritizing extremely high prices. TSMC has a similar waiting list, so does ever other tool manufacturer. Memory capacity is being increased approximately as fast as it can be.
You can find public statements in the last 12 months from Samsung and Hynix about their aversion to scaling out production or increasing supply because they are concerned about profits.
Hard to know who to believe, on one side there are the absurdly high prices and statements from companies and on the other hand there are strangers posting on HN that it's actually all fine.
The memory business has always been extremely cyclical, and I doubt this time will be any different. The manufacturers have learnt this, and therefore to avoid boom followed by bust they try build for what they perceive to be longer-term more sustainable demand, not just knee-jerk increase demand to meet an immediate shortage.
The risk to any of the manufacturers, especially those with most market share, would seem to be if they are not all acting in sync. Micron's consumer brand, Crucial, had a great reputation, but now they have exited that business in favor or devoting all their capacity to currently higher margin products... will this come back to bite them later if they want to re-enter the consumer business?
What's different this time is that LLMs require orders of magnitude more memory than we could have ever used before. And I think it's getting obvious by now that inference will move to local compute, So everyone's laptops and phones would need a lot of memory.
The question the memory manufacturers have is, sure, we have all these orders out to the next three years... but will they actually be there to pay when the memory is delivered?
The memory manufacturers have been through bubbles before.
I don't deny that AI is "real", in that there is certainly something there, but I would hate to be the guy betting billions or even trillions of dollars that we're not in a demand bubble well in excess of what is justified by that tech at the moment and expanding capacity is a great idea. Of course, I would also hate to be the guy saying "no we shouldn't expand capacity" when it turns out that, yes, we should have, but the staggering profits being made in the meantime would cushion that blow pretty well.
> LLMs require orders of magnitude more memory than we could have ever used before
That's true. But will anyone have any money to pay for the LLMs? The AI market is currently being heavily subsidised by, ultimately, everyone else. But the economy in general is looking extremely dire at the moment, and that seems unlikely to last.
Basically everyone I know who doesn't have a tech job is struggling for money / feels like the world is unaffordable for them atm. Those without jobs are struggling to get them. And those with jobs are scared to change even if they don't like them.
And from what I read in the media, a lot of the economic indicators are backing this up: discretionary spending is down, private debt levels are rising, etc. All signs that the slack in the system is disappearing.
That's what people thought about a year ago, but with the amazing strides that smaller models have taken over the past year, people are now starting to question that.
> And I think it's getting obvious by now that inference will move to local compute
From the bottom of my heart, I hope that's the future.
I'm not convinced that will be the case. The AI companies don't want you to have local control. They want you to subscribe to a service that they can change at any time.
I grit my teeth when I type this, but (god help us) I think Apple is maybe the best (least bad?) hope here. They are the only big player with the hardware chops and without a current vested interest in getting you addicted to monthly AI subscriptions. I'm not saying this is highly likely... I'm just saying that out of the current major players, they're the ones with the ability and motivation to move in this direction in the near future.
The next best hope is probably just nVidia or AMD catering to the consumer market once again, after the AI bubble bursts or the datacenter market reaches saturation.
The outside hope is that some startup makes a business out of burning open weight LLMs to silicon like ChatJimmy... although ChatJimmy was bought up by AMD AFAIK.
It doesn't really matter what the AI companies want, if they misjudge the market somebody will just start a competitor and take it.
The relevant questions are: 1) Where are the economies of scale in the technology stack? 2) What's "good enough" to consumers, and how does that stack up with the relevant computing power needed? 3) What are the transaction costs along various system boundaries?
I think that the biggest force keeping inference in large centralized services is simply that provides a better product for the average user who doesn't care about local control (and the average user doesn't care about local control; indeed, for most people it's a misfeature, as then they have to administer their own hardware). HN is full of nerds that want to own their own stack; they're willing to put up with a little loss of capability to run Qwen 3.8 locally. But from the folks I know that have tried it vs. Claude vs. Codex vs. Antigravity, the local models are still pretty weak compared to what you can get by paying for a service. Most people will just pay for the service until the performance becomes indistinguishable and the price becomes less.
>It doesn't really matter what the AI companies want, if they misjudge the market somebody will just start a competitor and take it.
The vast majority of people are uninterested in local vs non-local models (not just LLMs but ANNs in general), only the services the models are providing. Even if performant local models can actually be served on device quickly, if the frontier labs don't want to go that way, they simply can and there's little threat the market can be taken from them that way.
If you can serve local models, the 'advantage' is that they can be run on the users device without that incurring any cost to you thus allowing flexibilty in pricing, but if such models can be made then whatever cost would be incurred to labs running on their own servers would be a rounding error and also inconsequential to them. That is to say, they can also compete with you on pricing. So unless someone comes along with something the frontier labs can't figure out or replicate, then it doesn't matter. Open AI might be 'forced' to drum down prices but only in situations where costs have also been drummed down. In such scenarios, margins may even become better for them.
That is not true. Semiconductor industry saw a similar supply crunch during covid and spent heavily for capacity. And when the crunch tapered off, they were left holding the bag.
Micron would probably not even exist, or be a sticker brand for some Chinese conglomerate if not for this supply shock.
It has honestly gotten frustrating to even read anything on the internet these days. It’s just conspiracies upon conspiracies all the time, and it’s constantly in your face.
If only they were good conspiracies with some weight to them, but no, they fall apart if you have a rudimentary understanding of the subject.
Which was caused by the exact problem they’re trying to avoid by not going on a spending spree to increase manufacturing capacity.
They had built way too much capacity during the 90s and were losing money on every chip because they couldn’t run the fabs at full capacity. That’s why they colluded to increase the price to save their collective hides.
A new competitor wouldn’t be able to enter the market and lower the prices without also losing money. That’s not the case today, the Chinese are trying hard to do just that.
It's easier to bear that there might be a conspiracy, than that this outcome is the result of chaos and stupidity and greed. Also it makes you think you are smarter than anyone else.
Fundamentally, it comes down to whether or not you think "AI/ML" can disappear tomorrow and the world would continue to run as normal.
From what I have seen, it cannot. It has become foundational the same way Internet 1.0 (eg. Email, Static HTML, Web Directories, SSL) became foundational for a significant portion of the global economy in the 1990s and 2000s.
All the big techs have accelerated zilch in actual features released or user-facing bugs fixed. All that seems to have accelerated is cybersecurity offense and defense.
And these are supposedly the most brilliant corporations around, so other companies are seeing even less gain.
I stick with my theory that LLMs are GPU-like, not CPU-like. They will rarely solve a problem like an adept, but they can solve 1 000 000 million low-hanging fruit problems within a second in a way that there was just not enough eyeballs for before. They can also brute force problems.
AL/ML isn't going away, but what that means for demand for various types of memory remains to be seen. LLM are clearly a commodity, and once they are "good enough" then it'll just be about price, meaning that smaller and less memory hungry models will win.
It's a bit like the Sun/etc workstation market vs PCs, and high end expensive PCs vs cheap ones. Once the cheap PCs became "good enough" then they naturally dominated.
Not only could the world run as normal without AI as we have it today, it would run better. The tech we have is a net negative, slowing things down and making things harder without providing value.
> I would be very worried if I created any opening (even on the very long term) for competition to justify itself being built up to meet demand
The interesting question is, why aren't other companies looking at this from the other direction?
For example, AMD now has a trillion dollar market cap. They make consumer and enterprise GPUs with integrated GDDR/HBM and have got to be looking at APUs that do the same going forward. They don't benefit from the price of their complement being high. They sold off their logic fabs at a point when they were half dead and had fallen behind, but you don't have to be TSMC to make DRAM. A large DRAM fab costs around $25 billion, i.e. 2.5% of their market cap. Why aren't they building two of them? Best case supply is still tight when it opens and they have a huge win, worst case the market has crashed by then but they still have a productive asset worth several billion dollars and paid for with cash from the time when it was highly available to them. And since they have internal demand, the amount the price would have to crash before the investment is a net loss is significantly more than it is for other companies.
Likewise Apple, Google, Amazon, etc. who have even more capital. What are these companies doing? They have the option to spend an amount of money they can easily afford to write off if it goes south, to get a huge win if supply continues to be tight.
Google and Amazon have tied up all their capital (and then some) in data centre investments, they don’t have more money to spend on fabrication. Plus they have a lot of experience running data centres, and zero experience running fabs.
Even if they hold the same opinion as Micron’s CEO, spinning up such a fab to reasonable yields would take a couple of years.
It’s a known dilemma in farming. Your crops are losing, so you decide to switch them to crops that are profitable, but so do your six neighbors and by the time the new crops start producing, you have flooded the market and everyone’s losing again.
Which is a huge problem for a farmer who redirected 100% of their capital to a new crop that turns out to be unprofitable. Whereas if you do that with 5% of your capital and it doesn't work out, you shrug and hardly notice. But if it does work out, that 5% is now worth 20%.
Market cap is not capital. AMD's cash-on-hand is $13b, which is considerable, and they could very well bring external funding to such a project on attractive terms, but it's not pittance. And for what, an historically low-margin commodity like DRAM? Let the boys play it out.
If anyone should do that, though, it's Apple. They have the money, the easy to predict need, the vertical integration in manufacturing. It could open all sorts of avenues for proprietary SoC to RAM interfaces. It makes a lot of sense for them to do.
Market cap is de facto capital because companies can issue their own shares.
> And for what, an historically low-margin commodity like DRAM?
How are they enjoying paying the "low" margins to the incumbents right now?
> They have the money, the easy to predict need, the vertical integration in manufacturing. It could open all sorts of avenues for proprietary SoC to RAM interfaces. It makes a lot of sense for them to do.
Which of those is supposed to not apply to AMD or Nvidia or, for that matter, Amazon or Google who are massive DRAM customers and also make the likes of Graviton and Axion?
And "proprietary interfaces" are essentially nonsense in this context. When a company comes up with a better way to connect DRAM to processors, the industry is more than happy to put it in the next version of the standard, which is what you want because it allows you to source the industry standard components in the event that your own facilities can't meet demand for any reason.
The primary reason companies use proprietary interfaces for things like this is when they're trying to capture the upgrade market, even then they're still typically using standard chips with a purposely incompatible connector, and Apple solders everything now anyway.
Nvidia and AMD do not fall in "easy to predict". Half, or more, of Nvidia's demand base is like, what, 5 hyperscalers? Which themselves are driven by OpenAI and Anthropic? It's not a sustainable market. Meanwhile Apple's business and demand are very easy to predict, with near constant moderate growth. Apple is also much more sensitive to margins than the hyperscalers, which is why Nvidia can get away with 75% margin.
That would be a fine argument for why Nvidia or AMD shouldn't build DRAM fabs to supply 100% of their present-day demand. What's the argument supposed to be for why they shouldn't build any at all? Build enough to satisfy 25% of their current demand and they would still need all of that even if the demand crashed by 75%.
$50B for Micron to build two fabs on the same site, i.e. they're still $25B each.
> They don't have in-house expertise to do that.
This is a questionable assertion to begin with. The skillset for designing logic and fabricating it are closely related enough that companies -- including AMD -- have traditionally done both, and companies like Intel and Samsung still do.
But more than that, what if they don't? You're not going to reassign your existing staff who are already doing necessary work anyway, you're going to hire new people.
This is a bit like complaining that toilet paper manufacturers did not dramatically scale up production in the first months of COVID-19. Had they invested in additional factories and production lines, they would be now holding the bag.
The memory market isn't all that different, except it's even more capital-intensive. Roughly half of HN is convinced that the current AI boom is unsustainable even though the technology itself is amazing. So why should memory makers take the bullet here?
Ultimately, as much as this sucks, having an inexpensive gaming rig is not a human right. People with money decided that all that money should go to AI, and we have to suffer through the downstream consequences of that.
I think neither CEOs nor boards of the public companies care. They want to maximize their paychecks and bonuses in the short term before they leave to other companies.
Point well taken. Same issue as with shareholders since it's not like companies really belong to a family anymore with a reputation on the line, like their name being in the company name.
I don't think the semiconductor business is so much like that - it's become more ingrained in them that you need to manage the business for long-term sustainability.
Except that in this case it's no longer 'collect money, move on to the next one'. It's 'collect money, become dynastically-rich, retire in 3 quarters from now'. After that, who cares.
This seems like a distinction without a difference. In the context of CEOs sacrificing the greater good for their bonus, [basically] nobody has done that for a $250k bonus. We're often talking 8 figures, tens of millions of dollars.
Is this really a categorically different thing if we're talking about $85 million vs. $12 million?
I've met dozens of ~$20M net worth people, they are just normal folks - in rich countries thats not even upper class money, thats "a nice home but thats it" kind of cash. $100M+ is "i can retire at will", while $500M are like a different species, surrounded by yes-men and usually quite damaged mentally from all the ass-kissing. So yeah, i think there is a meaningful distinction.
It's worse, sometimes even governments fail due to overconfidence.
Brazil and the UK (well, technically Japan) caused 2 natural rubber price spikes that first led to the UK smuggling rubber tree seeds to be imported to Malaysia, where the UK took over global natural rubber production, marginalizing Brazil.
Then Japan conquered the Malaysia during WW2, at which point everyone started developing artificial rubber and natural rubber was reduced to a much smaller market.
Micron has several fabs and fab upgrades coming online or under construction right now. An expansion in Boise and Virginia. Plus a huge megafab project in New York that's going to be the biggest in the world AFAIK.
All the DRAM makers have new fabs or plant expansions underway right now - indicating they believe the demand increase to be sustainable.
How many years ago did they start construction on these projects? How many years before that did they greenlight the project? I'm quite skeptical the timelines are short enough to line up with the AI boom.
That's the death trap of silicon boom-bust cycles that have claimed so many in the industry. Demand surges, you start expanding, the bust arrives, you go out of business.
Micron is putting up over $100 billion for this expansion wave IIRC. You can't afford that unless you are certain the increase in demand is sustainable.
You want to assure investors that investment will pay off so you want to talk up your order book and market outlook.
You also want to take as much profit as you can now so if things turn against you it doesn't mean bankruptcy. A pile of cash fixes many problems :)
How much of this issue is self-inflicted? Micron is building a massive fab in Syracuse. They've spent like 4 years seeking approval, lobbying politicians, doing BS "environmental impact studies" (which are mostly just rent seeking and political handouts). They finally started building, but if they had just been able to put shovel in the ground we would already have this fab be producing memory.
It took just over 3 years to get approved, and the reason it took so long is that it is massive project (1400 acres) and has enormous infrastructure requirements that needed comprehensive study by reviewing agencies to make sure the area could handle them.
The first two fabs there will require 17 million gallons perday of water and when fully operations in the early 2040s the facility will need 48 million gallons per day. That is more than existing infrastructure can handle so the project also includes a new pipeline extension to Lake Ontario.
That water requires extensive treatment before it can be discharged. It will be pretreated on site, then go to a new $1 billion industrial wastewater treatment plant being for this project, and then finally discharged into the Oneida river.
The energy requirements when fully operation are 1850 MW drawn continuously 24/7 every day of the year. That's 16 TWh a year.
For comparison that is a little above the annual electricity use of Ft. Worth, TX or Charlotte, NC, and little below Indianapolis, IN. It is about 14% under San Francisco, CA. A little over half of Chicago, IL, about 1/3 of Los Angeles, CA, and about 30% of NYC.
That's way beyond the existing grid capacity, so lots of infrastructure upgrades there are taking place.
If only we didn't have 100+ years of factories shitting where we eat we might not have thousands of governments wanting assurances that companies don't harm people in their area.
Silicon Valley is littered with SuperFund sites of bankrupted companies who conveniently couldn't afford to clean up the toxic mess they left behind when they became the vessel full of debt after the corporate shell game moved on.
Do these environmental impact studies effectively stop that? could they effectively stop this issue with a more efficient process?
What happens now is that there are a large amount of studies, that pay consultants and organizations to generate thousands of pages of documentation. They then turn around, have some niche issues that they pay to go away. For example was Micron having to purchase a few acres of land 25 miles away as a park for some owls effectively helping stop superfund sites?
This is basically environmentalism security theater and political grifting.
Maybe if they tried to build some of these on existing dilapidated commercial or industrial plots that are vacant and languishing they wouldn't have to jump through so many procedural and political hoops. I'm told these exist all over rural america.
Instead they want to build their 100 acre Data-Center/Technology Industrial Park adjacent to the burrowing owl habitats or up against Grandma Martha's backyard.
Building on vacant industrial land is more expensive because you have to clean it up first. It might even be a superfund site, no one is interested in actually finding out. There hasn't been any room on compromise or cleanup assistance in most cases.
There is also additional risk here. If company X made an unknown superfund site, then company Y develops on it. Is company Y then liable if the superfund site becomes known?
All this speculation is pointless. Superfund sites haven't had their cleanup funded by the companies doing the pollution since 1995, when congress decided that the tax on companies to fund such cleanup (which had funded 70% of the cleanup of superfund sites up to that point) no longer needed to exist, and since then the taxpaying public of you and me have been paying every single dollar for cleaning up the pollutants these companies emit.
You didn't have to pay to cleanup a site before using it, because the expectation was that you paid while running your business and the site would be cleaned before you wanted to utilize it.
But now the "Superfund" has been underfunded for like 20 years. Biden's IRA set up a new tax for ten years but, not sure how well that's going.
We literally had the system set up well and working fine, but it cost companies a few dollars so it had to go when the "Greed is good" folks came along.
Oil is sky high, a recession might be near. Also, optimizations (like the current KV one) can yield a drastic reduction in RAM needs. It's a risk.
If I was trying so stabilize (just stabilize) funding for humongously expensive fabs: I'd be giving markets reassurances that my business will continue to be high margins for the mid-future.
What would the alternative be? If you just kept prices lower, you'd have shortages. There would still be the same motivation—if not more!—to spin up competing production.
And to build up a large amount of new capacity yourself, you need to invest a massive amount of capital into something relatively risky.
It had mostly come down to 3 (Samsung, SK Hynix, Micron), but the Chinese manufacture's market share is growing rapidly, especially as they ramp up due to sanctions. CXMT is currently 10% of global DRAM, having seen previous YOY growth of 1%->4%->8%. YMTC is now global #3 in NAND.
It wouldn't be surprising to also see Japan make a comeback in memory - they are investing heavily (both privately & government) in semiconductor manufacturing.
For now, there are only 3 that matter, but the reason is purely because the US government has sabotaged their competition as soon as it became apparent that the incumbents would lose market share.
Without what the US government did and still does, the prices of memory would have never increased so much and all the world would not have been forced to pay so much for memory.
« Without what the US government did and still does, the prices of memory would have never increased so much and all the world would not have been forced to pay so much »
If that is true (and given the reasons you provide below, and the general quality of your commenting, I assume it is) then the situation is similar (if less militarized) to that of the oil sector, where the cost increase is likely higher by perhaps three orders of magnitude.
I assume that some Americans have downvoted what I have written, but this is a fact that is true beyond any reasonable doubt.
A few years ago, the US government has added the Chinese vendors of DRAM and of flash memory on their sanctioned entity list exactly at the time when Apple and other major computer manufacturers were evaluating the replacement of products from Micron and the like with the Chinese alternatives. This was prevented by the "sanctions".
There is absolutely no doubt that the so-called sanctions did not have any other purpose than removing the competition for Micron.
The official reason is absolutely ridiculous, i.e. that they are among the companies that provide products to the Chinese military.
If that would have been the reason, then 100% of all existing Chinese companies from any domain of activity would have to be put on the "sanctioned entity" list, because any of them would sell to the military of their country, if given an opportunity, like all companies from USA will provide products and services to the US military, given the opportunity.
While the Chinese memory makers will obviously also sell to the military, almost all their production goes into consumer products, like smartphones and laptops, which are the main products affected by the US "sanctions" (which are no true sanctions, because true sanctions against China would have been accompanied and conditioned by some political requests, which is not the case for any of the US "sanctions").
There is also no doubt that without the US "sanctions" the Chinese memory makers would have bought the latest semiconductor manufacturing equipment and they would have had the production capacity to flood the market with their products when Micron, Samsung and Hynix stopped providing enough memory for anyone who is not among their top 5 customers.
Thus their is no doubt that the US government has stolen a lot of money from my own pocket when I was forced to update a couple of old computers this year, by artificially restricting the amount of memory that is produced and preventing competition in this market.
As I see it, any US citizen who supports the actions of their government to restrict the production around the world of various kinds of goods and to prevent competition in various global markets, actions that have increased the prices of many kinds of consumer products in all countries, is an accomplice to theft from a very large number of people all over the world, who are forced to pay inflated prices for such products, resulting in money that goes from many countries into the pockets of the shareholders of a few big transnational companies, including Qualcomm, Micron, Samsung and Hynix.
> I assume that some Americans have downvoted what I have written, but this is a fact that is true beyond any reasonable doubt.
I would guess that the reason for any downvotes is that even if what you wrote is completely true it isn't really relevant, because the current problem is not lack of competition. The current problem is a huge increase in demand over a short time.
Even if there were no restrictions on China and they had latest generation fabs, they would not have had latest generation fabs just sitting around idle. They would be unable to meet the current increased demand for exactly the same reason everybody else can't.
If I can sell my things at a higher volume for longer vs higher price and less volume now, I'll go with the first, but not shareholders focused on quarterly numbers.
Is your claim that businesses that have to spend billions or tens of billions of dollars and years to bring facilities online to make the products they sell are owned (and influenced) by shareholders focused on quarterly numbers?
This “quarterly, short term” reddit-ism or internet-ism is baffling since the businesses that have earned the most money for shareholders (to the tune of trillions of dollars, eclipsing all other businesses), all have leadership making very expensive, very long term bets.
> Is your claim that businesses that have to spend billions or tens of billions of dollars and years to bring facilities online to make the products they sell are owned (and influenced) by shareholders focused on quarterly numbers?
Yes? That doesn't mean they can't spend billions on R&D. If Apple didn't sell iPods like hotcakes, they wouldn't be able to R&D the iPhone. That doesn't mean nobody at Apple had long-term vision, quite the opposite.
>Yes? That doesn't mean they can't spend billions on R&D
How can business leaders and shareholders spend money on long term R&D and still be accused of focusing on the short term and next quarter? If they were focusing on making the next few quarters look good, they would reduce expenses, which include R&D, and make profits higher for the next few quarters.
>If Apple didn't sell iPods like hotcakes, they wouldn't be able to R&D the iPhone. That doesn't mean nobody at Apple had long-term vision, quite the opposite.
I can’t parse these statements. Are you or are you not claiming businesses like Apple and Micron shareholders are focused on the short term?
>If I can sell my things at a higher volume for longer vs higher price and less volume now, I'll go with the first, but not shareholders focused on quarterly numbers.
What if a business is maxed out on production capacity? What if a business can sell things for a higher price now, and sell things at a lower price later?
Well it would depend how much higher prices and for how long until competition could arrive. Any business that could sell stuff at 10x prices for 10 years would make enough extra money that it would be worth it even if they knew bankruptcy would happen in year 11.
This is a very difficult supply chain in which to compete, even nation states mostly can’t. The machines that build the chips are really hard to come by and even China is just hoping they can start to make them in 2030 and they likely cannot. Meanwhile demand is accelerating.
History does not drop financial gifts like this in one industry’s lap often. From a sheer numbers standpoint, what they’re doing absolutely is the only logical choice.
Wouldn’t shortages produce basically the same incentive structure regarding competition or national security requirements? Like the demand for RAM is actually way up, so if they don’t raise prices enough the inventory will just quickly become depleted.
Whether they did this or not, I think the exponential demand for memory would still be there. So there would still be conditions that encourage new competing manufacturers.
He's not prioritizing high prices. Demand is exploding and supply can't keep up. Everybody in the world that can is already trying to ramp up production.
I honestly expected that there would be more competition entering the memory market by now. I know it takes a lot of time and capital, but Europe has plenty of semiconductor know-how with ASML and ARM, and a strong desire to be less dependent on other parts of the world, so investment in a European semiconductor industry would make a lot of sense to me. And these high memory prices sound like the perfect opportunity.
What are those construction times in China? Not just for nuclear, but also memory or semiconductor plants, probably less than what it is in the West.
The three-headed A memory cartel should enjoy their brief moment in the sun. Because for the five years it will be over, and when I say over, it’ll be over for good…
Or it just be that demand really does seriously exceed supply. It's not fast to bring supply online, it's costly and risky to do so if the demand will not be sustained for a long time. IMO, there's no conspiracy. I have been telling folks for the last 3 years that the price of AI compute is not coming down for a good 5+ years. Possibly 10 years too. Every human in the world will need AI, the same way we all got on the internet and then mobile. However, outside of humans, for every human there's going to be ten or hundreds of agents also all demanding compute.
> Or it just be that demand really does seriously exceed supply. It's not fast to bring supply online, it's costly and risky to do so if the demand will not be sustained for a long time
This is exactly it. There's a lot of conspiracy theory in this thread without realizing that it's just what happens when an insanely large amount of capital is deployed: market distortion.
It's not just RAM, it's things like gas turbines and power transformers. I've seen suggestions that it's even affecting housing starts because the construction effort is pulling in workers, especially electricians.
However unlike you I'm not of the opinion that all that can be built, in that short timeframe. Even less so that it can be operated without pushing us over the 2C global warming threshold.
What is the cost of AI tokens? You can't just calculate cost of inference. There is a cost to train, there is a cost for staff to build it all, the money investors provided for it all. When you run a business, all those costs must be recovered. Sure, it takes $2 in material cost to make a pizza, but what's the cost of pizza? You have to take into account, energy, rent, staff, insurance, and profit too. Tokens/$ is subsidized if you are talking about APIs
The big AI companies like OpenAI and Anthropic may have rushed into it too fast, and in effect wasted a lot of money on early model generations that never paid for themselves, early model architectures that were not cost-efficient, experiments that never panned out, etc.
The AI pure-play companies like OpenAI and Anthropic are presumably in worst financial position, since others like Google, Microsoft, Amazon and Meta (also Alibaba, Baidu, Bytedance, Tencent, Xiaomi) have generally been more conservative and are treating AI as an incremental revenue source not a highly leveraged all-in bet.
If someone enters as competitor because of high prices/profit, they'll want for them to stay that way. It won't make it better for the consumer if the supply is constricted in such a way.
I think this is a point where governments should intervene and perhaps start fabs that would produce RAM at cost for working class and SMEs to use.
> because I prioritized extremely high prices for a few years that might have the effect of changing the arithmetic for potential competitors that wouldn't have done it before.
Or maybe they believe the demand is not sustainable, and they don't want to fund $10B ghost fabs.
There will be new competitors. And there will be one or two companies that will take it (memory) in house and eliminate the memory companies out of the picture.
The three greedy memory companies are changing the equations in tech industry and it won’t be in their favor in the long term.
The counterpoint is that that new, and existing, mortgage originators did not fare so well after the mortgage market collapsed in the GFC. During large parts of the housing boom deep in the money mortgages could not be refinanced as quickly as borrowers would like due to capacity constraints at the originator level. Obv, not including subprime originators, where there was arguably 10X the needed capacity.
> extremely high prices for a few years that might have the effect of changing the arithmetic for potential competitors that wouldn't have done it before.
The big problem is, competitors need money. Say a fab costs 10 billion $ a piece, that is a big chunk of investment needed that absolutely needs to be paid back because even a big bank consortium will have a hard time shouldering this kind of single risk.
Now, the fab needs around three years to build... which means if there suddenly comes a huge amount of new supply or alternatively, the AI bubble pops, prices will crash, the companies won't be able to service the loans, and the banks will be sitting on a very expensive piece of real estate and machinery.
Particularly in semiconductors, this "bust and boom" cycle, "pork cycle" or however you want to call it, is very very well known. Banks and other investors have been through that one more often than you can count, that's the reason why there are so few producers left after all.
The only place I think can reasonably pull it off to start a competitor is China because their economic policy is focusing on decades in the future and the CCP is willing (and able) to shoulder the risk - however China can't acquire EUV machinery due to sanctions and will likely take a few years until they can pull it off on their own, by the time TSMC et al will already be at the next generation.
Not enough people are prepared if this isn't a bubble. I think he's betting on this being a bubble which would imply building more supply would be a mistake, because they'll end up with a factory and no more customers.
It's starting to look more and more like we will always need more memory.
This is abdicating an entire market in the hopes that your ability to manipulate it later, which has worked so far unfortunately, will hold. It is so obviously market manipulation here. Sure, prices would go up when this much server demand comes in but it is so profitable to make DDR5 that completely abandoning that market is stupid and only works if all players move in unison. It would just take one of the big three deciding to be the consumer maker to dominate the market and reap massive profits now and later when the server market cools and the other two try to come back. The fact that they are all shying away from such a ridiculously profitable sector is clear evidence of the collusion. They need jail time, again, but this time not allow those going to jail to become the CEO when they get out. They should also have fines that actually stop this behavior in the future. Fines so punitive that it actually brings into question their ability to continue as a company.
Micron is the only one to bail, and they are the smallest footprint. No one else has bailed.
Samsung recently shifted some capacity from HBM back to DDR5 modules, freeing up around 80,000 wafers/month, so DDR5 margins are currently good enough to compete with HBM for wafers, or maybe they are anticipating a turn, or both.
Maybe I'm missing something but this doesn't really make sense? Right now it's more profitable to sell to datacenters because they're less price sensitive. Consumers and consumer PC manufacturers don't really care who they're buying RAM from -- they'll just buy it for the cheapest price. When it's no longer as profitable to sell to datacenters, they'll just switch to selling for consumers and will then compete against each other. Why would any RAM manufacturer choose to handicap themselves by selling to consumers today when they can earn much more selling to datacenters?
Suppose someone came to my town and started offering to hire all the programmers for, say, twice the market rate. Would I really need to assume a conspiracy to understand why they all started quitting their current jobs? Occam's Razor would suggest they're probably just deciding to accept an offer for more money.
I mean, yes, but they are also right. They benefit from the market conditions, and contribute to the price increase, but it’s pretty clear there won’t be that much available shovels for a very long time
Of course, because there was a HOA that became financially so big (on paper), that it secured 80% of the shovel supply for the next two years. Now everyone is fighting over the remaining 20%.
Shouldn't this be affected by the diggers from the Far Lands recently telling everyone how to angle the backhoe blade to dig 470x as much earth in one swing?
CXMT has entered the market, but for now they cannot increase their production fast enough to influence significantly the global prices, due to the US "sanctions", which prevent them to buy modern fabrication equipment.
So they must use older equipment, which has low productivity for state-of-the-art memories. Because of the "sanctions" there are now Chinese companies that have progressed significantly in semiconductor manufacturing equipment, but the effect of this will also take some more time to manifest.
Nonetheless, if the 3 big memory producers will not be able to increase the availability of memory until 2028, then until then CXMT will grab with certainty a major part of the market (they have already grown quickly until 10% of the market, from a negligible percent previously).
The rest of the world won’t care. They will use whatever memory they can get, and if the Chinese are willing to supply it over the next 5 to 10 years, they end up on top…
The greed and shortsightedness from both the memory makers and their big customers trying to lock out competitors from memory purchases is palpable. The first act of a Shakespearean or Greek Tragedy is where the characters take the actions that lock in the circumstances that precede their downfall. And this sets up everything for the next act where CXMT has everything it needs to establish itself on the market with buyers that would perhaps would not have considered it in other circumstances.
I thought your price for 32g was you getting ripped off. Holy crap. The price is super high now. Now I am lamenting missing out on getting 128GB ram when it was $700, thinking it was too high. It is outrageously expensive now.
Everyone, and I mean everyone, is gated by ASML machines. They can't make nearly enough EUV machines to satisfy the demand. ASML has also not scaled their production nearly as much because they're quite far away from the AGI-pilled companies in the supply chain.
Well, CXMT is not gated by ASML machines because they can't buy them. They can raise their capacity because they have access to the domestically produced steppers. They are not as good as the ASML ones, and if the memory prices were not bonkers, they'd never make a profit using them, but as it is the low yields and high die sizes don't matter.
The problem is that you’re giving them an in into the market, and as time goes on, they will get better at making memory. That’s like saying you’re going to exclude the Chinese the space station. Look what happened. They now have their own space station?
They are. They are using ASML DUV machines for memory. Samsung, SK, and Micron have moved onto EUV machines for memory because they are better. So CXMT not only has to rely on ASML, they have to rely on a worse machine to make modern memory chips.
Depends on what kind of memory you are talking about.
For DRAM, CXMT were only 1% of the global supply 3 years ago, and are now already 10%. They are growing extremely rapidly.
If you are concerned about consumer memory - smartphones, laptops, desktops, then this is what CXMT are making. CXMT do make HBM too, for AI accelerators, but their volume there is held back by sanctions.
They're 10% of the market now, and they've shown the ability to scale up manufacturing rapidly. They certainly won't singlehandedly meet demand, but they might at least be able to temper the price growth quite a bit.
Yeah, if anything, this makes it more clear to me that all this wonky supply/demand stuff will correct itself sooner rather than later. I've been wanting to upgrade my workstation but the DDR5 RDIMM sticks are unbelievably expensive, last few weeks I've almost pulled the trigger anyway, now I think I'll continue wait at least for the beginning of next year.
Install an adblocker. I'm not even kidding; it's been my secret to keeping older hardware with less (and slower) memory usable since the late '00s, and it's served me well.
That and, if you're into gaming, thank your lucky stars that there are all sorts of games from 15 years ago that are every bit as much fun now as they were then.
Don't be too hard on the CEO. It will be unavoidable and he will be deeply sorry and humbled by their dedication throughout the years. Also, people will receive 2 months of benefits and FREE job search consulting (through an AI agent).
Anthropic is expected to spend 500B of money it doesnt have. Based on these commitments, data-centers have ordered equipment which includes memory. So, a lot of demand for memory is built on a speculative foundation.
> Anthropic is expected to spend 500B of money it doesnt have.
... yet.
That is why the big AI corps all went public this year or are trying to do so. They are valued large enough that they will be included in the popular indices (NASDAQ, S&P 500, ...) from the start, which means an awful lot of money will flow into them each month automatically from all the retirement savings.
And that unfortunately completely prevents the core function that stock markets should offer - that bad actors get punished. Can't get punished when there is so, so much "dumb money" floating around and entering the fray every month.
I absolutely loathe Nvidia at this point. They've completely abandoned the core consumer that built Jensen's empire. Why not offer pre-orders of 5090s and MSRP? Why has every single drop at Best Buy gone primarily to botters and scalpers? Why not give end users half a fighting chance like EVGA used to by letting end users getinline before Nvidia crushed them?
When the AI bubble bursts, Nvidia is going to be Enron.
Because it would immediately go to scalpers anyway. Like, say the 5090 - I'm seeing prices of almost 7000€ for that thing [1]. Assuming without AI it would be priced at 1500€ - I could make about 3 months of rent just by immediately selling the GPU to the next reseller who then peddles it to some AI training farm. And in this economy, oh hell yes I would.
> When the AI bubble bursts, Nvidia is going to be Enron.
Nah. NVIDIA was a money printer way before AI was a thing. Post-collapse they will still be a money printer, just at lower speed.
The ones who are going to get fucked over hard when the AI bubble pops is pensioners, the generation who is just entering the workforce (2007ff had an aftereffect in wage depression lasting for up to a decade), and everyone who took up investments to service the AI demand (utilities, construction companies and the likes).
>> Because it would immediately go to scalpers anyway.
No they wouldn't, tens of thousands of unique users would be in line with dozens or hundreds of scalpers, if they're all allowed to buy one every 180 days per payment method, the end users actually get a fighting chance. Today they have no chance.
And it's not just a few people. The sophisticated scalping operations have hundreds of straw buyers in their network. For items that generate far less profit than a 5090 currently does.
I've watched vans dropping of an endless stream of people in the parking lot of my local Costco to buy out all the cases of Pokemon cards within 20 minutes of the "random" drop they put on the floor. They go in, buy the max quantity of the item with cash provided by their keeper, bring it back to the van with a receipt and change, and get the $20-30 per item pay they were promised. There were 2 or 3 'legit' buyers out of hundreds of obvious straw buyers.
If the profit per item is $4-5k, expect this to go into turbo mode and be done in a marketplace/automated fashion and go viral. Just sign up to your favorite scalper's platform, submit your information, get a quote and a prepaid label and drop the box you got from nVidia off at USPS the next day. That "tens of thousands" of individuals will turn into "millions" really quick once it gets known there is rent money involved by clicking a few buttons and it goes viral on all the social media platforms.
The only way out from that is manufacturer enough cards to meet demand, and at that point the retail channel doesn't really matter.
Everyone I know who signed up for EVGA cards for the 30 series got one eventually, even if it took a couple months of waiting. I don't know anyone with a 5090 FE at MSRP.
Yeah, I will never buy an Nvidia GPU again. I assume many of their former customers feel the same way. When the bubble pops, I think Nvidia is going to feel the pain as they will have pushed many of their former customers into the arms of AMD.
Because the LLMs were trained on average software which doesn't give a shit about performance and they'll revert to the mean for sufficiently complicated projects. While they can help with searching for performance improvements, they also are liable to cheat if you aren't reviewing them for correctness of any algorithmic changes. Also you need to be responsible for teaching them the right benchmarks to use, since, again, performance optimization has been niche and not the standard.
That's the one silver lining. I've always kept a lean desktop of mostly terminal based programs, and I've heard one too many times "just buy more RAM" as a defense for resource intensive graphical programs using Electron or modern browsers in general.
I doubt it will change much, inference (not training) is very profitable and demand for inference is quite high. See: Mistral serving GLM on their own servers.
The catchup models are all basically distillations of the sota models. Without the next training cycle things are going to stagnate. And as all those companies you mentioned are all linked to OpenAI and Anthropic and relying on hosting deals and things they are all going to be in the ringer when things to go south.
So on the one hand you have all the sota model makers doing investor expectation management in saying wet need a slowdown for safety (may or may not be true, but also means they don’t spend on the next training cycle before IPO? Could be making their books look better too?) and on the other we have a sense that the models aren’t yet at the stopping place where we can just not train another cycle and use distillations of the current generation?
They are probably too big to fail already (and too friendly with the current US gov). And even if they fail, open-weight models will take over as the usefulness of the tech behind has been proven.
I expect them to sell everything if they go bust to recoup what little they can for investors. That would double whack hardware manufacturers are their final consumers are now both leaving the markets as buyers and entering as competitors.
Only last year Microsoft admitted to having GPUs in inventory that they couldn't power, due to a lack of electricity. So much of this production capacity could very well go into making memory chips for GPUs sitting in a warehouse.
The amount of money spend on the AI hype train is crazy. Meanwhile we're wasting fab time making chips that might never be powered on. It's absolutely insane that there are more money to be made on hardware for AI that may never be used, rather than producing a product that consumers and businesses need right now.
> Acer CEO says memory makers are hyping 2030 shortage fears to protect margins — PC prices set to decline by late 2027, cheaper Chinese capacity coming online delivers lower memory prices
One thing is certain: in that world, we would not have to brace for a future where pretty much any form of human labor is at risk of getting replaced by just having the machines oxidize even more carbon and telling those humans to just eat cake if they fail to sell any their skills for bread.
This is why I (actually/literally) just purchased a Commodore 64; I miss the days of crazy optimization and hardware hacking to fit so much into so little :).
Micron profited immensely by exiting the retail RAM / SSD markets. All the big manufacturers used Crucial RAM. Apple, Dell, HP, Lenovo, etc. They are the number one reason RAM is so expensive. They decided it was more important to make trillions selling to the A.I. hyped data center buildouts. They weren't wrong, they made serious money. I'm still rather pissed off about it.
As an end user it’s difficult to not be bitter about the situation, particularly when you have execs coming out and saying, “yep these prices sure are terrible, better buy now before they get even worse!”
I’ve seen people say, “well that’s just the free market, there’s a customer that’s willing to pay more than you,” but that argument doesn’t hold water when industry incumbents are actively obstructing entrance of new competitors into their primary markets. Micron, etc can’t both refuse to carve off more of their capacity for consumer-destined NAND and also refuse to allow others to fill the vacuum they’ve so conveniently created. Either you’re serving the market or you’re not; you don’t get to have your cake and eat it too under properly functioning, healthy capitalism.
I thought that the bulk of memory is already "technically bought" by the AI hyperscalers in the form of the big (not so) "hidden" debt? As in: the supply should already be known.
Micron reportedly sold $100bn worth of DRAM last financial year, and reportedly has $40bn long term commitments. $40bn is more than their entire fiscal 2025 revenue, so it's a lot, but it's clear it'll be a minority of their available supply.
Micron stock has gone up 10x in the past 2 years in a sector where the barrier to entry is lower than GPU/TPUs. CEO will do everything he can to keep stock price from returning to the mean.
One factor constraining supply for RAM in the short-to-medium term is that memory manufacturers is that DDR5 (and assumedly DDR6) capacity is being moved to the much more profitable HBM.
I think supply will be tight until at least the AI market crashes, which could take years or happen within months. Nobody can really predict when this will happen but it's fairly clear (IMHO) that the revenue projections needed to justify the investments by OpenAI, Anthropic and SpaceX simply don't exist.
Without the AI boom, the economy is in recession and has been for awhile. We're in an affordability crisis and the energy crunch is coming.
Kudos to the memory supply-industry on converting their market in a way that a handful of large-scale customers moved from mere supply-forecasting to outbidding themselves to BUY future production of chips for years to come.
I would now like to open the floor to investigations in this industry to determine whether this is hindering free-market forces and is ripe for regulation, as no-one is able to actually compete with these handful of companies anymore.
...anyone?
...ah yeah, right, I forgot. They also found a circular financing scheme which enslaved the entire global economy and organically prevents governments from properly governing their industry.
That’s not the point. You’re speeding up the progress of them becoming a competitor, a relentless competitor worldwide.
And you are now encouraging other large pocket tech companies inside and outside China to give thought to taking memory in-house if that means making and selling devices at a reasonable price in a timely manner.
That's today. You have to think about what the industry is going to look like in a decade.
Building a DRAM company from scratch is hard, expensive, and risky. You're burning a fortune to get started, and the result is a commodity product. But the AI bubble allows CXMT to have terrible yields, sell a legacy product, and still make a healthy profit! This is literally a once-in-a-lifetime opportunity to catch up.
And yes, right now that means selling DRAM for 10% below the crazy AI-inflated prices the incumbents charge. But sooner or later either demand will drop or supply will catch up. CXMT, like many other Chinese companies, is undoubtedly willing to operate at razor-thin margins to increase market share. Either the other manufacturers move down as well, or they'll be made irrelevant by CXMT stealing all their customers.
Absolutely. far too many people are thinking about the next 2 to 4 years. How many times must they see the Chinese play the long game before they figure it out?
Oh sure, they'll take profits while they can, but the supply increase will start pulling the market down. It's how price signals work.
Only way to keep prices this high would be a multinational cartel of RAM makers agreeing never to compete on price. One defector kills it. Not likely you could keep a cartel like this given that the players are all on different sides of geopolitical tensions. China's not going to want to help keep Taiwan, Japan, Germany, or the USA in the money.
Not gonna happen with China in the world, they will sell to every place else on the face of the planet this memory bump is temporary, also keeping it high means other companies will reconfigure what they do. if you want to continue to sell millions of devices at a reasonable price, you’re not going to sit on the sidelines and wait for the three-headed cartel.
You are going to get busy. and design and engineer around the cartel. Is there one company in the world that might do that? A company that has the engineering, design and the money with the capacity to execute over time?
If you follow the money though much of this future “demand” is on fulling orders from the AI labs and at the moment the AI labs don’t have the cash to pay that bill. A rather important detail in this whole mix.
So either the labs need to start printing money, now, or there’s going to be a whole bunch of excess chips hitting the market in 2027-28.
Someone in the middle is going to get left taking delivery of all these chips but unable to collect payment for them. And someone else is going to get left holding the bag on data center loans to house those chips. That’s the storm ahead.
Do you not believe that Anthropic and/or OpenAI are going to be able to IPO? Once they do, that's a huge source of capital for them that could absolutely be used to fund hardware purchases.
Based on current reporting the IPO wouldn’t generate anywhere near enough cash to pay for the bills coming due. Just between those two they have over a trillion $ of commitments coming due that need cash payment. Thats actual dollars they need to pay and neither the IPO nor current cash flows appear to be generating that kind of money to pay these bills when they come due.
It’s a big problem which neither big lab has a great answer for right now.
I guess those estimations based on projection for data Center build out, which include 30GW+ in 2027 alone, more then 3 previous years combined. Surely every human on earth will start token-maxing tomorrow and bring all of their money to their AI labs overlords to justify that explosion in spending, right?
Honestly, FF became similar worse:
On my 16 GB machine, when surfing some hours FF has allocated around 10-11 GB and for whatever reason it does not free them, unless I exit and restart it
On linux do something like `systemd-run --user --scope -p MemoryHigh=XG -p MemoryMax=X+2G firefox` to constrain it since the firefox people refuse to implement the memory limits logic that say Edge has..
But at the same time, this isn't just a Firefox issue, although they may be worse than most. Its a fundamental tenant of GC languages. You trade CPU (for the garbage collection) for RAM utilization. IIRC This was apparent by the late 70s with Dijkstra, but very clearly indicated in say Hertz where they are explicitly scaling GC overhead vs memory footprint to find that it takes ~5x the ram to approach the efficiency of an explicit allocate/deallocate model. Lower footprints are achievable but the GC overhead goes way up just to reach 2x the ram footprint such that the GC tends to dominate the CPU utilization.
And there are lots of papers like this for long running processes. GC works well if your model is basically allocate as much as you like, exit the process before actually needing to do a proper GC pass. Especially for contended systems where the memory is returned to the OS. This is why say PHP tended not to have this problem, because the memory was largely freed at the end of every request.
The New Tab Suspender plugin has my 50+ Firefox tabs at 2.2GB of ram usage right now. Having every tab ready and waiting for you is of course going to have increased memory usage.
Perhaps we should blame this on shitty website design. So many of them load huge assets for no good reason. They get away with it because modern processors are that good, bandwidth is cheap, and until now, RAM has also been cheap.
- the described behaviour is the one I experience on my Win10 notebook (Lenovo X1)
- Firefox wasnt always that, thats what Im refering to
- I can see in ProcessExplorer how many threads it starts & keeps, and for whatever reason this is always a lot, even if I open only 1 tab after coming from new start
There’s always going to be consumer products, even if the value proposition is worse.
If there’s a silver lining, I hope it leads to more optimisation, less Electron, and less bloat.
Typed from my 8GB MacBook Neo; which is honestly plenty fast enough. It “feels” like a 16GB windows laptop, even tho of course I know it’s not, but there’s something about optimisation.
Lots of more niche products than Apple’s laptop won’t exist for a while due to the price of hardware. At a smaller scale the inflation hardware companies experience is really, really massive
While I definitely agree that there will be a slow down in purchase decisions for both consumers and businesses I struggle with this idea that there will be “no consumer products” in the market. There will be.
You’re making what seems like a zero sum assumption. I am still using my M1 MacBook for work. I have a first version Mac Studio at home and countless other devices laying around. The world will continue to move forward and yes purchasing will slow down but folks will still be using their existing phones, laptops and other devices.
I don’t think Apple is gonna go home because of the three-headed memory cartel, since they didn’t spend/burn billions (one trillion?) of dollars on AI, I think they will replace the cartel in-house, Intel is out and Qualcomm is on the way out, the three headed cartel can follow them.
The future is starting to resemble smartphones that use predominantly RAM for local compute and inference while all persistent storage is managed by the cloud.
And People believing that everyone will have a robot at home.. until they realize that a human is much cheaper to maintain, a bread and a liter of water is enough...
Man whose compensation package is strongly tied to the economic performance of his company says company's economic performance will be great in 2027 and 2028 (so people should invest now to benefit from that bump)
I like to tell people that "AI" is the "last bubble." It was VR, the metaverse, crypto, NFTs, AR... AI is the Last Great Hype. There's really nothing beyond it. When this bubble pops it'll be remembered for decades to come because nothing will replace it.
The LLM bubble may fizzle rather than pop in any dramatic fashion, but it's certainly not the last word in AI - much more the exact opposite.
We've had AI "bubbles" / hype-fests before (expert systems, Japanese 5th generation systems), and I don't expect this will be the last.
There will also be technology bubbles and hype-fests in the future outside of AI - it's human nature. Hard to predict what they will be of course (the future is like that).
Until governments stop printing money and sending it directly to people that have nothing to do with it but investing, people will keep inventing random useless stuff to invest into.
Make no mistake here: they also plan to retain these high prices. It's like the mafia now. There should be instant-jail time to cure these CEOs of their greed. Like, go to prison for a week, to help refresh your mind and stop thinking in terms of milking the customers dry.
If/when the bubble pops demand will fall so hard they'll have little ability to keep prices too high. Especially with Chinese companies pushing out a lot more product by then.
Micron reported full year revenue of $133bn, up 256%, $100bn of which was DRAM, with non-GAAP earnings last quarter of their financial year of $37.7bn on $54.23bn of revenue. Seems pretty healthy to me. Any reason to think they'll go out of business?
But only in 2028 because all the wafers rotting away in the warehouses are HBM which are too expensive to package or integrate into consumer electronics with tighter margins. Looking forward to having enthousiast-grade GPUs with HBM again though.
I'm not a business guy, but I think I would be very worried if I created any opening (even on the very long term) for competition to justify itself being built up to meet demand (or national security requirements of other countries) on an industry I mostly control, because I prioritized extremely high prices for a few years that might have the effect of changing the arithmetic for potential competitors that wouldn't have done it before.
It makes me consider whether there would be retaliation in later years against the US intellectual property system in countries other than China. While RAM is a tangible product, the areas it serves aren't really.
If I understand this comment correctly, it's saying that Micron and the others would be better off selling RAM at below the price they could get, in order to keep competitors from deciding to enter the market.
Such collusion might well be a violation of anti-trust law, but in any case, it wouldn't work. If they decide to sell at below-market price, they must (by definition) be preventing some customers from buying by some non-price mechanism (eg, only selling to companies run by other members of the CEO's family). Potential competitors would then see an opportunity to sell to these excluded customers.
Not exactly, I think it's saying that Micron would be better off increasing supply themselves (and thus lowering prices) than waiting for a competitor to come and provide more supply to meet the higher demand.
If there was enough existing competition, that might be the right move, because if they don't increase capacity, a competitor probably would in order to gain more market share. But I think the barrier of entry for making RAM is high enough that Micron can probably get away with keeping supply low (and thus prices high) for at least a few years before a new competitor appears. Assuming demand stays high that long.
They aren't keeping supply low.
They have two massive DRAM fabs (56000 m^2 cleanroom size each) that they started construction on in 2022 and 2023 in Idaho, with production starting in 2027 and 2028.
The point they were stating was that if they kept up the high prices, that leaves room for a low priced competitor, likely from China, to occupy the lower end (and larger) share of the market. Like what BYD did to the EV market everywhere except the US. Or what Chinese mobiles did to the smartphone market. Eventually they become so powerful that they start going upmarket, trying to catch everyone in the higher end segments too. And by the time the Western companies think they can retaliate, too bad, China's got them by the bollocks on raw material.
But the reason that semiconductors are so vulnerable to boom/bust cycles (going back decades) is that takes a very long time to build the fab- you have nothing for years, and then the fab completes and enormous numbers of chips get made. So if you get demand wrong either direction you have problems. Either you flood the market or prices go way up.
I don't think it is possible to beat the fab construction delays. This isn't about things being slow in the US: TSMC took four years to go from site selection to some production at their Fab 22 plant in Taiwan (it is currently only 60% built out, is my understanding). And TSMC is best in the world at getting frontier (logic) semiconductor fabs built, it took them slightly longer (five years) for Fab 21 in Arizona. Logic fabs are not the same as memory fabs, but they are close enough to be a good comparison for timeline.
Basically, unless you started fab construction (specifically to make memory chips) in 2023 you cannot change production for 2027 and 2028. The only thing that will affect prices between now and 2028 is demand, production can't go up because the fabs to make those chips don't exist. If you started now you might get memory chips in 2031, and can you bet 40 billion dollars that the demand will still be there? (A current generation logic fab costs about 20 billion. By 2031 a new frontier logic fab will cost about 40 billion). I don't have prices for a current memory fab, but as I understand it they are comparable.)
Lowering their prices would result in supply running out sooner. Their competitors would still get the same bite at the apple but would have a longer window where they are the only option. Or resellers would just arbitrage the surplus left on the table.
Which are exactly the two years they are threatening tighter supply.
He's "threatening" tighter supply the same way a weather forecaster "threatens" heavy rain.
Production starts != maxing out capacity of the fab
He's not threatening. Demand is exploding and supply just can't keep up.
Yeah there is no such thing as selling a commodity below its market price on the open market. Someone else will come and arbitrage that away, or it’s not an open market.
Supply is severely limited and there is not really any competition making RAM chips, so it is kind of not a commodity and not a market.
But you are right someone will arbitrage the price, like scalpers selling concert tickets.
Well, that's blatantly false...
A well-known anticompetitive tactic is for a big player with a vast bank account to move into an area and eat losses while waiting for their competitors to starve. Well observed in the 90's movie rental market, so certainly real.
The statement further up should have qualified it as "scarce commodity". Then it is correct.
To successfully perform dumping, you have to have access to excess supply capable of disrupting that market.
Selling below market price != selling below cost, which is a prerequisite for it to be anticompetitive
You don’t lower your prices on the off chance a competitor might enter the market in the future. You wait until they have invested capital but before they start making profits and are at their most vulnerable.
I don't think that would work. They wouldn't demolish the competitor's factory when it goes bankrupt. Instead the factory, being an asset, gets sold for cheap, and another company gets a shot at competing while having much lower debts. This process could repeat a few times until the debt is low enough to profitability compete.
So if that's the expected outcome, it might be better to prevent competitors from building factories in the first place.
Also why anti-dumping laws exist
Do those solve the problem just outlined?
(Legitimately asking: I don't know the answer ... but based on most US business law my strong suspicion is ... no, they don't.)
No, because they're hardly enforced. China has been blatantly dumping a buttload of products for decades and the U.S. has been asleep at the wheel as it continues to hollow out our domestic manufacturing industries even further.
Silicon Valley and other venture backed enterprises also regularly dump into various markets without consequences.
Are you seriously not aware of the existing history of the memory cartel? https://en.wikipedia.org/wiki/DRAM_industry_price_fixing
There are nicer ways to phrase that.
I think misleading people using confident language about an issue you don't understand is more unkind.
Also I was truly asking if they weren't aware in disbelief, since I assumed it was common knowledge.
Why
Micron started as the outsider busting in during a past RAM crunch. They more than any manufacturer should understand this.
Does 87% gross margins sound like that them lowering prices is tantamount to dumping?
It also ignores the giant CAPEX and OPEX required for any competition. This isn't banana stands or book publishing, or whatever.
These are billion dollar+ funded operations that take multi-years just to see if you can compete at the levels required.
All these comments show up every time, like there's a memory faerie land where you just put on your memory faerie hat and walla! Memory!
So bizarre.
Everything is a conspiracy. The elites refuse to wave their increase-manufacturing-capacity-instantly wands.
I mean, we can assume the people who make the memory love getting huge prices for it. But beyond that, I just dont see why someone who invest several multi billion dollars to squeeze into a potential window that may or may not last 3 years. Especially with how volatile the whole AI bubble is.
> Especially with how volatile the whole AI bubble is.
There's nothing volatile about the AI bubble, it's been going on for a while regardless of cost and that will continue because of politics - which is rather transparent.
> I just don't see why someone who invest several multi billion dollars to squeeze into a potential window that may or may not last 3 years.
Interesting, Chinese companies don't seem to know this, they're investing bigly in RAM and other silicon despite being sanctioned and tariffed up the wazoo. Should we tell them that it's a crunch for no more that 3 years so they don't waste themselves for nothing?
The more interesting question is why Chinese RAM is hugely tariffed, if not for preserving the crunch and its monopoly pricing?
> I mean, we can assume the people who make the memory love getting huge prices for it.
Your mistake is to think of the RAM makers instead of the systemic reasons for the problem, these would have to include tariffs, sanctions, subsidies, etc that drive market pricing, RAM manufacturers and their potential competitors.
> Interesting, Chinese companies don't seem to know this, they're investing bigly in RAM and other silicon despite being sanctioned and tariffed up the wazoo.
It's also possible they're preparing for the possibility that they get cut-off and so profit is not the primary motivation. As you mention, they're already sanctioned...
Also, CMXT started in 2016 so that gives some indication of the timelines involved. Even in Taiwan it's like a year and a half just to get the building made and these are experts operating in their own country working 24/7. That timelines doesn't include equipment into it, which is also in shortage, and getting lines qualified.
https://www.tomshardware.com/tech-industry/building-a-chipma...
> Such collusion might well be a violation of anti-trust law
The Memory cartel do not care about silly things like laws
Yes, one of those markets with infinite demand where supply has no impact on price. Those ones.
> Yes, one of those markets with infinite demand where supply has no impact on price. Those ones.
In economics, "demand" doesn't mean "I want this", it means paid, funded demand. 100s of millions of "free" LLM users aren't funded demand, they're subsidized. This leads to front-running and scalping the silicon market at an unreal scale, it also monopolizes access to silicon and compute. Nothing happens by chance here.
The argument I've seen about why someone else doesn't start selling RAM is it takes a year or two to ramp production. But here's a CEO saying he's going to make sure prices stay high enough, long enough that you can spin up a competitor
> But here's a CEO saying he's going to make sure prices stay high enough, long enough that you can spin up a competitor
2 years is not remotely enough time to spin up a competitor. It's not even enough time for a current leading competitor to spin up another fab.
If you are a newcomer you are talking a decade or more before mass production. If you are HK Hynix perhaps 5 years or so.
The majors all have large fabs under construction, but the majority of new production from those facilities won't be coming fully on-line until 2028 at the earliest. Even from the date of first wafer shipped most of these facilities take 6-12mo to reach mass production level output.
Chinese DRAM is interesting only because they started building capacity (expertise) a decade ago. Only now are the expected to have any material impact on the total volume of the market. And if they want to gain even more market share, they will be under the same ~5 years to spin up a brand new fabrication plant. Perhaps a bit less because they are China, but still multiple years.
And it's important to point out how long these fabs have been under construction.
Micron started building it's new fab in 2022. They are expecting their first chips to be out by 2027 (maybe). The projected cost, $50B.
This is a fundamental problem which the free market cannot handle. It takes multiple years and billions for competitors to spin up. No sane bank or business person would try and get into cutting edge memory fabrication. That means the players we have are the players we'll likely always have (With MAYBE china entering the arena).
The free market only works when the required upfront capital for a new player to enter a business is relatively low. In industries where the players are static or shrinking, you have a free market failure.
correct, which is why industrial policy is important.
> 2 years is not remotely enough time to spin up a competitor. It's not even enough time for a current leading competitor to spin up another fab.
But it is enough time for customers to consider competitors they may have avoided before.
So maybe their priced-out customers give Chinese DRAM a serious look, make some compromises, and never go back to Micron.
Then there is risk; who is to say that the current demand will be around in five to ten years by the time you ramp up.
Imagine being the sucker to go and do a big capex build out for a hypothetical demand in two years. Have fun explaining that to the investors who supported you. Also you probably need 5 years to cover your expenses at whatever the current costs are. If the volatile prices collapse you are in a world of hurt.
Someone will certainly take this bet - but it is a high risk one.
> But here's a CEO saying he's going to make sure prices stay high enough, long enough that you can spin up a competitor
Where does he say he's going to make sure prices stay high?
"We will produce as many Oreo's as we can, but soon Oreo supply will be much tighter than ever before"
- Oreo CEO, suggesting you to buy every Oreo you can get
Micron is a trillion dollar company. The CEO would never say something like that directly. He's talking in the "business dialect" of English and you have to read between the lines a bit.
The CEO is saying nothing that everyone who analyzes the situation and knows at least something about semiconductor production wouldn't say. Demand is way up, supply cannot keep up, it will take some time to increase supply.
Micron is working on increasing that supply. The started building a massive fab for DRAM in 2022 in Idaho. It should start production in 2027. They started a second massive fab for DRAM next to it in 2023 with production scheduled for 2028.
This is the info that people should be looking at. Building these facilities takes time. I'm guessing they are thinking that demand is not going to decline anytime soon to justify this buildout. EDIT both SKHynix and Samsung are also expanding their DRAM production lines and new plants. So they are all working on adding new capacity.
Not to be snarky, but the CEO is saying that the supply will be tight for years. I know Econ 101 is over simplified but I do recall that when you reduce supply and demand stays constant, or increases, the price goes up.
I also recall some Econ 101, but I seem to recall a little but more than you since I also recall that the price goes up when the supply stays constant or even increases but the demand increases faster than the supply increases.
The CEO is saying that that is the situation we are in.
Demand has increased massively over a short time. Supply cannot increase as rapidly because for big demand increases it involves building new factories and equipment which takes significant time.
The CEO says that their new stuff will start coming online in 2028 but they can't forecast when supply will catch up.
> but I seem to recall a little but more than you
or less, as you seem to forget what this whole discussion is about.
If they dont make sure the prices dont go up, they go up.
Once theyre up, and expected to stay up for the foreseeable future, competitors see a gigantic opportunity. Suddenly, the large amount of investment necessary to the ball rolling is worth it.
Once this opportunity is taken advantage of, the original market position of the company being able to freely set whatever price they wanted is eroded forever -- and in this particular case: it opens them up to being out-competed by the new competitor(s).
In this case those will will likely be coming out of china and be state backed, so theyll likely operate at a loss for maximum damage to get as much of the market as they can get.
in 5 years, Micron is likely going to go cry for state support to socialize their upcoming losses.
and before you say "but there are other competitors already": you should be aware that theyve previously been convicted of running a cartel. And the way theyve been behaving is exactly how they were behaving back then. While unproven, it is more then just likely that theyre still price fixing (and have been for years)
I'm curious what you think they could do to make sure prices don't up?
dont increase the price of the product?
the price didnt just go up for consumers, they themselves have deeply dipped into the opportunity to maximize profits.
its so hilariously outlandish that eg SK Hynix has to pay out >$450k to union employees (per employee - not in total!) because they have a contract that a percentage of profits get shared as a bonus across the company. this payout is usually not particularly high.
Would you prefer the money goes to middlemen? When supply is constrained and the manufacturer doesn't price to match, all that happens is scalping. Even if the memory die manufacturers gave away their products for free it wouldn't budge consumer RAM prices in this market environment.
youre taking a very consumer oriented look, while we're explicitly talking about the perspective of the producer in this thread.
the consumer price barely matters for the perspective of the company evaluating wherever they can enter the market, because thats just a tiny fraction of it. And the largest piece of the market buys directly from the producer, SK Hynex, Samsung and Micron.
For consumers wanting to buy RAM sticks the difference is academic, i give you that. But for the capital that decides wherever to invest millions if not billions into getting a factory up and running - the price they can set for the b2b sales is the more important factor.
If it wasnt already proven to be basically whatever number they feel like setting... the opportunity would be a lot less obvious.
Large tech companies like cloudflare are spending developer hours lowering memory usage because they feel the impact too.
The idea that companies should have priority price controlled access to memory puts us even deeper into a situation where we can never afford local compute ever again.
For stuff I wrote that I run on my server, I have been converting from Clojure to Rust. Sometimes mechanically, often by hand.
I actually like Clojure more but even with GraalVM AOT it still tends to take significantly more memory than even naive Rust code (e.g. using Clone everywhere). If you’re looking willing to spend time abusing the borrow checker you can often get memory usage down to like 5% of what you’d get out of Java stuff.
The price is up because demand is way above supply (and will be for a long time because it takes years to expand supply to match the current level of demand).
If the manufacturers decided to keep the price constant that would NOT stop the price that most people pay from going up. It would just mean that the people who managed to buy from the manufacturer before the supply ran out would resell on the secondary market for a giant profit.
"would resell on the secondary market for a giant profit."
But you've no evidence that that "giant profit" would be more than the increase that the suppliers would apply.
Arbitrage allows the market to find an agreeable price; a middleman arbiter with a warehouse of RAM has an incentive to agree a price with a buyer because he's stuck with that inventory if he doesn't shift it, and he has costs related to financing it.
A supplier has no such incentive, they can just jack the price up and if sales drop off they run the machines at a lower rate, or sack employees. But sales aren't going to drop off, because the baseline for demand has been lifted
Once this "RAM Crisis" is over, what reason would manufacturers have to reduce prices?
They have to cover their fixed costs somehow. Collusion and cartels only get you so far. OPEC and Mexican drug cartels are still exposed to market pressures.
> dont increase the price of the product?
Literal empty shelves. And resellers on Ebay making bank.
Exactly. Demand won't be going down for quite some time, so prices will stay high unless supply goes up. So where is the new competition entering the market?
If there's no competition entering the market, wouldn't that be a clear sign that this is not a free market but effectively a cartel?
It takes 3-5 years to build a new fab for current generation DRAM.
Better to get started then, instead of remaining dependent on a cartel that is now learning that people will pay anything.
> year or two to ramp production
More like 5-10
What they need to do is all raise prices for 3 years, dump below cost for 4-6 months and repeat.
Congrats, you just guaranteed that the startup can easily lock in contracts for memory production in order hedge against volatility.
The Big 3 (Samsung, Hynix, Micron) have been playing this game for 30+ years. There used to be many more DRAM companies but most of them went bankrupt when DRAM was cheap because the big 3 will intentionally drive the price down to bankrupt the competition. Now that they have a tri-opoly, they're milking the profits. New competition will emerge but they don't have the capital of the big 3 so they likely won't survive the next DRAM winter.
You seem to be adding a lot of intentionality to price swings that seem expected when supply is relatively fixed while demand fluctuates. (Maybe you’re right, but it’s not the simpler explanation.)
The supply isn't fixed. A new plant takes 3-5 years to build and cost tens of billions. DRAM companies don't always know where the demand will be when the next plant completes. 2 years ago there was too much supply so DRAM price hit rock bottom. So DRAM companies held back on investing in new plants so here we are now.
The comment said "relatively fixed" and a charitable reading that "relatively" here means in the short run.
Exactly this. Unfortunately, this strategy has become standard practice in every area of the economy and it's a huge corrupting force.
They are not prioritizing extremely high prices. TSMC has a similar waiting list, so does ever other tool manufacturer. Memory capacity is being increased approximately as fast as it can be.
You can find public statements in the last 12 months from Samsung and Hynix about their aversion to scaling out production or increasing supply because they are concerned about profits.
Hard to know who to believe, on one side there are the absurdly high prices and statements from companies and on the other hand there are strangers posting on HN that it's actually all fine.
The memory business has always been extremely cyclical, and I doubt this time will be any different. The manufacturers have learnt this, and therefore to avoid boom followed by bust they try build for what they perceive to be longer-term more sustainable demand, not just knee-jerk increase demand to meet an immediate shortage.
The risk to any of the manufacturers, especially those with most market share, would seem to be if they are not all acting in sync. Micron's consumer brand, Crucial, had a great reputation, but now they have exited that business in favor or devoting all their capacity to currently higher margin products... will this come back to bite them later if they want to re-enter the consumer business?
What's different this time is that LLMs require orders of magnitude more memory than we could have ever used before. And I think it's getting obvious by now that inference will move to local compute, So everyone's laptops and phones would need a lot of memory.
The question the memory manufacturers have is, sure, we have all these orders out to the next three years... but will they actually be there to pay when the memory is delivered?
The memory manufacturers have been through bubbles before.
I don't deny that AI is "real", in that there is certainly something there, but I would hate to be the guy betting billions or even trillions of dollars that we're not in a demand bubble well in excess of what is justified by that tech at the moment and expanding capacity is a great idea. Of course, I would also hate to be the guy saying "no we shouldn't expand capacity" when it turns out that, yes, we should have, but the staggering profits being made in the meantime would cushion that blow pretty well.
> LLMs require orders of magnitude more memory than we could have ever used before
That's true. But will anyone have any money to pay for the LLMs? The AI market is currently being heavily subsidised by, ultimately, everyone else. But the economy in general is looking extremely dire at the moment, and that seems unlikely to last.
> the economy in general is looking extremely dire at the moment
Can you elaborate on this? What about the economy is looking dire?
Basically everyone I know who doesn't have a tech job is struggling for money / feels like the world is unaffordable for them atm. Those without jobs are struggling to get them. And those with jobs are scared to change even if they don't like them.
And from what I read in the media, a lot of the economic indicators are backing this up: discretionary spending is down, private debt levels are rising, etc. All signs that the slack in the system is disappearing.
That's what people thought about a year ago, but with the amazing strides that smaller models have taken over the past year, people are now starting to question that.
> And I think it's getting obvious by now that inference will move to local compute
From the bottom of my heart, I hope that's the future.
I'm not convinced that will be the case. The AI companies don't want you to have local control. They want you to subscribe to a service that they can change at any time.
I grit my teeth when I type this, but (god help us) I think Apple is maybe the best (least bad?) hope here. They are the only big player with the hardware chops and without a current vested interest in getting you addicted to monthly AI subscriptions. I'm not saying this is highly likely... I'm just saying that out of the current major players, they're the ones with the ability and motivation to move in this direction in the near future.
The next best hope is probably just nVidia or AMD catering to the consumer market once again, after the AI bubble bursts or the datacenter market reaches saturation.
The outside hope is that some startup makes a business out of burning open weight LLMs to silicon like ChatJimmy... although ChatJimmy was bought up by AMD AFAIK.
It doesn't really matter what the AI companies want, if they misjudge the market somebody will just start a competitor and take it.
The relevant questions are: 1) Where are the economies of scale in the technology stack? 2) What's "good enough" to consumers, and how does that stack up with the relevant computing power needed? 3) What are the transaction costs along various system boundaries?
I think that the biggest force keeping inference in large centralized services is simply that provides a better product for the average user who doesn't care about local control (and the average user doesn't care about local control; indeed, for most people it's a misfeature, as then they have to administer their own hardware). HN is full of nerds that want to own their own stack; they're willing to put up with a little loss of capability to run Qwen 3.8 locally. But from the folks I know that have tried it vs. Claude vs. Codex vs. Antigravity, the local models are still pretty weak compared to what you can get by paying for a service. Most people will just pay for the service until the performance becomes indistinguishable and the price becomes less.
>It doesn't really matter what the AI companies want, if they misjudge the market somebody will just start a competitor and take it.
The vast majority of people are uninterested in local vs non-local models (not just LLMs but ANNs in general), only the services the models are providing. Even if performant local models can actually be served on device quickly, if the frontier labs don't want to go that way, they simply can and there's little threat the market can be taken from them that way.
If you can serve local models, the 'advantage' is that they can be run on the users device without that incurring any cost to you thus allowing flexibilty in pricing, but if such models can be made then whatever cost would be incurred to labs running on their own servers would be a rounding error and also inconsequential to them. That is to say, they can also compete with you on pricing. So unless someone comes along with something the frontier labs can't figure out or replicate, then it doesn't matter. Open AI might be 'forced' to drum down prices but only in situations where costs have also been drummed down. In such scenarios, margins may even become better for them.
That is not true. Semiconductor industry saw a similar supply crunch during covid and spent heavily for capacity. And when the crunch tapered off, they were left holding the bag.
Micron would probably not even exist, or be a sticker brand for some Chinese conglomerate if not for this supply shock.
They are a lot more reticent this time around.
Nobody is holding any post-COVID bags right now. Every fab and factory is either running balls-out or waiting on something.
There's no point bringing this up.
Whenever anyone does, there are dozens of posts acting conspiratorial on HN.
It has honestly gotten frustrating to even read anything on the internet these days. It’s just conspiracies upon conspiracies all the time, and it’s constantly in your face.
If only they were good conspiracies with some weight to them, but no, they fall apart if you have a rudimentary understanding of the subject.
https://en.wikipedia.org/wiki/DRAM_industry_price_fixing
Which was caused by the exact problem they’re trying to avoid by not going on a spending spree to increase manufacturing capacity.
They had built way too much capacity during the 90s and were losing money on every chip because they couldn’t run the fabs at full capacity. That’s why they colluded to increase the price to save their collective hides.
A new competitor wouldn’t be able to enter the market and lower the prices without also losing money. That’s not the case today, the Chinese are trying hard to do just that.
It's easier to bear that there might be a conspiracy, than that this outcome is the result of chaos and stupidity and greed. Also it makes you think you are smarter than anyone else.
Fundamentally, it comes down to whether or not you think "AI/ML" can disappear tomorrow and the world would continue to run as normal.
From what I have seen, it cannot. It has become foundational the same way Internet 1.0 (eg. Email, Static HTML, Web Directories, SSL) became foundational for a significant portion of the global economy in the 1990s and 2000s.
All the big techs have accelerated zilch in actual features released or user-facing bugs fixed. All that seems to have accelerated is cybersecurity offense and defense.
And these are supposedly the most brilliant corporations around, so other companies are seeing even less gain.
I stick with my theory that LLMs are GPU-like, not CPU-like. They will rarely solve a problem like an adept, but they can solve 1 000 000 million low-hanging fruit problems within a second in a way that there was just not enough eyeballs for before. They can also brute force problems.
AL/ML isn't going away, but what that means for demand for various types of memory remains to be seen. LLM are clearly a commodity, and once they are "good enough" then it'll just be about price, meaning that smaller and less memory hungry models will win.
It's a bit like the Sun/etc workstation market vs PCs, and high end expensive PCs vs cheap ones. Once the cheap PCs became "good enough" then they naturally dominated.
Not only could the world run as normal without AI as we have it today, it would run better. The tech we have is a net negative, slowing things down and making things harder without providing value.
> I would be very worried if I created any opening (even on the very long term) for competition to justify itself being built up to meet demand
The interesting question is, why aren't other companies looking at this from the other direction?
For example, AMD now has a trillion dollar market cap. They make consumer and enterprise GPUs with integrated GDDR/HBM and have got to be looking at APUs that do the same going forward. They don't benefit from the price of their complement being high. They sold off their logic fabs at a point when they were half dead and had fallen behind, but you don't have to be TSMC to make DRAM. A large DRAM fab costs around $25 billion, i.e. 2.5% of their market cap. Why aren't they building two of them? Best case supply is still tight when it opens and they have a huge win, worst case the market has crashed by then but they still have a productive asset worth several billion dollars and paid for with cash from the time when it was highly available to them. And since they have internal demand, the amount the price would have to crash before the investment is a net loss is significantly more than it is for other companies.
Likewise Apple, Google, Amazon, etc. who have even more capital. What are these companies doing? They have the option to spend an amount of money they can easily afford to write off if it goes south, to get a huge win if supply continues to be tight.
Google and Amazon have tied up all their capital (and then some) in data centre investments, they don’t have more money to spend on fabrication. Plus they have a lot of experience running data centres, and zero experience running fabs.
Google has a $4T market cap, Amazon >$2.5T. How hard would it be for them to raise $50B by selling new shares?
Even if they hold the same opinion as Micron’s CEO, spinning up such a fab to reasonable yields would take a couple of years.
It’s a known dilemma in farming. Your crops are losing, so you decide to switch them to crops that are profitable, but so do your six neighbors and by the time the new crops start producing, you have flooded the market and everyone’s losing again.
Which is a huge problem for a farmer who redirected 100% of their capital to a new crop that turns out to be unprofitable. Whereas if you do that with 5% of your capital and it doesn't work out, you shrug and hardly notice. But if it does work out, that 5% is now worth 20%.
Market cap is not capital. AMD's cash-on-hand is $13b, which is considerable, and they could very well bring external funding to such a project on attractive terms, but it's not pittance. And for what, an historically low-margin commodity like DRAM? Let the boys play it out.
If anyone should do that, though, it's Apple. They have the money, the easy to predict need, the vertical integration in manufacturing. It could open all sorts of avenues for proprietary SoC to RAM interfaces. It makes a lot of sense for them to do.
> Market cap is not capital.
Market cap is de facto capital because companies can issue their own shares.
> And for what, an historically low-margin commodity like DRAM?
How are they enjoying paying the "low" margins to the incumbents right now?
> They have the money, the easy to predict need, the vertical integration in manufacturing. It could open all sorts of avenues for proprietary SoC to RAM interfaces. It makes a lot of sense for them to do.
Which of those is supposed to not apply to AMD or Nvidia or, for that matter, Amazon or Google who are massive DRAM customers and also make the likes of Graviton and Axion?
And "proprietary interfaces" are essentially nonsense in this context. When a company comes up with a better way to connect DRAM to processors, the industry is more than happy to put it in the next version of the standard, which is what you want because it allows you to source the industry standard components in the event that your own facilities can't meet demand for any reason.
The primary reason companies use proprietary interfaces for things like this is when they're trying to capture the upgrade market, even then they're still typically using standard chips with a purposely incompatible connector, and Apple solders everything now anyway.
Nvidia and AMD do not fall in "easy to predict". Half, or more, of Nvidia's demand base is like, what, 5 hyperscalers? Which themselves are driven by OpenAI and Anthropic? It's not a sustainable market. Meanwhile Apple's business and demand are very easy to predict, with near constant moderate growth. Apple is also much more sensitive to margins than the hyperscalers, which is why Nvidia can get away with 75% margin.
That would be a fine argument for why Nvidia or AMD shouldn't build DRAM fabs to supply 100% of their present-day demand. What's the argument supposed to be for why they shouldn't build any at all? Build enough to satisfy 25% of their current demand and they would still need all of that even if the demand crashed by 75%.
> A large DRAM fab costs around $25 billion
$50B for Micron.
> Why aren't they building two of them?
They don't have in-house expertise to do that.
> $50B for Micron.
$50B for Micron to build two fabs on the same site, i.e. they're still $25B each.
> They don't have in-house expertise to do that.
This is a questionable assertion to begin with. The skillset for designing logic and fabricating it are closely related enough that companies -- including AMD -- have traditionally done both, and companies like Intel and Samsung still do.
But more than that, what if they don't? You're not going to reassign your existing staff who are already doing necessary work anyway, you're going to hire new people.
This is a bit like complaining that toilet paper manufacturers did not dramatically scale up production in the first months of COVID-19. Had they invested in additional factories and production lines, they would be now holding the bag.
The memory market isn't all that different, except it's even more capital-intensive. Roughly half of HN is convinced that the current AI boom is unsustainable even though the technology itself is amazing. So why should memory makers take the bullet here?
Ultimately, as much as this sucks, having an inexpensive gaming rig is not a human right. People with money decided that all that money should go to AI, and we have to suffer through the downstream consequences of that.
I think neither CEOs nor boards of the public companies care. They want to maximize their paychecks and bonuses in the short term before they leave to other companies.
Like termites, moving on to the next tree. One day we'll figure out how to be rid of these parasites, then we'll have healthier trees.
Say the creatures that depend on termites for habitat.
But didn't in generations past and therefore do not need to now.
What eats CEO's?
The SEC?
When
Point well taken. Same issue as with shareholders since it's not like companies really belong to a family anymore with a reputation on the line, like their name being in the company name.
I don't think the semiconductor business is so much like that - it's become more ingrained in them that you need to manage the business for long-term sustainability.
Yeah CAPEX for semiconductors is terrifying. They cannot just handwave away large fractions of a decade, especially with the lead time for equipment.
Except that in this case it's no longer 'collect money, move on to the next one'. It's 'collect money, become dynastically-rich, retire in 3 quarters from now'. After that, who cares.
This seems like a distinction without a difference. In the context of CEOs sacrificing the greater good for their bonus, [basically] nobody has done that for a $250k bonus. We're often talking 8 figures, tens of millions of dollars.
Is this really a categorically different thing if we're talking about $85 million vs. $12 million?
I've met dozens of ~$20M net worth people, they are just normal folks - in rich countries thats not even upper class money, thats "a nice home but thats it" kind of cash. $100M+ is "i can retire at will", while $500M are like a different species, surrounded by yes-men and usually quite damaged mentally from all the ass-kissing. So yeah, i think there is a meaningful distinction.
Stop smoking the stash
It's worse, sometimes even governments fail due to overconfidence.
Brazil and the UK (well, technically Japan) caused 2 natural rubber price spikes that first led to the UK smuggling rubber tree seeds to be imported to Malaysia, where the UK took over global natural rubber production, marginalizing Brazil.
Then Japan conquered the Malaysia during WW2, at which point everyone started developing artificial rubber and natural rubber was reduced to a much smaller market.
> They want to maximize their paychecks and bonuses in the short term before they leave to other companies.
You could say that about pretty much any employee - and you'd be completely wrong. What makes you right in the case of CEOs and boards?
> You could say that about pretty much any employee - and you'd be completely wrong. What makes you right in the case of CEOs and boards?
https://www.cnbc.com/2026/04/30/us-ceo-pay-grew-20-times-fas...
Micron has several fabs and fab upgrades coming online or under construction right now. An expansion in Boise and Virginia. Plus a huge megafab project in New York that's going to be the biggest in the world AFAIK.
All the DRAM makers have new fabs or plant expansions underway right now - indicating they believe the demand increase to be sustainable.
How many years ago did they start construction on these projects? How many years before that did they greenlight the project? I'm quite skeptical the timelines are short enough to line up with the AI boom.
The Boise fab was a result of the CHIPS act (so tech sovereignty rather than because of AI), and they recently tacked on an R&D expansion to that due to the AI boom. https://boisedev.com/news/2026/08/20/micron-to-build-10-bill...
That's the death trap of silicon boom-bust cycles that have claimed so many in the industry. Demand surges, you start expanding, the bust arrives, you go out of business.
Micron is putting up over $100 billion for this expansion wave IIRC. You can't afford that unless you are certain the increase in demand is sustainable.
You want to assure investors that investment will pay off so you want to talk up your order book and market outlook.
You also want to take as much profit as you can now so if things turn against you it doesn't mean bankruptcy. A pile of cash fixes many problems :)
How much of this issue is self-inflicted? Micron is building a massive fab in Syracuse. They've spent like 4 years seeking approval, lobbying politicians, doing BS "environmental impact studies" (which are mostly just rent seeking and political handouts). They finally started building, but if they had just been able to put shovel in the ground we would already have this fab be producing memory.
It took just over 3 years to get approved, and the reason it took so long is that it is massive project (1400 acres) and has enormous infrastructure requirements that needed comprehensive study by reviewing agencies to make sure the area could handle them.
The first two fabs there will require 17 million gallons per day of water and when fully operations in the early 2040s the facility will need 48 million gallons per day. That is more than existing infrastructure can handle so the project also includes a new pipeline extension to Lake Ontario.
That water requires extensive treatment before it can be discharged. It will be pretreated on site, then go to a new $1 billion industrial wastewater treatment plant being for this project, and then finally discharged into the Oneida river.
The energy requirements when fully operation are 1850 MW drawn continuously 24/7 every day of the year. That's 16 TWh a year.
For comparison that is a little above the annual electricity use of Ft. Worth, TX or Charlotte, NC, and little below Indianapolis, IN. It is about 14% under San Francisco, CA. A little over half of Chicago, IL, about 1/3 of Los Angeles, CA, and about 30% of NYC.
That's way beyond the existing grid capacity, so lots of infrastructure upgrades there are taking place.
If only we didn't have 100+ years of factories shitting where we eat we might not have thousands of governments wanting assurances that companies don't harm people in their area.
Silicon Valley is littered with SuperFund sites of bankrupted companies who conveniently couldn't afford to clean up the toxic mess they left behind when they became the vessel full of debt after the corporate shell game moved on.
Do these environmental impact studies effectively stop that? could they effectively stop this issue with a more efficient process?
What happens now is that there are a large amount of studies, that pay consultants and organizations to generate thousands of pages of documentation. They then turn around, have some niche issues that they pay to go away. For example was Micron having to purchase a few acres of land 25 miles away as a park for some owls effectively helping stop superfund sites?
This is basically environmentalism security theater and political grifting.
Maybe if they tried to build some of these on existing dilapidated commercial or industrial plots that are vacant and languishing they wouldn't have to jump through so many procedural and political hoops. I'm told these exist all over rural america.
Instead they want to build their 100 acre Data-Center/Technology Industrial Park adjacent to the burrowing owl habitats or up against Grandma Martha's backyard.
Building on vacant industrial land is more expensive because you have to clean it up first. It might even be a superfund site, no one is interested in actually finding out. There hasn't been any room on compromise or cleanup assistance in most cases.
There is also additional risk here. If company X made an unknown superfund site, then company Y develops on it. Is company Y then liable if the superfund site becomes known?
All this speculation is pointless. Superfund sites haven't had their cleanup funded by the companies doing the pollution since 1995, when congress decided that the tax on companies to fund such cleanup (which had funded 70% of the cleanup of superfund sites up to that point) no longer needed to exist, and since then the taxpaying public of you and me have been paying every single dollar for cleaning up the pollutants these companies emit.
You didn't have to pay to cleanup a site before using it, because the expectation was that you paid while running your business and the site would be cleaned before you wanted to utilize it.
But now the "Superfund" has been underfunded for like 20 years. Biden's IRA set up a new tax for ten years but, not sure how well that's going.
We literally had the system set up well and working fine, but it cost companies a few dollars so it had to go when the "Greed is good" folks came along.
Oil is sky high, a recession might be near. Also, optimizations (like the current KV one) can yield a drastic reduction in RAM needs. It's a risk.
If I was trying so stabilize (just stabilize) funding for humongously expensive fabs: I'd be giving markets reassurances that my business will continue to be high margins for the mid-future.
What would the alternative be? If you just kept prices lower, you'd have shortages. There would still be the same motivation—if not more!—to spin up competing production.
And to build up a large amount of new capacity yourself, you need to invest a massive amount of capital into something relatively risky.
This reminds me of an economist joke.
3 businessmen are in jail, they ask each other what they are there for.
Businessman 1: "I made my prices higher than my competitors and was jailed for price gouging!"
Businessman 2: "I made my prices lower than my competitors and was jailed for price dumping!"
Businessman 3: "I made my prices the same as my competitors and was jailed for price collusion!"
There is 3 companies that make memory - they ALL want to keep memory prices high for as long as possible.
It had mostly come down to 3 (Samsung, SK Hynix, Micron), but the Chinese manufacture's market share is growing rapidly, especially as they ramp up due to sanctions. CXMT is currently 10% of global DRAM, having seen previous YOY growth of 1%->4%->8%. YMTC is now global #3 in NAND.
It wouldn't be surprising to also see Japan make a comeback in memory - they are investing heavily (both privately & government) in semiconductor manufacturing.
For now, there are only 3 that matter, but the reason is purely because the US government has sabotaged their competition as soon as it became apparent that the incumbents would lose market share.
Without what the US government did and still does, the prices of memory would have never increased so much and all the world would not have been forced to pay so much for memory.
« Without what the US government did and still does, the prices of memory would have never increased so much and all the world would not have been forced to pay so much »
If that is true (and given the reasons you provide below, and the general quality of your commenting, I assume it is) then the situation is similar (if less militarized) to that of the oil sector, where the cost increase is likely higher by perhaps three orders of magnitude.
I assume that some Americans have downvoted what I have written, but this is a fact that is true beyond any reasonable doubt.
A few years ago, the US government has added the Chinese vendors of DRAM and of flash memory on their sanctioned entity list exactly at the time when Apple and other major computer manufacturers were evaluating the replacement of products from Micron and the like with the Chinese alternatives. This was prevented by the "sanctions".
There is absolutely no doubt that the so-called sanctions did not have any other purpose than removing the competition for Micron.
The official reason is absolutely ridiculous, i.e. that they are among the companies that provide products to the Chinese military.
If that would have been the reason, then 100% of all existing Chinese companies from any domain of activity would have to be put on the "sanctioned entity" list, because any of them would sell to the military of their country, if given an opportunity, like all companies from USA will provide products and services to the US military, given the opportunity.
While the Chinese memory makers will obviously also sell to the military, almost all their production goes into consumer products, like smartphones and laptops, which are the main products affected by the US "sanctions" (which are no true sanctions, because true sanctions against China would have been accompanied and conditioned by some political requests, which is not the case for any of the US "sanctions").
There is also no doubt that without the US "sanctions" the Chinese memory makers would have bought the latest semiconductor manufacturing equipment and they would have had the production capacity to flood the market with their products when Micron, Samsung and Hynix stopped providing enough memory for anyone who is not among their top 5 customers.
Thus their is no doubt that the US government has stolen a lot of money from my own pocket when I was forced to update a couple of old computers this year, by artificially restricting the amount of memory that is produced and preventing competition in this market.
As I see it, any US citizen who supports the actions of their government to restrict the production around the world of various kinds of goods and to prevent competition in various global markets, actions that have increased the prices of many kinds of consumer products in all countries, is an accomplice to theft from a very large number of people all over the world, who are forced to pay inflated prices for such products, resulting in money that goes from many countries into the pockets of the shareholders of a few big transnational companies, including Qualcomm, Micron, Samsung and Hynix.
> I assume that some Americans have downvoted what I have written, but this is a fact that is true beyond any reasonable doubt.
I would guess that the reason for any downvotes is that even if what you wrote is completely true it isn't really relevant, because the current problem is not lack of competition. The current problem is a huge increase in demand over a short time.
Even if there were no restrictions on China and they had latest generation fabs, they would not have had latest generation fabs just sitting around idle. They would be unable to meet the current increased demand for exactly the same reason everybody else can't.
>There is 3 companies
There are 3 companies
There be 3 companies
(Who said that we can only talk like a pirate one day a year?)
Thanks for you're help
I also like to keep prices high for as long as possible for the things I sell. I bet most people I ask would answer the same.
If I can sell my things at a higher volume for longer vs higher price and less volume now, I'll go with the first, but not shareholders focused on quarterly numbers.
Is your claim that businesses that have to spend billions or tens of billions of dollars and years to bring facilities online to make the products they sell are owned (and influenced) by shareholders focused on quarterly numbers?
This “quarterly, short term” reddit-ism or internet-ism is baffling since the businesses that have earned the most money for shareholders (to the tune of trillions of dollars, eclipsing all other businesses), all have leadership making very expensive, very long term bets.
> Is your claim that businesses that have to spend billions or tens of billions of dollars and years to bring facilities online to make the products they sell are owned (and influenced) by shareholders focused on quarterly numbers?
Yes? That doesn't mean they can't spend billions on R&D. If Apple didn't sell iPods like hotcakes, they wouldn't be able to R&D the iPhone. That doesn't mean nobody at Apple had long-term vision, quite the opposite.
>Yes? That doesn't mean they can't spend billions on R&D
How can business leaders and shareholders spend money on long term R&D and still be accused of focusing on the short term and next quarter? If they were focusing on making the next few quarters look good, they would reduce expenses, which include R&D, and make profits higher for the next few quarters.
>If Apple didn't sell iPods like hotcakes, they wouldn't be able to R&D the iPhone. That doesn't mean nobody at Apple had long-term vision, quite the opposite.
I can’t parse these statements. Are you or are you not claiming businesses like Apple and Micron shareholders are focused on the short term?
>If I can sell my things at a higher volume for longer vs higher price and less volume now, I'll go with the first, but not shareholders focused on quarterly numbers.
What if a business is maxed out on production capacity? What if a business can sell things for a higher price now, and sell things at a lower price later?
Well it would depend how much higher prices and for how long until competition could arrive. Any business that could sell stuff at 10x prices for 10 years would make enough extra money that it would be worth it even if they knew bankruptcy would happen in year 11.
This is a very difficult supply chain in which to compete, even nation states mostly can’t. The machines that build the chips are really hard to come by and even China is just hoping they can start to make them in 2030 and they likely cannot. Meanwhile demand is accelerating.
History does not drop financial gifts like this in one industry’s lap often. From a sheer numbers standpoint, what they’re doing absolutely is the only logical choice.
> prioritized extremely high prices
Wouldn’t shortages produce basically the same incentive structure regarding competition or national security requirements? Like the demand for RAM is actually way up, so if they don’t raise prices enough the inventory will just quickly become depleted.
Whether they did this or not, I think the exponential demand for memory would still be there. So there would still be conditions that encourage new competing manufacturers.
Building capacity with looming bust around the corner has its own drawbacks. Damned if you do damned if you dont situation for them.
He's not prioritizing high prices. Demand is exploding and supply can't keep up. Everybody in the world that can is already trying to ramp up production.
I honestly expected that there would be more competition entering the memory market by now. I know it takes a lot of time and capital, but Europe has plenty of semiconductor know-how with ASML and ARM, and a strong desire to be less dependent on other parts of the world, so investment in a European semiconductor industry would make a lot of sense to me. And these high memory prices sound like the perfect opportunity.
It takes too long! It's like all the people going "why don't we just build nuclear", without looking at the construction times of recent plants.
What are those construction times in China? Not just for nuclear, but also memory or semiconductor plants, probably less than what it is in the West.
The three-headed A memory cartel should enjoy their brief moment in the sun. Because for the five years it will be over, and when I say over, it’ll be over for good…
First line of your post was enough.
Or it just be that demand really does seriously exceed supply. It's not fast to bring supply online, it's costly and risky to do so if the demand will not be sustained for a long time. IMO, there's no conspiracy. I have been telling folks for the last 3 years that the price of AI compute is not coming down for a good 5+ years. Possibly 10 years too. Every human in the world will need AI, the same way we all got on the internet and then mobile. However, outside of humans, for every human there's going to be ten or hundreds of agents also all demanding compute.
> Or it just be that demand really does seriously exceed supply. It's not fast to bring supply online, it's costly and risky to do so if the demand will not be sustained for a long time
This is exactly it. There's a lot of conspiracy theory in this thread without realizing that it's just what happens when an insanely large amount of capital is deployed: market distortion.
It's not just RAM, it's things like gas turbines and power transformers. I've seen suggestions that it's even affecting housing starts because the construction effort is pulling in workers, especially electricians.
However unlike you I'm not of the opinion that all that can be built, in that short timeframe. Even less so that it can be operated without pushing us over the 2C global warming threshold.
Aren't we already at 2.5 C?
> I have been telling folks for the last 3 years that the price of AI compute
What does that even mean ?
The cost of AI itself (tokens/$ for a given level of intelligence) has been falling rapidly, and shows no sign of stopping.
What is the cost of AI tokens? You can't just calculate cost of inference. There is a cost to train, there is a cost for staff to build it all, the money investors provided for it all. When you run a business, all those costs must be recovered. Sure, it takes $2 in material cost to make a pizza, but what's the cost of pizza? You have to take into account, energy, rent, staff, insurance, and profit too. Tokens/$ is subsidized if you are talking about APIs
The big AI companies like OpenAI and Anthropic may have rushed into it too fast, and in effect wasted a lot of money on early model generations that never paid for themselves, early model architectures that were not cost-efficient, experiments that never panned out, etc.
The AI pure-play companies like OpenAI and Anthropic are presumably in worst financial position, since others like Google, Microsoft, Amazon and Meta (also Alibaba, Baidu, Bytedance, Tencent, Xiaomi) have generally been more conservative and are treating AI as an incremental revenue source not a highly leveraged all-in bet.
they operate at a loss (they pretend to profit by excluding key costs in financial reports). they won't do it forever.
If someone enters as competitor because of high prices/profit, they'll want for them to stay that way. It won't make it better for the consumer if the supply is constricted in such a way.
I think this is a point where governments should intervene and perhaps start fabs that would produce RAM at cost for working class and SMEs to use.
> because I prioritized extremely high prices for a few years that might have the effect of changing the arithmetic for potential competitors that wouldn't have done it before.
Or maybe they believe the demand is not sustainable, and they don't want to fund $10B ghost fabs.
There will be new competitors. And there will be one or two companies that will take it (memory) in house and eliminate the memory companies out of the picture.
The three greedy memory companies are changing the equations in tech industry and it won’t be in their favor in the long term.
The counterpoint is that that new, and existing, mortgage originators did not fare so well after the mortgage market collapsed in the GFC. During large parts of the housing boom deep in the money mortgages could not be refinanced as quickly as borrowers would like due to capacity constraints at the originator level. Obv, not including subprime originators, where there was arguably 10X the needed capacity.
> extremely high prices for a few years that might have the effect of changing the arithmetic for potential competitors that wouldn't have done it before.
The big problem is, competitors need money. Say a fab costs 10 billion $ a piece, that is a big chunk of investment needed that absolutely needs to be paid back because even a big bank consortium will have a hard time shouldering this kind of single risk.
Now, the fab needs around three years to build... which means if there suddenly comes a huge amount of new supply or alternatively, the AI bubble pops, prices will crash, the companies won't be able to service the loans, and the banks will be sitting on a very expensive piece of real estate and machinery.
Particularly in semiconductors, this "bust and boom" cycle, "pork cycle" or however you want to call it, is very very well known. Banks and other investors have been through that one more often than you can count, that's the reason why there are so few producers left after all.
The only place I think can reasonably pull it off to start a competitor is China because their economic policy is focusing on decades in the future and the CCP is willing (and able) to shoulder the risk - however China can't acquire EUV machinery due to sanctions and will likely take a few years until they can pull it off on their own, by the time TSMC et al will already be at the next generation.
Not enough people are prepared if this isn't a bubble. I think he's betting on this being a bubble which would imply building more supply would be a mistake, because they'll end up with a factory and no more customers.
It's starting to look more and more like we will always need more memory.
This is abdicating an entire market in the hopes that your ability to manipulate it later, which has worked so far unfortunately, will hold. It is so obviously market manipulation here. Sure, prices would go up when this much server demand comes in but it is so profitable to make DDR5 that completely abandoning that market is stupid and only works if all players move in unison. It would just take one of the big three deciding to be the consumer maker to dominate the market and reap massive profits now and later when the server market cools and the other two try to come back. The fact that they are all shying away from such a ridiculously profitable sector is clear evidence of the collusion. They need jail time, again, but this time not allow those going to jail to become the CEO when they get out. They should also have fines that actually stop this behavior in the future. Fines so punitive that it actually brings into question their ability to continue as a company.
Micron is the only one to bail, and they are the smallest footprint. No one else has bailed.
Samsung recently shifted some capacity from HBM back to DDR5 modules, freeing up around 80,000 wafers/month, so DDR5 margins are currently good enough to compete with HBM for wafers, or maybe they are anticipating a turn, or both.
Maybe I'm missing something but this doesn't really make sense? Right now it's more profitable to sell to datacenters because they're less price sensitive. Consumers and consumer PC manufacturers don't really care who they're buying RAM from -- they'll just buy it for the cheapest price. When it's no longer as profitable to sell to datacenters, they'll just switch to selling for consumers and will then compete against each other. Why would any RAM manufacturer choose to handicap themselves by selling to consumers today when they can earn much more selling to datacenters?
Suppose someone came to my town and started offering to hire all the programmers for, say, twice the market rate. Would I really need to assume a conspiracy to understand why they all started quitting their current jobs? Occam's Razor would suggest they're probably just deciding to accept an offer for more money.
Don't hold your breath. Trump has said he has a micron Position.
Shovel seller says shovels will be expensive in winter so make sure to buy as many as you can today.
Shovel seller has access to knowledge about the market and processes that you could never know, though.
I'm pretty sure they are sold out for the foreseeable future.
and please all invest in the shovel making company!
Exactly. Memory seller says memory prices will go higher and remain high forever - got it.
I mean have you looked at housing for the last decade? Im glad I didn't wait for it to come down lower than it started the decade...
Just because he is the one selling it doesn't mean he isn't telling the truth.
I mean, yes, but they are also right. They benefit from the market conditions, and contribute to the price increase, but it’s pretty clear there won’t be that much available shovels for a very long time
Of course, because there was a HOA that became financially so big (on paper), that it secured 80% of the shovel supply for the next two years. Now everyone is fighting over the remaining 20%.
I mean they already ate all the shovel supply for this year, so unless the bubble pops now, the shovel supply for next year will be used by that HOA.
Especially when the equipment the shovel seller uses to produce shovels has been retooled to produce backhoes instead because of a digging bubble.
Shouldn't this be affected by the diggers from the Far Lands recently telling everyone how to angle the backhoe blade to dig 470x as much earth in one swing?
It would be severely affected if Jevon's paradox weren't terrorizing the globe right now.
I have huge hopes fox CXMT entering the market and at least offsetting this nightmare. Just this week i've paid ~$1k USD for 2x32G DDR5 ECC, insane.
CXMT has entered the market, but for now they cannot increase their production fast enough to influence significantly the global prices, due to the US "sanctions", which prevent them to buy modern fabrication equipment.
So they must use older equipment, which has low productivity for state-of-the-art memories. Because of the "sanctions" there are now Chinese companies that have progressed significantly in semiconductor manufacturing equipment, but the effect of this will also take some more time to manifest.
Nonetheless, if the 3 big memory producers will not be able to increase the availability of memory until 2028, then until then CXMT will grab with certainty a major part of the market (they have already grown quickly until 10% of the market, from a negligible percent previously).
The rest of the world won’t care. They will use whatever memory they can get, and if the Chinese are willing to supply it over the next 5 to 10 years, they end up on top…
The greed and shortsightedness from both the memory makers and their big customers trying to lock out competitors from memory purchases is palpable. The first act of a Shakespearean or Greek Tragedy is where the characters take the actions that lock in the circumstances that precede their downfall. And this sets up everything for the next act where CXMT has everything it needs to establish itself on the market with buyers that would perhaps would not have considered it in other circumstances.
> Just this week i've paid ~$1k USD for 2x32G DDR5 ECC, insane.
https://pcpartpicker.com/trends/price/memory/#ram.ddr5.5600....
I thought your price for 32g was you getting ripped off. Holy crap. The price is super high now. Now I am lamenting missing out on getting 128GB ram when it was $700, thinking it was too high. It is outrageously expensive now.
Everyone, and I mean everyone, is gated by ASML machines. They can't make nearly enough EUV machines to satisfy the demand. ASML has also not scaled their production nearly as much because they're quite far away from the AGI-pilled companies in the supply chain.
Well, CXMT is not gated by ASML machines because they can't buy them. They can raise their capacity because they have access to the domestically produced steppers. They are not as good as the ASML ones, and if the memory prices were not bonkers, they'd never make a profit using them, but as it is the low yields and high die sizes don't matter.
The problem is that you’re giving them an in into the market, and as time goes on, they will get better at making memory. That’s like saying you’re going to exclude the Chinese the space station. Look what happened. They now have their own space station?
Given the way that Chinese industrial policy has gone over the last decade, I'm skeptical that this made much of a difference.
They are. They are using ASML DUV machines for memory. Samsung, SK, and Micron have moved onto EUV machines for memory because they are better. So CXMT not only has to rely on ASML, they have to rely on a worse machine to make modern memory chips.
You don't need ASML machines to make RAM. BS.
All modern DDR5 and HBM are made using EUV machines.
The exception is CXMT, which uses ASML DUV machines in a less efficient process.
They will end up in five years being the dominant memory maker in the world.
They don't have the infrastructure to meet global demand.
Depends on what kind of memory you are talking about.
For DRAM, CXMT were only 1% of the global supply 3 years ago, and are now already 10%. They are growing extremely rapidly.
If you are concerned about consumer memory - smartphones, laptops, desktops, then this is what CXMT are making. CXMT do make HBM too, for AI accelerators, but their volume there is held back by sanctions.
I would never bet against a Chinese company ramping up production quickly.
They'll probably join the party and hype about tight demand by that time.
In communist China, the party joins you.
They're 10% of the market now, and they've shown the ability to scale up manufacturing rapidly. They certainly won't singlehandedly meet demand, but they might at least be able to temper the price growth quite a bit.
Yet. Maybe they will by 2028, making this prediction incorrect.
Right now, but that’s gonna change rapidly over the next five years. Once again, they didn’t have a space station. now they do…
doesn't seem unlikely, i'd probably ask someone who won't massively benefit from people believing that it will be tighter though
Yeah he doesn't want people delaying purchasing decisions
Yeah, if anything, this makes it more clear to me that all this wonky supply/demand stuff will correct itself sooner rather than later. I've been wanting to upgrade my workstation but the DDR5 RDIMM sticks are unbelievably expensive, last few weeks I've almost pulled the trigger anyway, now I think I'll continue wait at least for the beginning of next year.
Exactly - this kind of borders on market influence.
Effectively: "Now's the time to invest..."
The war on general purpose computing is ramping up. They are coming for your computer....get ready.
How are you getting ready?
Is there nothing we can do but pay the absurd prices and be happy about it? There is no hope left?
I'm genuinely afraid.
Install an adblocker. I'm not even kidding; it's been my secret to keeping older hardware with less (and slower) memory usable since the late '00s, and it's served me well.
That and, if you're into gaming, thank your lucky stars that there are all sorts of games from 15 years ago that are every bit as much fun now as they were then.
> How are you getting ready?
Finding out how to make the best use of the large amount of DDR3 RAM I have.
Those Chinese C612-based boards for the E5-2696v3, et al. (which support DDR3 and DDR4) seem OK with the custom BIOSen out there.
Installing linux on old hw is helpful. I got lucky and hoarded a few hundred GB of RAM before prices went up.
build the biggest beowulf cluster you can!
remind me in a few years when the same CEO is writing long post on Linkedin explaining why they need to let go half of the company.
Don't be too hard on the CEO. It will be unavoidable and he will be deeply sorry and humbled by their dedication throughout the years. Also, people will receive 2 months of benefits and FREE job search consulting (through an AI agent).
And it will be "the hardest thing he's ever done"
...and he will send many thanks for the great work all of you did throughout all the hard years!
In the past, when you were family!
Anthropic is expected to spend 500B of money it doesnt have. Based on these commitments, data-centers have ordered equipment which includes memory. So, a lot of demand for memory is built on a speculative foundation.
> Anthropic is expected to spend 500B of money it doesnt have.
... yet.
That is why the big AI corps all went public this year or are trying to do so. They are valued large enough that they will be included in the popular indices (NASDAQ, S&P 500, ...) from the start, which means an awful lot of money will flow into them each month automatically from all the retirement savings.
And that unfortunately completely prevents the core function that stock markets should offer - that bad actors get punished. Can't get punished when there is so, so much "dumb money" floating around and entering the fray every month.
I absolutely loathe Nvidia at this point. They've completely abandoned the core consumer that built Jensen's empire. Why not offer pre-orders of 5090s and MSRP? Why has every single drop at Best Buy gone primarily to botters and scalpers? Why not give end users half a fighting chance like EVGA used to by letting end users get in line before Nvidia crushed them?
When the AI bubble bursts, Nvidia is going to be Enron.
> Why not offer pre-orders of 5090s and MSRP?
AI bots (probably powered by NVIDIA chips) would gobble those up in seconds.
> Why not offer pre-orders of 5090s and MSRP?
Because it would immediately go to scalpers anyway. Like, say the 5090 - I'm seeing prices of almost 7000€ for that thing [1]. Assuming without AI it would be priced at 1500€ - I could make about 3 months of rent just by immediately selling the GPU to the next reseller who then peddles it to some AI training farm. And in this economy, oh hell yes I would.
> When the AI bubble bursts, Nvidia is going to be Enron.
Nah. NVIDIA was a money printer way before AI was a thing. Post-collapse they will still be a money printer, just at lower speed.
The ones who are going to get fucked over hard when the AI bubble pops is pensioners, the generation who is just entering the workforce (2007ff had an aftereffect in wage depression lasting for up to a decade), and everyone who took up investments to service the AI demand (utilities, construction companies and the likes).
[1] https://www.pccomponentes.de/msi-geforce-rtx-5090-ventus-3x-...
>> Because it would immediately go to scalpers anyway.
No they wouldn't, tens of thousands of unique users would be in line with dozens or hundreds of scalpers, if they're all allowed to buy one every 180 days per payment method, the end users actually get a fighting chance. Today they have no chance.
Scalpers use other people to purchase high demand items to then scalp.
There is enough profit to pay people to buy stuff and resell it.
And it's not just a few people. The sophisticated scalping operations have hundreds of straw buyers in their network. For items that generate far less profit than a 5090 currently does.
I've watched vans dropping of an endless stream of people in the parking lot of my local Costco to buy out all the cases of Pokemon cards within 20 minutes of the "random" drop they put on the floor. They go in, buy the max quantity of the item with cash provided by their keeper, bring it back to the van with a receipt and change, and get the $20-30 per item pay they were promised. There were 2 or 3 'legit' buyers out of hundreds of obvious straw buyers.
If the profit per item is $4-5k, expect this to go into turbo mode and be done in a marketplace/automated fashion and go viral. Just sign up to your favorite scalper's platform, submit your information, get a quote and a prepaid label and drop the box you got from nVidia off at USPS the next day. That "tens of thousands" of individuals will turn into "millions" really quick once it gets known there is rent money involved by clicking a few buttons and it goes viral on all the social media platforms.
The only way out from that is manufacturer enough cards to meet demand, and at that point the retail channel doesn't really matter.
Everyone I know who signed up for EVGA cards for the 30 series got one eventually, even if it took a couple months of waiting. I don't know anyone with a 5090 FE at MSRP.
Yeah, I will never buy an Nvidia GPU again. I assume many of their former customers feel the same way. When the bubble pops, I think Nvidia is going to feel the pain as they will have pushed many of their former customers into the arms of AMD.
I guess it is back to the old days, when home computers had the memory they had and that was about it.
When are we getting the wave from REIN (Rewrite Electron In Native)?
I look forward to memory optimizations in software design.
You joke but I don't see why mass optimization shouldn't be possible given the cost of mass refactoring went to effectively zero at this point.
Because the LLMs were trained on average software which doesn't give a shit about performance and they'll revert to the mean for sufficiently complicated projects. While they can help with searching for performance improvements, they also are liable to cheat if you aren't reviewing them for correctness of any algorithmic changes. Also you need to be responsible for teaching them the right benchmarks to use, since, again, performance optimization has been niche and not the standard.
Mass refactoring will always compete with mass new product launches and mass new feature production, so it's likely to always lose.
This has absolutely been a dream of mine. We act like it's the 80s-90s again and optimize all the software.
That's the one silver lining. I've always kept a lean desktop of mostly terminal based programs, and I've heard one too many times "just buy more RAM" as a defense for resource intensive graphical programs using Electron or modern browsers in general.
It'd be a shame if their two main sources of income (openai and anthropic) were to go bust...
I doubt it will change much, inference (not training) is very profitable and demand for inference is quite high. See: Mistral serving GLM on their own servers.
They're too big to fail at this point. US taxpayers will bail them out if push comes to shove.
I certainly won't be. The US Government that allocates my tax dollars might though.
In that unlikely case, open-weight hosting providers (and Google and Meta and SpaceXAI) would pick up the slack.
The only chance of memory demand going down would be breakthroughs in model size reduction.
The catchup models are all basically distillations of the sota models. Without the next training cycle things are going to stagnate. And as all those companies you mentioned are all linked to OpenAI and Anthropic and relying on hosting deals and things they are all going to be in the ringer when things to go south.
So on the one hand you have all the sota model makers doing investor expectation management in saying wet need a slowdown for safety (may or may not be true, but also means they don’t spend on the next training cycle before IPO? Could be making their books look better too?) and on the other we have a sense that the models aren’t yet at the stopping place where we can just not train another cycle and use distillations of the current generation?
> The catchup models are all basically distillations of the sota models
Source/citation?
Also gpu speedup. If vera Rubin is 7x faster then to serve same number of tokens you need 1/7 the memory. Did I get that right?
They are probably too big to fail already (and too friendly with the current US gov). And even if they fail, open-weight models will take over as the usefulness of the tech behind has been proven.
Sure, and they'll be cognizant of that risk when considering expansion. The lessons of the dotcom fiber boom aren't that distant in memory.
If AI companies have allocation of memory, can't they just sell the memory allocation if they are tight on cash?
I expect them to sell everything if they go bust to recoup what little they can for investors. That would double whack hardware manufacturers are their final consumers are now both leaving the markets as buyers and entering as competitors.
Well, but to who? And for pennies on the dollar, most likely. Also not all ram is made equal. HBM for AI cards isn't LPDDR5 in iPhones.
I don’t think they will. There is too much money in it now.
Only last year Microsoft admitted to having GPUs in inventory that they couldn't power, due to a lack of electricity. So much of this production capacity could very well go into making memory chips for GPUs sitting in a warehouse.
The amount of money spend on the AI hype train is crazy. Meanwhile we're wasting fab time making chips that might never be powered on. It's absolutely insane that there are more money to be made on hardware for AI that may never be used, rather than producing a product that consumers and businesses need right now.
Don't worry, the Trump administration is addressing this by lifting pollution limits for data centers.
But wait, I thought we all agreed that capitalism is the most efficient way to distribute goods in an economy!
I'd argue that this is a form of financial engineering that is somewhat removed from "true" capitalism.
By goods you mean wealth vertically to the oligarchs, then yes.
> Acer CEO says memory makers are hyping 2030 shortage fears to protect margins — PC prices set to decline by late 2027, cheaper Chinese capacity coming online delivers lower memory prices
If only we'd all made do with 640k!
One thing is certain: in that world, we would not have to brace for a future where pretty much any form of human labor is at risk of getting replaced by just having the machines oxidize even more carbon and telling those humans to just eat cake if they fail to sell any their skills for bread.
This is why I (actually/literally) just purchased a Commodore 64; I miss the days of crazy optimization and hardware hacking to fit so much into so little :).
Do you run LLMs locally?
I do not, no; can you give me some context for the question? :)
This but unironically.
Micron profited immensely by exiting the retail RAM / SSD markets. All the big manufacturers used Crucial RAM. Apple, Dell, HP, Lenovo, etc. They are the number one reason RAM is so expensive. They decided it was more important to make trillions selling to the A.I. hyped data center buildouts. They weren't wrong, they made serious money. I'm still rather pissed off about it.
As an end user it’s difficult to not be bitter about the situation, particularly when you have execs coming out and saying, “yep these prices sure are terrible, better buy now before they get even worse!”
I’ve seen people say, “well that’s just the free market, there’s a customer that’s willing to pay more than you,” but that argument doesn’t hold water when industry incumbents are actively obstructing entrance of new competitors into their primary markets. Micron, etc can’t both refuse to carve off more of their capacity for consumer-destined NAND and also refuse to allow others to fill the vacuum they’ve so conveniently created. Either you’re serving the market or you’re not; you don’t get to have your cake and eat it too under properly functioning, healthy capitalism.
How they refuse competitors to enter the market?
Micron has been lobbying to ban CXMT from the US market after Apple was entertaining using them as a supplier.
Memory production is restricted by patents. CXMT only exists because they bought patents from another company.
Yeah, sure the rest of the world, outside the West, isn’t going to care.
I thought that the bulk of memory is already "technically bought" by the AI hyperscalers in the form of the big (not so) "hidden" debt? As in: the supply should already be known.
Why the need for speculation?
Micron reportedly sold $100bn worth of DRAM last financial year, and reportedly has $40bn long term commitments. $40bn is more than their entire fiscal 2025 revenue, so it's a lot, but it's clear it'll be a minority of their available supply.
per yahoo finance:
MICRON INSIDER TRANSACTIONS (LAST 6 MOS)
Purchases 887
Sales 337,619
Seems like the Execs aren't seeing this as a good thing for their company
Sounds like I'm not buying a computer or a phone for 2 years. I hope my Pixel 6 lasts that long.
Micron stock has gone up 10x in the past 2 years in a sector where the barrier to entry is lower than GPU/TPUs. CEO will do everything he can to keep stock price from returning to the mean.
The ripple effects are going to be big. Its going to be rough for any company building products with RAM in it. People will hold on to what they got.
If I made a similar statement in my industry, I would be jailed or at least heavily fined for market manipulation.
I think as a manufacturer I'd be less worried about competition than about vertical integration by large players.
It makes sense to de-risk your supply chain and at this point memory is a key supply risk.
One factor constraining supply for RAM in the short-to-medium term is that memory manufacturers is that DDR5 (and assumedly DDR6) capacity is being moved to the much more profitable HBM.
I think supply will be tight until at least the AI market crashes, which could take years or happen within months. Nobody can really predict when this will happen but it's fairly clear (IMHO) that the revenue projections needed to justify the investments by OpenAI, Anthropic and SpaceX simply don't exist.
Without the AI boom, the economy is in recession and has been for awhile. We're in an affordability crisis and the energy crunch is coming.
The cartel has spoken.
Kudos to the memory supply-industry on converting their market in a way that a handful of large-scale customers moved from mere supply-forecasting to outbidding themselves to BUY future production of chips for years to come.
I would now like to open the floor to investigations in this industry to determine whether this is hindering free-market forces and is ripe for regulation, as no-one is able to actually compete with these handful of companies anymore.
...anyone?
...ah yeah, right, I forgot. They also found a circular financing scheme which enslaved the entire global economy and organically prevents governments from properly governing their industry.
CXMT is coming for this market, and this creates a huge opening for it.
> Chinese CXMT DRAM doesn't look like the budget savior many were expecting — new modules enter the market, but prices still track the big three
https://www.tomshardware.com/pc-components/dram/chinese-cxmt...
That’s not the point. You’re speeding up the progress of them becoming a competitor, a relentless competitor worldwide.
And you are now encouraging other large pocket tech companies inside and outside China to give thought to taking memory in-house if that means making and selling devices at a reasonable price in a timely manner.
CXMT isn't going to just magically price memory like it is 5 years ago. Maybe 10-20% lower, but not some integer multiple lower.
That's today. You have to think about what the industry is going to look like in a decade.
Building a DRAM company from scratch is hard, expensive, and risky. You're burning a fortune to get started, and the result is a commodity product. But the AI bubble allows CXMT to have terrible yields, sell a legacy product, and still make a healthy profit! This is literally a once-in-a-lifetime opportunity to catch up.
And yes, right now that means selling DRAM for 10% below the crazy AI-inflated prices the incumbents charge. But sooner or later either demand will drop or supply will catch up. CXMT, like many other Chinese companies, is undoubtedly willing to operate at razor-thin margins to increase market share. Either the other manufacturers move down as well, or they'll be made irrelevant by CXMT stealing all their customers.
Absolutely. far too many people are thinking about the next 2 to 4 years. How many times must they see the Chinese play the long game before they figure it out?
I agree with everything here, but lots of people have magical thinking that CXMT will just instantly erase the run-up in memory prices.
Oh sure, they'll take profits while they can, but the supply increase will start pulling the market down. It's how price signals work.
Only way to keep prices this high would be a multinational cartel of RAM makers agreeing never to compete on price. One defector kills it. Not likely you could keep a cartel like this given that the players are all on different sides of geopolitical tensions. China's not going to want to help keep Taiwan, Japan, Germany, or the USA in the money.
Not gonna happen with China in the world, they will sell to every place else on the face of the planet this memory bump is temporary, also keeping it high means other companies will reconfigure what they do. if you want to continue to sell millions of devices at a reasonable price, you’re not going to sit on the sidelines and wait for the three-headed cartel.
You are going to get busy. and design and engineer around the cartel. Is there one company in the world that might do that? A company that has the engineering, design and the money with the capacity to execute over time?
Are you telling me that Intel will corner the market on https://www.intel.com/content/www/us/en/history/virtual-vaul...
Because that would be smart. And they have the packaging tech to add compute to it.
Obviously, because they are making a killing.
I just got some windfall. Maybe I will pickup the Macbook Pro I so desperately need!
If you follow the money though much of this future “demand” is on fulling orders from the AI labs and at the moment the AI labs don’t have the cash to pay that bill. A rather important detail in this whole mix.
So either the labs need to start printing money, now, or there’s going to be a whole bunch of excess chips hitting the market in 2027-28.
Someone in the middle is going to get left taking delivery of all these chips but unable to collect payment for them. And someone else is going to get left holding the bag on data center loans to house those chips. That’s the storm ahead.
Do you not believe that Anthropic and/or OpenAI are going to be able to IPO? Once they do, that's a huge source of capital for them that could absolutely be used to fund hardware purchases.
Based on current reporting the IPO wouldn’t generate anywhere near enough cash to pay for the bills coming due. Just between those two they have over a trillion $ of commitments coming due that need cash payment. Thats actual dollars they need to pay and neither the IPO nor current cash flows appear to be generating that kind of money to pay these bills when they come due.
It’s a big problem which neither big lab has a great answer for right now.
They can IPO all they want it's not gonna save them from the Chinese labs selling equivalent models at 80% discount
These labs made a bold bet that cost trillions, but those trillions in research are a box of concrete around their feet now
Folks... If token demand goes up 3x and vera Rubin boosts tokens per second by 7, then wouldn't you need less memory?
Or did I mess that up?
We'll simply use more tokens then.
https://en.wikipedia.org/wiki/Jevons_paradox
So instead of consumer's not being able to buy memory, consumers won't be able to buy memory?
makes sense. large backlog + sustained demand + constant manufacture time = larger, faster-filling backlog. i'm sure this has a fancy economics term
Ah, so the world is going to end in RAM chip maximizers, not paperclips.
I guess those estimations based on projection for data Center build out, which include 30GW+ in 2027 alone, more then 3 previous years combined. Surely every human on earth will start token-maxing tomorrow and bring all of their money to their AI labs overlords to justify that explosion in spending, right?
In the short term, what `Neil44` mentioned "Yeah he doesn't want people delaying purchasing decisions" makes sense.
In the long term, the big comapnies are pushign to have consumers to adopt "terminals" back in old days.
- we use crappy devices, while cloud got beefy machines.
Glad I got received my M5 ultra studio already, I smell a price increase coming.
and here I hoped to suspect that my DRAM prices were all fake news run by my MITM ... goood, damn it
this would be the time to revive and evolve some of that Made In Germany Engineering ...
Not good. Prices are never going down
Of course they will come down. Supply will rise to meet any sustained demand - it always does.
No one will use AI products if there won't be consumer products because of RAM shortage.
The plan is to make us all use shitty hardware like a chromebook and rent compute from the cloud for anything heavier than a web browser.
> anything heavier than a web browser
The way Chrome feasts on RAM, might need the cloud for that too.
Honestly, FF became similar worse: On my 16 GB machine, when surfing some hours FF has allocated around 10-11 GB and for whatever reason it does not free them, unless I exit and restart it
On linux do something like `systemd-run --user --scope -p MemoryHigh=XG -p MemoryMax=X+2G firefox` to constrain it since the firefox people refuse to implement the memory limits logic that say Edge has..
But at the same time, this isn't just a Firefox issue, although they may be worse than most. Its a fundamental tenant of GC languages. You trade CPU (for the garbage collection) for RAM utilization. IIRC This was apparent by the late 70s with Dijkstra, but very clearly indicated in say Hertz where they are explicitly scaling GC overhead vs memory footprint to find that it takes ~5x the ram to approach the efficiency of an explicit allocate/deallocate model. Lower footprints are achievable but the GC overhead goes way up just to reach 2x the ram footprint such that the GC tends to dominate the CPU utilization.
And there are lots of papers like this for long running processes. GC works well if your model is basically allocate as much as you like, exit the process before actually needing to do a proper GC pass. Especially for contended systems where the memory is returned to the OS. This is why say PHP tended not to have this problem, because the memory was largely freed at the end of every request.
The New Tab Suspender plugin has my 50+ Firefox tabs at 2.2GB of ram usage right now. Having every tab ready and waiting for you is of course going to have increased memory usage.
Same. I have to restart FF once it reaches the 10GB mark.
Perhaps we should blame this on shitty website design. So many of them load huge assets for no good reason. They get away with it because modern processors are that good, bandwidth is cheap, and until now, RAM has also been cheap.
I do not get why this gets so many downvotes:
- the described behaviour is the one I experience on my Win10 notebook (Lenovo X1)
- Firefox wasnt always that, thats what Im refering to
- I can see in ProcessExplorer how many threads it starts & keeps, and for whatever reason this is always a lot, even if I open only 1 tab after coming from new start
But the cloud is becoming unaffordable as well…
There’s always going to be consumer products, even if the value proposition is worse.
If there’s a silver lining, I hope it leads to more optimisation, less Electron, and less bloat.
Typed from my 8GB MacBook Neo; which is honestly plenty fast enough. It “feels” like a 16GB windows laptop, even tho of course I know it’s not, but there’s something about optimisation.
And a 4gb Linux machine.
Lots of more niche products than Apple’s laptop won’t exist for a while due to the price of hardware. At a smaller scale the inflation hardware companies experience is really, really massive
While I definitely agree that there will be a slow down in purchase decisions for both consumers and businesses I struggle with this idea that there will be “no consumer products” in the market. There will be.
There will be, but the question who will be able to afford it.
It’s like saying that there is no housing crysis in several countries because the market is full of houses, apartments, etc.
You’re making what seems like a zero sum assumption. I am still using my M1 MacBook for work. I have a first version Mac Studio at home and countless other devices laying around. The world will continue to move forward and yes purchasing will slow down but folks will still be using their existing phones, laptops and other devices.
I don’t think Apple is gonna go home because of the three-headed memory cartel, since they didn’t spend/burn billions (one trillion?) of dollars on AI, I think they will replace the cartel in-house, Intel is out and Qualcomm is on the way out, the three headed cartel can follow them.
The future is starting to resemble smartphones that use predominantly RAM for local compute and inference while all persistent storage is managed by the cloud.
They can sell the AI products to write memory-optimized software to replace React apps.
Tragedy of the commons.
Neither consumer products nor RAM are really commons?
https://en.wikipedia.org/wiki/Tragedy_of_the_commons
If anything, it's the tragedy of econ 101.
The AI companies will sell those products to us. Locked down and all.
And People believing that everyone will have a robot at home.. until they realize that a human is much cheaper to maintain, a bread and a liter of water is enough...
- i have the exact opposite to say
- one of those "frontier" model IPOs s gonna tank badly and take out the entire AI hype with it
can't tank if you keep pushing the IPO lol
Man whose compensation package is strongly tied to the economic performance of his company says company's economic performance will be great in 2027 and 2028 (so people should invest now to benefit from that bump)
Gonna be crazy to watch this bubble pop.
I like to tell people that "AI" is the "last bubble." It was VR, the metaverse, crypto, NFTs, AR... AI is the Last Great Hype. There's really nothing beyond it. When this bubble pops it'll be remembered for decades to come because nothing will replace it.
Biotech has a lot of potential there, at least in principle.
Everyone has said this about every bubble while in the midst of it though.
"Crypto is the last 'big thing' because it'll revolutionize currency and everyday transactions"
"Metaverse is the last 'big thing' because everything you own will be virtual in the future'
Realistically I see AI as no different.
It wasn't the first SV-techbro hype technology, and it won't be the last.
The LLM bubble may fizzle rather than pop in any dramatic fashion, but it's certainly not the last word in AI - much more the exact opposite.
We've had AI "bubbles" / hype-fests before (expert systems, Japanese 5th generation systems), and I don't expect this will be the last.
There will also be technology bubbles and hype-fests in the future outside of AI - it's human nature. Hard to predict what they will be of course (the future is like that).
Until governments stop printing money and sending it directly to people that have nothing to do with it but investing, people will keep inventing random useless stuff to invest into.
Well like the very short 3D-TV fad, I can see a big sudden goldrush into AI powered robots (more than companies are in now).
It is also possible they all pivot to copying Neural Link and upselling the whole world on brain implants that the oligarchs can control.
Sounds like the gov needs to spin up another “Project Warp Speed” like they did during covid.
Can we get the South American Cartels on it ..
They are milking the market now.
Make no mistake here: they also plan to retain these high prices. It's like the mafia now. There should be instant-jail time to cure these CEOs of their greed. Like, go to prison for a week, to help refresh your mind and stop thinking in terms of milking the customers dry.
If/when the bubble pops demand will fall so hard they'll have little ability to keep prices too high. Especially with Chinese companies pushing out a lot more product by then.
That's probably planned but it will probably be in agreement with other producers.
Micron may be out of business by then.
Micron reported full year revenue of $133bn, up 256%, $100bn of which was DRAM, with non-GAAP earnings last quarter of their financial year of $37.7bn on $54.23bn of revenue. Seems pretty healthy to me. Any reason to think they'll go out of business?
How? They weren't doing poorly before 2026, and as far as I can tell, their profits have risen dramatically.
What if Oracle, Anthropic and OpenAI can't pay their infrastructure bill?
Then memory goes back to pre-2026 pricing and memory is still profitable with a cartel market of less than 5 manufacturers.
But only in 2028 because all the wafers rotting away in the warehouses are HBM which are too expensive to package or integrate into consumer electronics with tighter margins. Looking forward to having enthousiast-grade GPUs with HBM again though.
I'd imagine the quadruple revenue they collected this year would be able to tide them over for a manufacturing cycle or two.