Finally secure a research grant and had budgeted money for a new compute workstation and the RAM prices are killing my budget, because I'd love to get at least 256 GB of RAM and that's 4-5 grand now. Ridiculous, especially since I'd like to have a GPU in it with at least 32Gb of VRAM for medical image model training, so that is also like 4-5 grand too.
Do I wait two years and hope, or pay, isthe question I keep asking myself.
At least in Europe we are building a common super-computer infrastructure exactly for this. Tier 0 (EuroHPC) to tier 3 (University clusters) now have considerable amount of GPU (both due to demand, but also since classical CPU-heavy FLOPs have becime even more expensive due to RAM prices). Our system will be in a ISO 27001 / TISAX level 3 scope (rather relevant for engineering). The system work up to a PUE of 1.1 and will also solve part of heating decarbonization problem. We need souvereignity in research. But we at least in research we do not need each lab buying their own workstations anymore. The systems are actually also now also offered to start-ups and small enterprises that don't have the capex/opex for efficient and secure compute themselves.
That is a third possibility, because there are Federal supportrd resources for compute, see [1]. And I can and will probably use at some point, but there is something about not having to deal with other people's systems and having that bare metal right there.
More like building shared infra to get at a better and more efficent end of scaling laws, than having many small clusters all over. I don't think it really is a privacy question but just a way more economical solution.
Basically make your own hpc provider for academic purposes, and rent out extra capacity to smaller commercial interests to utilise that spare capacity and also prop up the local economy with usable research compute.
Here it is not the privacy concern of the user but that of third parties (e.g. those in medical images). And yes I am a strong advocate of using shared infrastructure for this rather than letting some PhD student handle this a part of their research (which is the reality with lab clusters)
Information sources vary per country, but there should be lists like this one [0] for Germany. Access condition are a bit of a mess. EuroHPC was started to harmonize things a bit [1] but also here the reality is complicated at least for the 50% funded by national money.
It's possibility that I've considered, am still considering. Mostly because I am used to working locally, but also because I am not sure which works out cheapest in the end. For my use cases, if I need anything with more than 24 or 32 GB, it definitely makes sense to rent I thing. Those cards are extra pricy. Also, I keep the GPU and analysis computer for the lab after the grant ends, but I don't keep my budget for online conpute..that would go back to the Federal government, so it makes sense for some level of hardware to be around for lean times
I mean, Runpod is relatively cheap at that VRAM tier, it might be a good interim solution so you don't have to completely sit on your hands while you wait.
> On 5 April 2006, Sun Woo Lee, Senior Manager of DRAM at Samsung Electronics, entered into a plea bargain with the US Government for his involvement in the price fixing conspiracy. Following the plea agreement he was sentenced to 8 months in prison and fined US$250,000. Lee was subsequently promoted to President of Samsung Germany in 2009, and then President of Samsung Europe in 2014.
how many successful careers include 8 months in prison
You might want to read the court docket attached at the bottom of the page. The plaintiffs present substantial circumstantial evidence, consistent with tacit collusion (which is not illegal). It's for the courts to investigate and decide if there was explicit collusion, which is illegal.
Tacit collusion and just working within the same market can also look the same. I have no skin in the game (other than wanting prices to go down again), but there is also way too much emotion involved in the discussions about this because it is hurting our wallets.
- Cutting DRAM production in 2022 as a result of low prices/demand made sense if they didn't expect the market to recover as quickly as it did. A lot of the recovery is also likely caused by the start of the AI boom.
- The deals OpenAI made with SK Hynix and Samsung was a smart move by OpenAI where they drove up prices for everybody else while securing "cheap" supply themselves. I dont think its clear whether SK Hynix and Samsung came out ahead in that deal as the demand likely would have risen either way and it wouldn't have needed to go up by as much as it has now.
- Moving production to HBM also makes perfect sense when that is what the market demands and is willing to pay a premium for.
Every billionaire has incentive to make public statements influencing their competitors while not doing what they say. If everyone else reduces production, you get to take the excess.
See the repeated prisoner's dilemma, where the fact that you keep repeating the game completely changes what it means to have an optimal strategy. A company doesn't make a choice to collude or not once, they make it over and over again.
Billionaires and corporations collude every day via Signal, Telegram, WhatsApp, etc. We've just made it almost impossible for them to leave a paper trail, and for us to prove it. Thus we get lots of people defending their actions as "just business" or something: plausible deniability is all most people need.
> but there is also way too much emotion involved in the discussions about this because it is hurting our wallets
That's a matter of course. This goes way beyond our wallets. They are actively pricing us out of computing. They are deliberately pushing computers back to the big iron mainframe era.
Computers are subversive: they democratized cryptography, copyright infringement, all sorts of things. Now that humanity has developed AI, they will not cannot tolerate mere citizens having access to powerful computers any longer.
If we don't fight this somehow, it's over for us. Everything the word "hacker" ever stood for will be destroyed. We need some kind of second amendment but for computers.
I was going to say the same, so I'll just add to your comment: it's possible too that there's a brave new world of local LLMs just around the corner that this kind of thing is retarding.
(Speaking of which—I've seen people suggest a seti@home-like approach to creating a frontier-level, fully open LLM. C'mon, guys, let's vibe code this thing!)
> The judge found that long-term HBM lock-ups, refusals of long-term conventional DRAM contracts, Micron's Crucial closure, fab delays, the failure to expand despite record prices, near-identical customer vetting and the companies' prior cartel history supported a plausible inference of conspiracy.
It makes sense that an industry historically prone to boom/bust cycles wouldn’t immediately go all in on expansion for a possibly transitory demand spike.
Tacit collusion is only legal in specific circumstances. If the companies involved seem to be acting against their own economic self interests, are communicating indirectly, or are collectively using some external mechanism (like realpage or whatever it was called) to set prices, it can be illegal.
And they were proven to be true. From the Wikipedia article:
“To date, five manufacturers have pleaded guilty to their involvement in an international price-fixing conspiracy between July 1, 1998, and June 15, 2002, including Hynix, Infineon, Micron Technology, Samsung, and Elpida.[3]“
I have to disagree with you there, the parent comment is claiming that this has happened before and they cite that Wikipedia article. ;-)
In my opinion, the fact that this has happened before, and that the companies involved eventually plead guilty, is an important piece of information for people following this story. Especially in light of the statements from the companies involved today.
The comment said "this isn't the first time", present tense, meaning they are currently being found to have done it again. That's not true (the finding has not happened yet), and what the person you're going back forth with is arguing.
The situations are quite different though. The DRAM industry price fixing scandal came in the aftermath of them heavily scaling up production. The manufacturers had all scaled up the production so much that they couldn't sell it all. They would all have had to sell at a loss if they were to compete.
It is the complete opposite of the current situation.
Collusion is a viable strategy, regardless of the market situation. They've been shown to apply price fixing in the past; presumably they still know each other's phone numbers.
Having done all three over my career (local machines as student, University's HPC cluster as grad/postgrad and eventually cloud in the private sector), I'd 100% rent out compute. Buying a custom rig for local ML training makes zero sense in 2026. With renting you'll be able to train bigger models/datasets faster on much better hardware while paying far less in total. Especially if you're doing project-based work.
If you use a custom disk image too, you can have it boot, apply security updates, reboot, then ssh in within minutes.
Work for however long, shut down. I do this with linode and their GPU VMs. If you shutdown nightly and on weekends, the costs drop dramatically on a monthly spend.
I only need it a few hours per day, so very cheap for custom compute.
The usage pattern i've seen on flexible compute is more aligned with 'I need this process to execute, run it when it's cheap' not 'I'm going to kick off this process, stop it if it get's expensive.' for example AWS Spot Instances
Friends of mine wanted to buy a GPU server last year for their new project. Management told them no, that there was a deal with cloud providers and that investing the budget in cloud compute was better.
The project ran out of budget roughly three months in. From what I heard, status meetings consist of everyone getting together once a month, asking "have we gotten any more budget yet? no?", and then wait until the project runs out.
Cloud providers need to make a profit too, and they've gotten very good at ensuring that their clients forget this point. The whole advantage of cloud compute is that you don't pay for the time you're not using it, but once you reach a certain point you're the one covering the overhead for everybody else.
As a scientist with a big workstation and access to several HPCs I can say that there is nothing better than having a powerful workstation. Yes, doing some sort of long runs is better on a cluster, but being able to do things locally without queue delays and various limitations imposed by HPCs is extremely valuable for initial development/experiments.
HPC ops teams are either really good or they're a dumping ground for grumpy DBAs/Operations people who have no skin in the game of getting anything done and it's a complete nightmare working with them.
As former HPC ops albeit in public service, we also tend to have funding for everything but ops personnel. This means you also often cannot afford the personnel you might be wishing for as they tend to go more towards better paying fields like devops etc.
It might (I say might) be worth your looking at xc [1] as a way to get more performance out of local/LAN hardware - assuming your program is data-parallel anyway.
Exactly. The number of times you go, "oops, it would be better if I changed X first" matters less if you can restart right away without waiting to restarts (and also using up your credits on the aborted run).
Simultaneously everyone needs to be doing AI research but no one can afford to do it except 5 companies. It's like no one remembers how major discoveries happened in the past.
I would love to believe the price will be lower in two years, but I doubt it. Why would it be lower? There’s still a shortage and most of the stuff it goes into hasn’t even been built yet.
Most of the stuff it goes into hasn’t even been built y̶e̶t̶.
Kevin O'Leary's 20 GW data centre is never getting built. Maybe he'll do a "phase 1" at a few hundred MW, but he's not going to build the largest power station in the world.
The only consumer GPU with 32GB of RAM in the current generation is the RTX 5090. NVidia claims it hasn't been discontinued but it's also impossible to buy and the resale prices have spiked to $5k+. So for a PC with 256GB of RAM and a 32GB GPU you're over $10k just for those two parts. Many believe, myself included, NVidia's new "flagship" will be announced in early 2026 and will be the RTX 5080 Super with 24GB of RAM, also probably costing $2k+ anyway.
What you probably want instead is a Mac Studio with an M5 Ultra. The 256GB model is ~$11k. The problem? Demand is so high that the lead time last I checked was March. But this will give you 256GB of shared memory, well above the 32GB you'll get on a 5090.
If you can live with 96GB (or less) on a Mac Studio it's quite a lot cheaper and the lead times are far quicker.
> The only consumer GPU with 32GB of RAM in the current generation is the RTX 5090.
Well not really, there is the AMD Radeon AI PRO 9700, which is a RX 9070 (or possibly the XT, dunno) with 32GB VRAM. It does use slower GDDR6, but it has "held its price better" (for some definition of better) and is still available for like 2.3k€ (this includes taxes). Though I remember when I was thinking of buying one when it was 1.4k not too long ago and didn't :( now I just don't bother looking anymore.
Slower in what specifically? It wouldn't surprise me if the card is up to 3x slower (it has roughly 3x less bandwidth), but do note that we are comparing a 2.3k€ card vs a +9k€ card here (plus on top of that the power usage on the RTX 5090 can be up to 2x under load).
The RX 7900 XTX has slightly higher bandwidth at almost 1kGBps compared to AI PROs almost 650GBps but has only 24GB instead of 32GB, though I can't find the card on sale anywhere rn.
In the end I see it mostly as a "is the massive jump in price worth it, or do you risk it and maybe get multiple AI 9700s instead and deal with ROCm instead". I can't tell how good ROCm has gotten in regards to training and such, but for inference Vulkan on my 9070 XT is pretty solid.
Someone should try vibecoding a massive backplane machine that takes e-waste cards like Tesla K80 as RAM sticks. Doesn't work opex wise but eBay sellers would love a bulk deal at this point.
You might want to consider DDR4, it's much cheaper, and make up for it with a lot of memory channels (Epyc or Threadripper Pro). Those older platforms can be had for a relative song on eBay.
Depends on what you're doing with it, though, obviously.
CPUs, yes. Motherboards, no. Expect to pay 1.1-2k for a 2017-2020 motherboard.
The market has gotten so tight that antiquated hardware has become competitive. Too many of us are trying to buy the same aging platforms for exactly this reason.
It's been a couple months since I looked, but when I did, you could get Threadripper Pro/mobo combos in that range on eBay. There were stupid prices listed in various places, hoping someone will fall for it, similar to the $30k 6000 RTX listings, but that wasn't the best price.
I don't think you get how specialised these facilities are. They can't convert from one process to another, for the same reason they can't just build a new one on the new process. The physical building is not even slightly the bottleneck. If they can afford a new process they can afford a new building for it.
There's a story where a fab company tried to open a fab in America but the chips kept coming out broken, and they tried everything and the only thing that worked was disassembling the whole factory and rebuilding it in the exact floor layout of their Taiwanese one.
Sometimes they do salvage one or more machines as both processes use the same one. But that had better be worth shutting down the old process, financially.
This production is constrained by the machines in the facility. The ddr5 production line does not magically produce more silicon by staffing it with more workers.
A high-end chip fabrication line is a very highly-tuned system. It's not a case of just buying the right machines and slapping them together, even doing a copy and paste does not guarantee the right results. You need to spend a lot of time adjusting the process variables and troubleshooting issues to bring things under control, and then keep monitoring them to maintain that control because things will drift over time or further issues will occur. You also need to bring a whole new set of people up to speed with how the fab works, and they also need to learn the quirks of that specific one as they go. These fabs are already optimized for near 100% utilization of the machinery in them, so more people doesn't mean more capacity.
Is it just about the factories though? I thought a big part of the current shortage was there literally not being enough silicon wafers to meet demand so much of that supply being redirected towards HBM led to a shortage of DDR4/DDR5 production.
The other things that use the wafers aren't 1000% more expensive.
On RAM side it's purely "the factories that turn wafers into RAM", and big part of the problem is that HBM just yields less bits per wafer (the number I've heard is between extra size and lower yield, about half) so more of the capacity moved to making HBM also lowers total bit count made
China are keeping their pricing high too for now. ChangXin Memory Technologies (CXMT) chips are entering the market but they don't seem to be reducing prices. Additionally I expect China will serve their domestic market preferentially. Hopefully that still drives prices down long term.
Memory is different because it's largely not a consumer market, but instead an input to other products and those products are not sold to consumers but to enterprise customers. I'm guessing there are plenty of domestic enterprise customers to buy the memory produced.
Of course they did, in my opinion.
They knew exactly what would happen. These people making choices arent morons. SK Hynix, Micron, Samsung all saw the opportunity as long as they are in sync. Then they all sit back raising MSRP with each other as "fake competition".
They could literally reopen manufacturing. But why? They know its free profit through inaction as long as none of them fix consumer DRAM market.
This act of playing dumb and ignorant has never been easier. Getting really old
They "literally" couldn't, there's no idle fab. They could redirect manufacturing to make more consumer RAM instead of HBM but why eat into your profits by doing that? The shareholders won't be happy.
I'm not a fan of this situation but unless there's some WhatsApp texts the investigators will find to point to collusion this is dead in the water. The filing itself is based on a lot of circumstantial evidence.
One can only hope that this insanity finally ends. That being said, all the neoclouds and nvidia too have been on a stock descend for a few months now, so this might all be part of the bubble slowly breaking.
The argument that this is a bubble rests upon the premise that the amount being invested is greater than the returns that can be expected. This is not a calculation where adultery should be a factor. Perhaps price fixing, certainly not people being held to account for price fixing.
It's a 3 way prisoners dilemma market where the prisoners have been caught explicitly colluding before. Now they do not need explicit colluding anymore as the pattern and outcome has been established already and the "jail" is as good as closed to new entrants. Clearly the expected "market" is not working in this niche, so something will have to be done about it.
I honestly don’t understand how can anyone conclude that. All 3 players have their entire inventory sold + reliable reports that their entire next year’s inventory sold.
If you really are thinking from a game theory perspective, in order for it to be the prisoner dilemma, they each need to benefit from lowering prices individually (to cheat). But given that demand outstrips supply so much, they each just lose if they lower price, regardless of what the other players do.
They are making the thing. What they cannot do is see the future in all its particulars for anticipating how much of that thing to make. We alao don't want situations were we use up too many resources and end up with more ram than other chips (which would cause a different shortage).
These things are complicated, and people act like the mere existence of increased prices is evidence of malpractice. This is a naiive and juvenile attitude.
It isn't 'seeing the future', it is exercising your power in the present to control what futures are possible.
Like, for instance, forcing your customers to buy more chips than they can install, lest they be cut off entirely. Or increasing hypothetical demand artificially by breaking international smuggling laws using your girlfriend's shell company.
It is not 'naiive' (or naïve, even) to remember previous instances of crimes committed by exactly these companies using exactly the same tactics.
Unregulated markets create these predatory incentives, it shouldn't shock your or anyone that this has come to pass again.
They cut supply of DDR4 around the same-ish time because DDR4 prices were dropping significantly (due to oversupply) and Intel and AMD processors were going to DDR5.
If the AI bubble hadn't come around buying up all the RAM, most new ram purchased today would be DDR5. The shift towards selling more HBM is also probably a bit unexpected.
> Judge Wise found that near-identical customer-vetting practices all three companies allegedly adopted in January 2026 "strongly support a plausible inference of conspiracy." She rejected the companies' "obvious and non-conspiratorial explanation" of booming AI demand, noting that the vetting began after the AI boom and the HBM pivot were underway.
Note, it's not just one thing, it's a combination of practices that make the collusion more likely explanation than otherwise
> All 3 players have their entire inventory sold + reliable reports that their entire next year’s inventory sold.
If they had started building new fabs earlier, they might have been able to sell even more inventory _and_ gotten back into the consumer market that they now neglect. Instead they've been way too slow to respond to demand or, in this case, purposefully didn't respond to demand.
But memory was dead cheap just a year ago. About 10 times cheaper. I don’t know if it’s reasonable to assume collusion here. It’s far more reasonable to assume they (and everybody else) didn’t anticipate the shortage.
No one anticipated that OpenAI would put in an order for 40% of the global supply of memory in 2025. And then it became a mad dash for everyone and that’s why we are where we are today.
I really doubt that it was a surprise. The big cloud providers talk to suppliers (and potential suppliers) frequently to telegraph future demand, both the amount of parts and specs.
No it was a huge surprise. I bought a 512 GB Mac Studio for $9k in April 2025 which was market. Even as recently as Jan 2026 I bought a 5070Ti for $750, the retail price. Now it’s $1500. No one anticipated the effects.
The analogy/classification works if you assume they engineered the scarcity through collusion. Then its "lower prices to win the market" vs "work together to artificially lower supply as a smoke screen to squeeze on prices."
The sense of scarcity factors into the ability to negotiate these contracts. Cooking the reports has a direct impact on contracts and does increase real demand.
The reason there are only three is that other players in their industry have been wiped out by investing in new production capacity to meet demand that doesn’t exist by the time they come online. It is very much an open question whether HBM demand growth really continues as forecast - it could collapse dramatically for various different reasons.
You'd need years of contracts - at prices no can predict are going to be good - to pay off the investment. Do you think the companies who failed or left the industry did so because they didn't know what contracts or futures were?
I am curious to see what will be uncovered in the trial. I know way too little of the market and tech involved to come to a conclusion (beyond anything motivated from wanting cheaper memory again).
Nearly everything could have a non-colluding explanation depending on the tech and market. (for example I would be skeptical to build up extremly expensive production if i follow the widespread belief that Ai is a huge bubble)
However, collusion is not magically out of the question with previous instances and some unexplained indicators. (which imo is what this dismissal means)
AI being a bubble doesn't change that it is here to stay and thus there is a need for more RAM either way. There was a railway bubble in the 1840s, but it didn't lead to trains going away, either.
The point is that while AI might stay, that an AI bubble popping will massivelys reduce demand for RAM. So any production capacity you might have built up is not able to earn back its cost.
If you think demand will break down before you can get a return on your investemtn, you should not build.
If you think of a different aspect of "the AI bububle popping", where demand for RAM does not significantly decrease, that's fine of course. But RAM demand by AI going down massively seems to be a major part of what a lot of people imagine when they talk about the bubble going away.
It will absolutely earn back the cost, just not at 10x it is now.
Even if we got back to pre-AI bubble, hell, especially if we got back to pre-ai bubble pricing, we'd see a lot of push for local AI and 32 GB+ GPUs/128GB+ SOC systems
> It will absolutely earn back the cost, just not at 10x it is now.
I put up as an example and prefaced it with me not knowing enough about the market and tech. I have not done the calculations, because I have no access to any of the relevant numbers, do not have a relevant model of the market dynamics I would expect, and do not have to suffer the losses if it doesn't play out.
It is great that you have more knowledge to be able to confidently state those things, however, I have to say I have no reason to actually believe you do.
Why would it reduce demand for RAM? Dot com bubble collapsing didn't reduce the demand for the Internet. Railway bubble collapsing didn't reduce demand for trains. And if the housing bubble would collapse it would also not reduce the demand for housing.
A bubble means something is overvalued. The stock prices will go down if the bubble pops, but not the demand.
I think your definition of a bubble is narrower than the use of most people. One possible reason for an expectation of reduced demand is that reduced investments can lead to companies having to reduce hardware investments themselves.
One example of people not sharing your definition are the many comments of the form "can't wait for the Ai bubble to burst and ram/gpus getting affordable again".
Also I am not sure about your examples. Same as the other commenter: I have no idea about the rail bubble. But I believe the dot-com bubble burst did in fact reduce demand on compute infrastructure. (Not long term of course.) eg https://www.theregister.com/on-prem/2002/01/11/server-sales-...
> reduced investments can lead to companies having to reduce hardware investments themselves
Yahoo kind of went away, but Google replaced them. I really don't understand why people would think that anything would change even if all the known AI companies went under tomorrow. We are still going to use AI more and more and we are still needing more and more hardware for that. Doesn't matter which company it is that is involved in all that.
> Not long term of course
Building more fabs is very much a long term thing, though.
If you look at passenger rail maps of the United States, trains definitely went away.
Passenger rail versus freight rail also demonstrates how the railroad bubble and aftermath ended up proving that not every use of rail is inherently profitable. In the long term, passenger rail is rarely profitable and almost always has to be subsidized by the state in some way with only a handful of exceptions.
It’s like George Carlin said, all these people attend to the same social environment, they know what’s good for them, they don’t need to meet in a hideout to fix prices to collude.
It's good that the courts finally look at the price fixing of
the RAM cartel. I think it must become a much bigger lawsuit.
It can not be that these de-facto monopolies keep on fixing
the prices upwards. Even in a very capitalistic society such
as the USA was once (before the orange mafia hijacked the
country, even though in a more recent election the mafia lost
big), capitalism assumes that free market competition lowers
the prices and affordability for the average Joe, as a buyer.
Clearly this is NOT the case here in the RAM cartel. You now
pay like 3x as much as you did two or three years ago - and
the high prices are maintained. That's a cartel. That has nothing
to do with a free market.
I also think mandatory jail times for the CEOs need to happen. It
can not be that they can buy their way out of punishments and
damage they cause against society.
Capitalism does not assume anything. If anything, unfettered capitalism naturally gravitate towards monopolies and rent-seeking, since they provide large and stable margins. The technological progress is an accidental byproduct of efficiency drives, which is an occasional byproduct of economic systems in specific historical circumstances.
Capitalism assume that the government uses force to enforce contracts. Without that capital accumulates with those that have military power instead of those that invests, which is much worse.
in a world where production facilities take multiple years to create, but demand increases a ton in less than that time, I would expect high prices for multiple years, with or without monopoly. The existence of a higher price than in the past is simply not evidence of a monopoly nor something going wrong.
I think an investigation is fine but I also think they might actually have a reasonable defense with demand around HBM.
The way this is packaged seems to hinge on the expectation that there isn't some kind of AI revolution under way. Everyone just spontaneously decided to fuck over dram production and instead manufacture nearly impossible assemblies that yield like shit. The semiconductor industry. Famously risk adverse. I really don't buy it.
If the dram price went up and there wasnt something like HBM demand to explain it away, then perhaps we have something more serious to work with.
Sure, courts don't declare people innocent, but rather "not guilty" given the evidence and societal standards of conduct.
It gets a little dubious, when the former CEO of Nvidia is broadcast on international news about making friendly recommendations to SK Hynix to focus on HBM over a consumer market they already hit 90% profit margins on. Then a year later hitting 600% growth in a market circularly financed by Nvidia.
These people ruined a lot of other peoples fun, and while not a crime they still burned their brands goodwill with consumers. In court that will matter whether they like it or not. =3
"Pride goes before destruction, And a haughty spirit before a fall." (proverb 16:18)
Not exactly related but I was thinking why don’t gov make rule like 40% of DRAM production has to be allocated for consumer electronics sectors but no AI regulation, wonder that would make compute distributed and no chicken and chicken man situation
Nor does the "market will take care of it" concept seem to be working that well nowadays but maybe there's a compromise in between extremes that will produce the most good.
No one complained about the “market” when prices cratered from oversupply, putting or nearly putting companies out of business. There is a lead time for new entrants and people are rightfully gun-shy going all in on a commodity that has had many boom/bust cycles.
With that said, the market will sort it out. Either the AI companies run out of capital or more production comes online.
Or it won’t, and this goes on so long that the market will introduce a new option.
rented compute could become so much the norm that owning compute can’t bridge the gap again.
Game consoles are shifting to discless as we speak, so no used game market. Phone leases are a thing now so console leases are probably coming. They maybe even already here if you consider game streaming. In this very thread people are suggesting that renting makes so much more sense than building that it’s almost treated as a questionable practice to own.
But once ownership of compute is no longer tenable, it doesn’t take much for the market to be coerced via money into pushing agendas via marketing. Not just political agendas. "Please drink a verification can" meme, for example.
There will be too much money to be made in leasing compute that it would never make market sense to restore ownership to the people.
> Or it won’t, and this goes on so long that the market will introduce a new option.
Micron had a gross margin of almost 87% last quarter for a commodity. Yes, it takes time to build a factory and money to buy the machines, but memory is a commodity. There is no 'secret' around creating it. That type of margin simply will not last.
What do you prefer, RAM being really expensive or RAM being out of stock?
Of course, it would be nice if there was enough supply to meet demand, but there currently is not, so it has to be one or the other. I still prefer expensive over not available.
If I can't get any RAM then it's immaterial whether it's because I'm poor or because I'm unlucky. But if the second case ensures that (say) research institutions get as good a chance as OpenAI then that's (IMO) better than "we'd love to research your idea but Elon Musk bought all the RAM for Grok".
It wouldn't be down to luck, you just wouldn't be able to buy it. All the RAM would have been sold out about a year ago and you wouldn't have been able to buy any RAM at whatever price and wouldn't be able to for maybe years to come.
You can still buy 64GB DDR5 RAM today. It just costs a little over a thousand dollars. This price discourages enough buyers for there to be stock left.
Of course, this assumes there is no price fixing going on, which hasn't been determined, yet.
Supply goes to those willing to pay the highest price (because they can use them at a high productivity).
RAN has a long lead time, so supply isn’t created overnight. That’s not a “market problem” that a physics problem.
It’s not a “market failure” because some customer can buy it at the price they want. Those customers always exist (people who wouldn’t buy 2 years ago unless it was cheaper)
An unexpected change happens and we get in an adjustment process, and people like you will throw the baby out with the bathwater because you naiively imagine the system has to work perfectly.
> The “planned economy” concept didn’t really work out last time it was tried.
It's currently being tried in France and the results are, unsurprisingly, devastating. France officially entered a recession (two trimesters in a row with negative GDP growth). It's a country where close to 60% of the GDP is public spending: a circular economy where the government confiscates all the money to give it to public servants and public aids, which then goes back to the government doesn't work.
There's an article, the main article, frontpage today on the biggest (leftist of course) french newspaper:
"Le lent declin economique de la France face au reste de l' Europe"
"The slow economical decline of France compared to the rest of Europe"
As a sidenote I'll point out the most of Europe, including the Eurozone, has seen basically zero growth inflation adjusted since the Euro was introduced, while at the same time the GDP of both the US and China have experience tremendous growth.
So that headline is saying: "A planned economy is eff'ing France in an already eff'ed Europe".
A potentially better idea would be to implement profit margin caps for industries with a small amount of competitors. They would become defacto public utilities.
If you cap profit margins, it encourages volume, since the only way to make more money is to increase sales.
Perhaps it could encourage new entrants into the market if a new entrant meant that the entire industry's profit margins were uncapped.
It's pretty easy to create the illusion of no profit though. Look at Hollywood accounting for example. Or how multinationals happen to make no profit in high tax countries. The general technique is to split your entity into multiple parts - the one whose profit is to be minimised pays arbitrary fees to the other, which is in a different jurisdiction.
The funny part here is that the free market supporters also tend to be the people who are against regulations that would help combat price fixing and other similar regulations that would help protect the free market. They usually do this by calling the proposed regulation socialism and the like, and lobby against it accordingly.
Free market supporters being against regulations is just part of being a free market supporter.
Regulations are inherently against a free market. This is not an ideological stance, it's just what the words mean.
Sometimes there is an argument that conditions do not favor a free market and regulation is required to fix that defect, but that's rather situational.
I mean, it depends how you define a "free market" doesn't it? A market filled with monopolies and price fixing schemes is not "free" in my personal experience with it. None of the assumptions about Capitalism (the good assumptions, about supply and demand, new competition being able to enter and undercut, innovation being a way to gain a piece of the pie, etc), can be said to faithfully apply to a market that isn't at least minimally regulated to get rid of price fixing and monopolies and cartels.
I mean, we had this discussion before a very long time ago and then it seems like libertarianism took off in start up tech spaces...who are relying on the lack of monopolies and price fixing specifically for their own businesses to be able to be successful...so I don't think the libertarian attitude of a lot of HNers really fits, given the amount of effort the government of the past has put into making the market actually "free" in a real sense and not just a liberal philosophical sense of it technically being possible for you to start a company to try to enter a business (only to get eaten alive by an entrenched oligopoly, for instance).
In some types of businesses, I think there is also good reason to not want it to be a free market. For instance, the nuclear weapons business. So we accept that the "free market" shouldn't apply to everything already.
I think with chip companies, the danger is that it is incredibly risky and expensive to enter the business, so you need to be very wary and on the lookout for things like price fixing and oligopolies...it is in the interests of the established players to do those things. And we haven't really seen very many new entrants to these markets for perhaps good reason (being very difficult to get the amount of VC needed for hardware startups and general wariness about hardware startups in the VC world)...but also perhaps if there was a price fixing schemes...that would also be a reason we haven't seen any new entrants.
Alright, but let's think about the example that started this. So, we have these RAM manufacturers doing (or alleged to be doing) price fixing. Now, price fixing is considered to be antithetical to a free market system due to subverting both competition and supply and demand.
So let's say we have regulation meant to combat price fixing or perhaps a judgement from a court of law against parties engaging in price fixing, like what happened with DRAM manufacturers 20-ish years beforehand. Well, by the logic you presented, those are also detrimental to a free market. However this leads to a contradiction, as both regulating and not regulating is a detrimental choice. Either you don't regulate, in which case price fixing -- and also we of course have to take into account that a similar argument argument against regulations also works for (unregulated) monopolies, monopsonies, and other sundry cartel actions -- will negatively undermine the free market. On the other hand, regulating price fixing, monopolies, monopsonies etc. will also negatively undermine the free market.
So, we have a logical inconsistency here as presented. So we realistically have three possibilities:
a) the free market itself is logically contradictory. I mean, many socialists in particular have argued for this position, but I at least don't find this consequent very attractive. It _could_ be true, but let's look at the other two.
b) that price fixing is not detrimental to the free market. Well mayhaps, but that really doesn't seem to correspond to the consensus around price fixing and other similar anticompetitive practices. And as said, the same argument could be made for monopolies and such, so you'd basically have to argue that none of them are detrimental to the market either.
c) that regulation isn't necessarily detrimental to the free market. You kinda already conceded this point, as you said
> Sometimes there is an argument that conditions do not favor a free market and regulation is required to fix that defect, but that's rather situational.
And if we assume this, then the question is no longer whether regulation is always bad for the market, but where the line is between regulation that is bad for the market and which is either neutral or positive for the market. But then the words no longer mean what they should mean by surface analysis, and that's a travesty!
So where does that leave us? Well, for myself the answer seems obviously to be that regulation can be non-detrimental to the free market, etymology be damned! And so in this case, at least from where I'm sitting, a free market advocate should be for legislation against price fixing. But they might think differently, in which case they should justify their views.
Not really. Wikipedia defines it as "an economic system in which the prices of goods and services are determined by supply and demand expressed by sellers and buyers. Such markets, as modeled, operate without the intervention of government or any other external authority." That means that sellers are free to do whatever they want.
A free market assumes that there is healthy competition among sellers, so they do not have incentive to collude. In a healthy market, if five sellers colluded to keep the price high, a sixth seller should pop up and undercut the colluding sellers to capture business from the overpriced colluding sellers.
Sure, but that's really naive in practice, isn't it? Because "a new competitor comes in" really isn't as easy as it sounds, especially in chips for example. Getting to the point where you can actually compete is really, really, really difficult and requires a lot of investment (money, time and resources).
Like many aspects of idealized Capitalism, I think it doesn't work in practice nearly as well as it does in theory. Which is why we made laws about this sort of stuff a long time ago. But they've not really been super strictly enforced for a while has been my impression. Perhaps they should be again?
Difficult when Capital is global and the companies making all this stuff are global. But I think they'd accept a US court decision given the centrality of that market...
If the government doesn't intervene, there's no control, so sellers can do whatever they please. Hence price fixing, dumping, monopolies, etc., are all possible in a free market. There's no contradiction.
> [If] price is not set by supply and demand, [...] you do not have a free market
You can't even define what "set by supply and demand" means. Supply and demand are variables in some equation that does not and cannot capture reality.
Having read most of the posts for this subject, I got to thinking: wouldn't it be cool and beneficial to the HN community, if we tried to anticipate equipment price spikes? We tend to be electronic parts focused, so why not post trends as we see them.
Example: I, like I'm sure others did, saw the M.2 SSD prices starting to rise. Even though I didn't see the data center build up coming, I bought a small supply of SSDs. Unfortunately I didn't realize the future memory situation!
If this isn't cool with the Hacker News bylaws, I withdraw this idea :)
You notice the trend, but can you predict how long will it last? What if it ends in a few months and prices go back to "normal" or even less? You made "strategic" reserves of RAM and SSD for nothing.
I think there is a greater argument that points to hoarding of the global supply of these components by three or four cloud companies, which in turn forces consumers to rent their services.
Well apparently the last price fixing conviction was only made after Micron blew a whistle, so if they just keep their lips zipped they're probably good.
They’ll fine the companies and that money goes back into the government coffers. Meanwhile the companies have to jack up prices to pay the fine. A win for everyone!
A good portion of what's on the front page at any given time is LLM output. I think the site explicitly made a decision that this is OK when they banned AI-generated comments but not AI-generated submissions.
I think that most people here just don't mind, or even favor this if they're AI accelerationists. Case in point are all the high-scoring hobby projects that are just "I asked Claude to make me X and then write an article about it".
If they couldn't bother to spend the time to write the writeup about a subject they're supposedly passionate about why should any of us bother spending time reading it?
-- particularly given that it's easy to use AI to do the skut work that it's good at, like digging up citation links, without using it to write which its extremely poor at.
I just find myself instantly closing without reading a large percentage of HN submissions these days.
I read the original complaint in this case when it was filed and I think it's a nothing burger. They admit outright that the reduction in supply is because the suppliers are making HBM instead, where they have infinite demand and are making more money.
Going after the memory suppliers is fashionable because they've gotten in trouble for collusion before-- but the claim's own text really undermines the collusion argument.
A more interesting question is probably: did large AI companies buy out the ram supply using their fountains of investor money in order to suppress competition in AI ... but that isn't a question that will be explored by this case.
If these companies are innocent, they have the best lawyers money can buy and plenty of money to defend themselves. No one here needs to spend their limited time on Earth defending the actions of these corporations for free. nd if they're guilty you shouldn't defend them at all. So don't worry about it, let the corporations defend themselves, folks. The corporations will never, ever defend you.
moribvndvs | 18 hours ago
minraws | 17 hours ago
briffle | 17 hours ago
brendoelfrendo | 16 hours ago
internetter | 15 hours ago
somenameforme | 17 hours ago
Woodi | 16 hours ago
M95D | 16 hours ago
Same here. I'll probably be so old I'd be dead by then.
SubiculumCode | 17 hours ago
Finally secure a research grant and had budgeted money for a new compute workstation and the RAM prices are killing my budget, because I'd love to get at least 256 GB of RAM and that's 4-5 grand now. Ridiculous, especially since I'd like to have a GPU in it with at least 32Gb of VRAM for medical image model training, so that is also like 4-5 grand too.
Do I wait two years and hope, or pay, isthe question I keep asking myself.
mdorazio | 17 hours ago
doc_ick | 16 hours ago
riedel | 16 hours ago
SubiculumCode | 15 hours ago
[1] https://allocations.access-ci.org/resources
thaumasiotes | 13 hours ago
You're building a common infrastructure to address privacy concerns about the use of shared infrastructure??
gunalx | 11 hours ago
Basically make your own hpc provider for academic purposes, and rent out extra capacity to smaller commercial interests to utilise that spare capacity and also prop up the local economy with usable research compute.
subscribed | 11 hours ago
riedel | 5 hours ago
linker3000 | 11 hours ago
moffkalast | 10 hours ago
riedel | 5 hours ago
[0] https://gauss-allianz.de/en/hpc-ecosystem
[1] https://access.eurohpc-ju.europa.eu/
SubiculumCode | 16 hours ago
moffkalast | 10 hours ago
reactordev | 10 hours ago
SubiculumCode | 6 hours ago
woadwarrior01 | 14 hours ago
https://en.wikipedia.org/wiki/DRAM_industry_price_fixing
jareklupinski | 12 hours ago
how many successful careers include 8 months in prison
timoshishi | 11 hours ago
moffkalast | 11 hours ago
KlutzySofa | 12 hours ago
The allegations haven't been proven to be true. Thats what the courts will decide on when the class action continues.
woadwarrior01 | 11 hours ago
KlutzySofa | 10 hours ago
- Cutting DRAM production in 2022 as a result of low prices/demand made sense if they didn't expect the market to recover as quickly as it did. A lot of the recovery is also likely caused by the start of the AI boom.
- The deals OpenAI made with SK Hynix and Samsung was a smart move by OpenAI where they drove up prices for everybody else while securing "cheap" supply themselves. I dont think its clear whether SK Hynix and Samsung came out ahead in that deal as the demand likely would have risen either way and it wouldn't have needed to go up by as much as it has now.
- Moving production to HBM also makes perfect sense when that is what the market demands and is willing to pay a premium for.
cyanydeez | 10 hours ago
It made less sense back when billionaires needed to get on trains and spend hours to collude. Now they just make podcast appearances.
bluGill | 9 hours ago
See the prisoner's dilemma.
zardo | 8 hours ago
baublet | 8 hours ago
matheusmoreira | 8 hours ago
That's a matter of course. This goes way beyond our wallets. They are actively pricing us out of computing. They are deliberately pushing computers back to the big iron mainframe era.
Computers are subversive: they democratized cryptography, copyright infringement, all sorts of things. Now that humanity has developed AI, they will not cannot tolerate mere citizens having access to powerful computers any longer.
If we don't fight this somehow, it's over for us. Everything the word "hacker" ever stood for will be destroyed. We need some kind of second amendment but for computers.
JKCalhoun | 6 hours ago
(Speaking of which—I've seen people suggest a seti@home-like approach to creating a frontier-level, fully open LLM. C'mon, guys, let's vibe code this thing!)
jmercouris | 3 hours ago
JKCalhoun | 56 minutes ago
matheusmoreira | an hour ago
I'm one of them! Every once in a while I ping the community about this idea.
https://news.ycombinator.com/item?id=49077295
https://news.ycombinator.com/item?id=48695813
https://news.ycombinator.com/item?id=37721361
https://news.ycombinator.com/item?id=34536043
GeekyBear | 5 hours ago
> The judge found that long-term HBM lock-ups, refusals of long-term conventional DRAM contracts, Micron's Crucial closure, fab delays, the failure to expand despite record prices, near-identical customer vetting and the companies' prior cartel history supported a plausible inference of conspiracy.
philipallstar | 5 hours ago
GeekyBear | 5 hours ago
They didn't even announce plans to expand production until after the news of this class action lawsuit broke, despite massive DRAM price increases.
thfuran | 3 hours ago
omgwtfbyobbq | 7 hours ago
cmiles74 | 8 hours ago
“To date, five manufacturers have pleaded guilty to their involvement in an international price-fixing conspiracy between July 1, 1998, and June 15, 2002, including Hynix, Infineon, Micron Technology, Samsung, and Elpida.[3]“
KlutzySofa | 6 hours ago
What hasn't happened is that they have been found guilty in the latest class action lawsuit. Which is what this discussion is about.
So far the court has only decided that the lawsuit can proceed, which indicates there is some substance to it.
cmiles74 | 6 hours ago
In my opinion, the fact that this has happened before, and that the companies involved eventually plead guilty, is an important piece of information for people following this story. Especially in light of the statements from the companies involved today.
hn_go_brrrrr | 5 hours ago
cmiles74 | 4 hours ago
martin8412 | 6 hours ago
It is the complete opposite of the current situation.
sdenton4 | 5 hours ago
jscyc | 14 hours ago
sigmoid10 | 13 hours ago
b112 | 12 hours ago
Work for however long, shut down. I do this with linode and their GPU VMs. If you shutdown nightly and on weekends, the costs drop dramatically on a monthly spend.
I only need it a few hours per day, so very cheap for custom compute.
LoganDark | 12 hours ago
b112 | 12 hours ago
The custom volume keeps your persistent install, configs, work in progress, etc.
You can programmatically kick off a linode, attach the volume, then boot, as well as delete, all via curl or python.
So not variable rates, but no rates when the beast is deleted.
LoganDark | 12 hours ago
newComputerWhoT | 4 hours ago
LoganDark | 12 hours ago
probably_wrong | 11 hours ago
The project ran out of budget roughly three months in. From what I heard, status meetings consist of everyone getting together once a month, asking "have we gotten any more budget yet? no?", and then wait until the project runs out.
Cloud providers need to make a profit too, and they've gotten very good at ensuring that their clients forget this point. The whole advantage of cloud compute is that you don't pay for the time you're not using it, but once you reach a certain point you're the one covering the overhead for everybody else.
sega_sai | 11 hours ago
sidewndr46 | 11 hours ago
Tanjreeve | 11 hours ago
c0balt | 5 hours ago
spacedcowboy | 10 hours ago
1: https://github.com/ThrudTheBarbarian/xc
SubiculumCode | 6 hours ago
SubiculumCode | 6 hours ago
thatsabadlook | 12 hours ago
michelb | 12 hours ago
Ygg2 | 10 hours ago
The glut of AI chips will lower prices meaning consumer GPUs start looking high margin. And if they pile into it again, prices will come down.
post-it | 8 hours ago
Kevin O'Leary's 20 GW data centre is never getting built. Maybe he'll do a "phase 1" at a few hundred MW, but he's not going to build the largest power station in the world.
jmyeet | 11 hours ago
What you probably want instead is a Mac Studio with an M5 Ultra. The 256GB model is ~$11k. The problem? Demand is so high that the lead time last I checked was March. But this will give you 256GB of shared memory, well above the 32GB you'll get on a 5090.
If you can live with 96GB (or less) on a Mac Studio it's quite a lot cheaper and the lead times are far quicker.
NekkoDroid | 10 hours ago
Well not really, there is the AMD Radeon AI PRO 9700, which is a RX 9070 (or possibly the XT, dunno) with 32GB VRAM. It does use slower GDDR6, but it has "held its price better" (for some definition of better) and is still available for like 2.3k€ (this includes taxes). Though I remember when I was thinking of buying one when it was 1.4k not too long ago and didn't :( now I just don't bother looking anymore.
reenorap | 9 hours ago
NekkoDroid | 8 hours ago
The RX 7900 XTX has slightly higher bandwidth at almost 1kGBps compared to AI PROs almost 650GBps but has only 24GB instead of 32GB, though I can't find the card on sale anywhere rn.
In the end I see it mostly as a "is the massive jump in price worth it, or do you risk it and maybe get multiple AI 9700s instead and deal with ROCm instead". I can't tell how good ROCm has gotten in regards to training and such, but for inference Vulkan on my 9070 XT is pretty solid.
SubiculumCode | 6 hours ago
numpad0 | 6 hours ago
xnx | 6 hours ago
ericd | 4 hours ago
Depends on what you're doing with it, though, obviously.
sleight42 | 3 hours ago
The market has gotten so tight that antiquated hardware has become competitive. Too many of us are trying to buy the same aging platforms for exactly this reason.
ericd | 2 hours ago
mesm7903 | 17 hours ago
weikju | 14 hours ago
tjpnz | 12 hours ago
sysguest | 17 hours ago
those upgrades ARE ACTUAL UPGRADES, and if exiting ddr4 helps on producing more ddr5, why not?
(and the cheap chinese makers can produce plenty of those ddr4 chips, so it's NOT that "monopoly destroying the market" scenario)
cowlevel | 16 hours ago
sysguest | 16 hours ago
you can just sell the production line and re-purpose human workers for ddr5?
cowlevel | 15 hours ago
There's a story where a fab company tried to open a fab in America but the chips kept coming out broken, and they tried everything and the only thing that worked was disassembling the whole factory and rebuilding it in the exact floor layout of their Taiwanese one.
Sometimes they do salvage one or more machines as both processes use the same one. But that had better be worth shutting down the old process, financially.
defmacr0 | 13 hours ago
rcxdude | 9 hours ago
gbnwl | 16 hours ago
InsidiousWombat | 12 hours ago
On RAM side it's purely "the factories that turn wafers into RAM", and big part of the problem is that HBM just yields less bits per wafer (the number I've heard is between extra size and lower yield, about half) so more of the capacity moved to making HBM also lowers total bit count made
poglet | 15 hours ago
fartfeatures | 14 hours ago
TacticalCoder | 12 hours ago
Something China literally never did though? They seem to love flooding the world with what they produce.
intrasight | 9 hours ago
doom2 | 9 hours ago
Woodi | 16 hours ago
"more akin to hawks flying in formation – an activity that hawks ordinarily do not do"
Just strange nowhere in article DDR5 is mentioned...
tenat10 | 15 hours ago
They could literally reopen manufacturing. But why? They know its free profit through inaction as long as none of them fix consumer DRAM market.
This act of playing dumb and ignorant has never been easier. Getting really old
cbg0 | 14 hours ago
They "literally" couldn't, there's no idle fab. They could redirect manufacturing to make more consumer RAM instead of HBM but why eat into your profits by doing that? The shareholders won't be happy.
I'm not a fan of this situation but unless there's some WhatsApp texts the investigators will find to point to collusion this is dead in the water. The filing itself is based on a lot of circumstantial evidence.
phyzix5761 | 15 hours ago
rf15 | 13 hours ago
Lerc | 12 hours ago
ppierre | 15 hours ago
PeterStuer | 14 hours ago
gpt5 | 10 hours ago
If you really are thinking from a game theory perspective, in order for it to be the prisoner dilemma, they each need to benefit from lowering prices individually (to cheat). But given that demand outstrips supply so much, they each just lose if they lower price, regardless of what the other players do.
vhold | 9 hours ago
bethekidyouwant | 9 hours ago
lkey | 8 hours ago
Failure to enforce this kind of thing gives us shortages of eggs and insulin too; It is a real source of social harm.
sophrosyne42 | 7 hours ago
These things are complicated, and people act like the mere existence of increased prices is evidence of malpractice. This is a naiive and juvenile attitude.
lkey | 4 hours ago
https://en.wikipedia.org/wiki/DRAM_industry_price_fixing has been linked elsewhere, but I'll link it again.
It isn't 'seeing the future', it is exercising your power in the present to control what futures are possible.
Like, for instance, forcing your customers to buy more chips than they can install, lest they be cut off entirely. Or increasing hypothetical demand artificially by breaking international smuggling laws using your girlfriend's shell company.
It is not 'naiive' (or naïve, even) to remember previous instances of crimes committed by exactly these companies using exactly the same tactics.
Unregulated markets create these predatory incentives, it shouldn't shock your or anyone that this has come to pass again.
jayd16 | 6 hours ago
sophrosyne42 | 7 hours ago
jayd16 | 6 hours ago
toast0 | 3 hours ago
If the AI bubble hadn't come around buying up all the RAM, most new ram purchased today would be DDR5. The shift towards selling more HBM is also probably a bit unexpected.
braiamp | 9 hours ago
Note, it's not just one thing, it's a combination of practices that make the collusion more likely explanation than otherwise
doom2 | 9 hours ago
If they had started building new fabs earlier, they might have been able to sell even more inventory _and_ gotten back into the consumer market that they now neglect. Instead they've been way too slow to respond to demand or, in this case, purposefully didn't respond to demand.
gpt5 | 9 hours ago
reenorap | 9 hours ago
lokar | 8 hours ago
OAI would be doing the same.
reenorap | 8 hours ago
lokar | 6 hours ago
kasabali | 7 hours ago
sophrosyne42 | 7 hours ago
jayd16 | 6 hours ago
The sense of scarcity factors into the ability to negotiate these contracts. Cooking the reports has a direct impact on contracts and does increase real demand.
jeremyjh | 6 hours ago
londons_explore | 4 hours ago
jeremyjh | 4 hours ago
ketzu | 14 hours ago
Nearly everything could have a non-colluding explanation depending on the tech and market. (for example I would be skeptical to build up extremly expensive production if i follow the widespread belief that Ai is a huge bubble)
However, collusion is not magically out of the question with previous instances and some unexplained indicators. (which imo is what this dismissal means)
OroPla | 12 hours ago
ketzu | 12 hours ago
If you think demand will break down before you can get a return on your investemtn, you should not build.
If you think of a different aspect of "the AI bububle popping", where demand for RAM does not significantly decrease, that's fine of course. But RAM demand by AI going down massively seems to be a major part of what a lot of people imagine when they talk about the bubble going away.
InsidiousWombat | 12 hours ago
Even if we got back to pre-AI bubble, hell, especially if we got back to pre-ai bubble pricing, we'd see a lot of push for local AI and 32 GB+ GPUs/128GB+ SOC systems
ketzu | 11 hours ago
I put up as an example and prefaced it with me not knowing enough about the market and tech. I have not done the calculations, because I have no access to any of the relevant numbers, do not have a relevant model of the market dynamics I would expect, and do not have to suffer the losses if it doesn't play out.
It is great that you have more knowledge to be able to confidently state those things, however, I have to say I have no reason to actually believe you do.
OroPla | 9 hours ago
A bubble means something is overvalued. The stock prices will go down if the bubble pops, but not the demand.
ketzu | 6 hours ago
One example of people not sharing your definition are the many comments of the form "can't wait for the Ai bubble to burst and ram/gpus getting affordable again".
Also I am not sure about your examples. Same as the other commenter: I have no idea about the rail bubble. But I believe the dot-com bubble burst did in fact reduce demand on compute infrastructure. (Not long term of course.) eg https://www.theregister.com/on-prem/2002/01/11/server-sales-...
OroPla | 4 hours ago
Yahoo kind of went away, but Google replaced them. I really don't understand why people would think that anything would change even if all the known AI companies went under tomorrow. We are still going to use AI more and more and we are still needing more and more hardware for that. Doesn't matter which company it is that is involved in all that.
> Not long term of course
Building more fabs is very much a long term thing, though.
RunSet | 7 hours ago
Neither does its being here to stay does not mean it is good or desirable.
In any case "It is here to stay." does not resemble a falsifiable claim. If you were wrong how could you tell?
Grombobulous | 5 hours ago
Passenger rail versus freight rail also demonstrates how the railroad bubble and aftermath ended up proving that not every use of rail is inherently profitable. In the long term, passenger rail is rarely profitable and almost always has to be subsidized by the state in some way with only a handful of exceptions.
OroPla | 4 hours ago
itopaloglu83 | 52 minutes ago
shevy-java | 14 hours ago
I also think mandatory jail times for the CEOs need to happen. It can not be that they can buy their way out of punishments and damage they cause against society.
toyg | 14 hours ago
Jensson | 11 hours ago
OroPla | 12 hours ago
Capitalism is defined by:
- the means of production (factories, tools, employees, land,..)
- being owned privately (as opposed to being state owned)
- for profit (so you do not have to live paycheck to paycheck)
A free market is defined by:
- price of goods is set by supply and demand
- the government doesn't intervene (for example through subsidies)
Price fixing is defined by:
- all the sellers (or buyers) agreeing on and therefore setting the price
- price is not set by supply and demand, therefore you do not have a free market
sophrosyne42 | 6 hours ago
beloch | 14 hours ago
bob1029 | 14 hours ago
The way this is packaged seems to hinge on the expectation that there isn't some kind of AI revolution under way. Everyone just spontaneously decided to fuck over dram production and instead manufacture nearly impossible assemblies that yield like shit. The semiconductor industry. Famously risk adverse. I really don't buy it.
If the dram price went up and there wasnt something like HBM demand to explain it away, then perhaps we have something more serious to work with.
Joel_Mckay | 9 hours ago
It gets a little dubious, when the former CEO of Nvidia is broadcast on international news about making friendly recommendations to SK Hynix to focus on HBM over a consumer market they already hit 90% profit margins on. Then a year later hitting 600% growth in a market circularly financed by Nvidia.
These people ruined a lot of other peoples fun, and while not a crime they still burned their brands goodwill with consumers. In court that will matter whether they like it or not. =3
"Pride goes before destruction, And a haughty spirit before a fall." (proverb 16:18)
badgersnake | 14 hours ago
born-jre | 14 hours ago
andrewstuart | 13 hours ago
Kirth | 13 hours ago
bajard | 13 hours ago
matwood | 13 hours ago
With that said, the market will sort it out. Either the AI companies run out of capital or more production comes online.
cheschire | 12 hours ago
rented compute could become so much the norm that owning compute can’t bridge the gap again.
Game consoles are shifting to discless as we speak, so no used game market. Phone leases are a thing now so console leases are probably coming. They maybe even already here if you consider game streaming. In this very thread people are suggesting that renting makes so much more sense than building that it’s almost treated as a questionable practice to own.
But once ownership of compute is no longer tenable, it doesn’t take much for the market to be coerced via money into pushing agendas via marketing. Not just political agendas. "Please drink a verification can" meme, for example.
There will be too much money to be made in leasing compute that it would never make market sense to restore ownership to the people.
matwood | 6 hours ago
Micron had a gross margin of almost 87% last quarter for a commodity. Yes, it takes time to build a factory and money to buy the machines, but memory is a commodity. There is no 'secret' around creating it. That type of margin simply will not last.
OroPla | 12 hours ago
Of course, it would be nice if there was enough supply to meet demand, but there currently is not, so it has to be one or the other. I still prefer expensive over not available.
probably_wrong | 11 hours ago
OroPla | 8 hours ago
You can still buy 64GB DDR5 RAM today. It just costs a little over a thousand dollars. This price discourages enough buyers for there to be stock left.
Of course, this assumes there is no price fixing going on, which hasn't been determined, yet.
newswasboring | 11 hours ago
InsidiousWombat | 12 hours ago
But since Bell systems breakup billionaire caste have been heavy at work to never allow government to wield that kind of power
refurb | 12 hours ago
Supply goes to those willing to pay the highest price (because they can use them at a high productivity).
RAN has a long lead time, so supply isn’t created overnight. That’s not a “market problem” that a physics problem.
It’s not a “market failure” because some customer can buy it at the price they want. Those customers always exist (people who wouldn’t buy 2 years ago unless it was cheaper)
sophrosyne42 | 7 hours ago
newsclues | 13 hours ago
noosphr | 13 hours ago
Communism has brought more people out of poverty in the last 100 years than capitalism did in 500.
TacticalCoder | 12 hours ago
It's currently being tried in France and the results are, unsurprisingly, devastating. France officially entered a recession (two trimesters in a row with negative GDP growth). It's a country where close to 60% of the GDP is public spending: a circular economy where the government confiscates all the money to give it to public servants and public aids, which then goes back to the government doesn't work.
There's an article, the main article, frontpage today on the biggest (leftist of course) french newspaper:
https://www.lemonde.fr/economie/article/2026/11/15/le-lent-d...
"Le lent declin economique de la France face au reste de l' Europe"
"The slow economical decline of France compared to the rest of Europe"
As a sidenote I'll point out the most of Europe, including the Eurozone, has seen basically zero growth inflation adjusted since the Euro was introduced, while at the same time the GDP of both the US and China have experience tremendous growth.
So that headline is saying: "A planned economy is eff'ing France in an already eff'ed Europe".
dieortin | 11 hours ago
> while at the same time the GDP of both the US and China have experience tremendous growth
On the other hand, China is a planned economy
sophrosyne42 | 7 hours ago
hahn-kev | 12 hours ago
layer8 | 11 hours ago
gus_massa | 12 hours ago
ion_trapper | 8 hours ago
ozgrakkurt | 7 hours ago
Not fighting on the other side of the world would be a good starting move to avoid increasing energy prices.
m4rtink | 5 hours ago
Donwe let them buy 40% of food production next, so that they can sell it back at insane prices to starving people ?
Grombobulous | 5 hours ago
If you cap profit margins, it encourages volume, since the only way to make more money is to increase sales.
Perhaps it could encourage new entrants into the market if a new entrant meant that the entire industry's profit margins were uncapped.
IshKebab | 4 hours ago
andrewstuart | 14 hours ago
……… “denied”
amelius | 12 hours ago
OroPla | 12 hours ago
sham1 | 11 hours ago
breakwaterlabs | 10 hours ago
Regulations are inherently against a free market. This is not an ideological stance, it's just what the words mean.
Sometimes there is an argument that conditions do not favor a free market and regulation is required to fix that defect, but that's rather situational.
retwakm4 | 10 hours ago
I mean, we had this discussion before a very long time ago and then it seems like libertarianism took off in start up tech spaces...who are relying on the lack of monopolies and price fixing specifically for their own businesses to be able to be successful...so I don't think the libertarian attitude of a lot of HNers really fits, given the amount of effort the government of the past has put into making the market actually "free" in a real sense and not just a liberal philosophical sense of it technically being possible for you to start a company to try to enter a business (only to get eaten alive by an entrenched oligopoly, for instance).
In some types of businesses, I think there is also good reason to not want it to be a free market. For instance, the nuclear weapons business. So we accept that the "free market" shouldn't apply to everything already.
I think with chip companies, the danger is that it is incredibly risky and expensive to enter the business, so you need to be very wary and on the lookout for things like price fixing and oligopolies...it is in the interests of the established players to do those things. And we haven't really seen very many new entrants to these markets for perhaps good reason (being very difficult to get the amount of VC needed for hardware startups and general wariness about hardware startups in the VC world)...but also perhaps if there was a price fixing schemes...that would also be a reason we haven't seen any new entrants.
sham1 | 9 hours ago
So let's say we have regulation meant to combat price fixing or perhaps a judgement from a court of law against parties engaging in price fixing, like what happened with DRAM manufacturers 20-ish years beforehand. Well, by the logic you presented, those are also detrimental to a free market. However this leads to a contradiction, as both regulating and not regulating is a detrimental choice. Either you don't regulate, in which case price fixing -- and also we of course have to take into account that a similar argument argument against regulations also works for (unregulated) monopolies, monopsonies, and other sundry cartel actions -- will negatively undermine the free market. On the other hand, regulating price fixing, monopolies, monopsonies etc. will also negatively undermine the free market.
So, we have a logical inconsistency here as presented. So we realistically have three possibilities:
a) the free market itself is logically contradictory. I mean, many socialists in particular have argued for this position, but I at least don't find this consequent very attractive. It _could_ be true, but let's look at the other two.
b) that price fixing is not detrimental to the free market. Well mayhaps, but that really doesn't seem to correspond to the consensus around price fixing and other similar anticompetitive practices. And as said, the same argument could be made for monopolies and such, so you'd basically have to argue that none of them are detrimental to the market either.
c) that regulation isn't necessarily detrimental to the free market. You kinda already conceded this point, as you said
> Sometimes there is an argument that conditions do not favor a free market and regulation is required to fix that defect, but that's rather situational.
And if we assume this, then the question is no longer whether regulation is always bad for the market, but where the line is between regulation that is bad for the market and which is either neutral or positive for the market. But then the words no longer mean what they should mean by surface analysis, and that's a travesty!
So where does that leave us? Well, for myself the answer seems obviously to be that regulation can be non-detrimental to the free market, etymology be damned! And so in this case, at least from where I'm sitting, a free market advocate should be for legislation against price fixing. But they might think differently, in which case they should justify their views.
amelius | 9 hours ago
baublet | 8 hours ago
tgv | 11 hours ago
mtlynch | 11 hours ago
Systemerror7A69 | 11 hours ago
So it doesn't just work like that in practice.
retwakm4 | 10 hours ago
Difficult when Capital is global and the companies making all this stuff are global. But I think they'd accept a US court decision given the centrality of that market...
OroPla | 9 hours ago
- price of goods is set by supply and demand
- the government doesn't intervene (for example through subsidies)
Price fixing is defined by:
- all the sellers (or buyers) agreeing on and therefore setting the price
- price is not set by supply and demand, therefore you do not have a free market
tgv | 8 hours ago
If the government doesn't intervene, there's no control, so sellers can do whatever they please. Hence price fixing, dumping, monopolies, etc., are all possible in a free market. There's no contradiction.
> [If] price is not set by supply and demand, [...] you do not have a free market
You can't even define what "set by supply and demand" means. Supply and demand are variables in some equation that does not and cannot capture reality.
OroPla | 4 hours ago
None of this is meant to capture reality. In reality the market is not free. The government very much intervenes.
amelius | 9 hours ago
bloomingeek | 10 hours ago
Example: I, like I'm sure others did, saw the M.2 SSD prices starting to rise. Even though I didn't see the data center build up coming, I bought a small supply of SSDs. Unfortunately I didn't realize the future memory situation!
If this isn't cool with the Hacker News bylaws, I withdraw this idea :)
Xenoamorphous | 9 hours ago
bloomingeek | 2 hours ago
niltecedu | 9 hours ago
rationalist | 9 hours ago
bloomingeek | 2 hours ago
lotsofpulp | 9 hours ago
chihuahua | 7 hours ago
ciupicri | 26 minutes ago
bparsons | 9 hours ago
bagacrap | 9 hours ago
taxmantakes | 9 hours ago
lpghatguy | 8 hours ago
uptrails | 7 hours ago
I think that most people here just don't mind, or even favor this if they're AI accelerationists. Case in point are all the high-scoring hobby projects that are just "I asked Claude to make me X and then write an article about it".
ultrarunner | 7 hours ago
nullc | 6 hours ago
-- particularly given that it's easy to use AI to do the skut work that it's good at, like digging up citation links, without using it to write which its extremely poor at.
I just find myself instantly closing without reading a large percentage of HN submissions these days.
beej71 | 6 hours ago
I flag all such articles. I can generate them myself if I want them.
nullc | 6 hours ago
Going after the memory suppliers is fashionable because they've gotten in trouble for collusion before-- but the claim's own text really undermines the collusion argument.
A more interesting question is probably: did large AI companies buy out the ram supply using their fountains of investor money in order to suppress competition in AI ... but that isn't a question that will be explored by this case.
dev1ycan | 5 hours ago
Shellban | 5 hours ago
burnte | an hour ago