No projections/information/analysis. The statement from someone with a vested interest in this demand lasting an extended period of time (including the statement that demand will continue past 2030).
I remember in the 90's there was some sort of shortage reportedly due to fires. In the end it was found the shortage was due to collusion by the manufacturers.
After the first two times, do they even need to collude?
Can't they count on each other doing the profitable thing? Especially when they PR it:
> Furthermore, focused investment in HBM is creating a virtuous cycle that improves profitability in the general-purpose memory market. As resources are diverted to HBM, the supply-demand balance for general DRAM is improving.
No because, some companies know they'll need ram, but they may have some leeway on timing (i.e. doing more to prolong the life of systems they'd typically replace). That means to guarantee ram they need to enter an agreement to buy the ram now. By setting the expectation that these prices continue or get worse companies that may have put off agreeing to these terms locking in prices based on demand rather than something closer to the cost of production. The agreements that these companies are being forced to make are upfront payment and then a top up to meet current market prices before delivery, there is no downside for these ram companies whether the price rises or falls (at worst they end up with supply that was already payed for by a bust company at a healthy margin and a glut in the market that would arrive anyway)
Competitors are going to come online anyway because of the threat of chip embargoes against China that mean they are investing in production regardless of what the outside world is doing.
> If demand is projected continue for a long time, this may encourage more competitors to enter this space, so supply will increase.
There are no competitors that can muster the capital to make decent RAM... other than China but China is under a heavy sanctions and tariffs regime, when Apple mentioned using Chinese RAM to ease pricing they were quickly shushed by above.
The current monopolists aren't building new capacity either or are simply pretending to do it. It's a planned monopoly market for reasons I can only speculate about.
Aren't they printing unlimited money from this? I'm sure the chip shortage lasts as long as they're capable of extending it without falling to competitors.
I am confident it will be their last big payday for the Cartel the Chinese certainly will design around it and move on and probably will go onto dominate the worldwide market after 2028. Afterwards Micron like Intel will go around begging for support.
China will eventually do similar things if they're allowed to become a unitary supplier, but while they're establishing their foothold obviously they can't.
Whom exactly? Korea and Taiwan are already saturated. EU may take 5 yearrs just to approve plans formnew factory. US is interested in super fast super duper AIG, ddr5 is way too boring.
Lawyers/politicians in the EU wait years before they do anything usually it’s after the damage has already been done. The only thing they will do is issue a fine after the fact.
DDR5 might be boring, but the high prices are strangling the tech market, including AI. It can’t keep this up much longer. Someone needs to step up and make more. I’m surprised the USA hasn’t offered money to TSMC, Intel, or Micron to expand manufacturing under the CHIPS act or whatever to help ease this shortage.
All the players are investing hundreds of billions of dollars (US) into new fabs. But SK Hynix said customers are signing 5 year contracts which used to be unheard of.
In my opinion, Apple will design around the memory cartel. They possess the financial resources, design, and engineering capabilities in-house (Apple Silicon).
Apple acquired Infineon from Intel, which was struggling, and developed an in-house modem. They certainly have the expertise to design memory and SSDs (PA Semi, Intrinsity, Anobit).
I believe Apple must implement this strategy if they intend to build their devices, and advance into the future in a timely manner, and Apple has demonstrated its ability to execute over time when necessary. This is undoubtedly one of those critical times.
I anticipate that the memory cartel will eventually experience its final significant financial gain, and those who can will design around them.
Yes, it will take time 2 to 4 years but that is nothing took 13 years to replace Intel and six years to replace Qualcomm which has already warned its investors on Apple, leaving them.
With RAM so profitable it makes sense to invest in new factories to build it. This should in time increase supply and decrease price but it takes a good 5 years for a new RAM fab to start producing.
Do you have a source for more info? That seems odd to me - LLMs are particularly resilient to bitflips, so if anything the demand would pull on DDR5 in general, not on ECC specifically.
>There is plenty of DDR5, its not even very expensive,
My 2x16GB of DDR5-6000 that I bought for 250€ is now 650€. 2x32GB that cost 200€ last July now goes for close to 1250€. Even my shitty ass DDR4-3000 sticks from over ten years ago would be selling for close to 300€.
DDR5 is horribly expensive, the only variants that are not as expensive are low frequencies or low amounts.
Honestly surprised by the degree of skepticism about this on HN.
The rack assemblers like Dell, HPE, Supermicro, and Lenovo have 18 to 24 month backlogs on deliveries at the moment, and even the neoclouds have 6-9 month backlogs to onboard customers onto their SaaS platforms.
Also, 2030 is barely 3 years away from now. It's almost September 2026, so 2027 is around the corner. Yes we are ALL old now. Using priors from 2015-2019 is the equivalent of thinking your experience in 1995-1999 should inform decisions made in 2008-2013.
You've been here awhile, this forum is default incredibly skeptical and critical, especially when it comes to AI. It has been to their predictive detriment for many years now. AI bubble conviction is over 2 years now, $300b NVDA profits since. AI tool adoption has completely encompassed any tech savvy office and isn't going anywhere. Broken records.
I think that a lot of people miss key aspects of the AI boom. Even discounting the skeptics, and the bubblers, crashers, etc.
There are a bunch of things that happen in parallel to the AI boom:
a) everyone and their mother is building out capacity. For each GB of VRAM installed, you generally want at least 1 GB of RAM, if not more. So just take nvda's numbers and go from there. And that's just for the GPU servers. But you also need ancillary services around GPU farms. You need some compute for filtering, preprocessing, you need data storage, and so on. And they're all RAM hungry processes.
b) RL is giving lots of capabilities improvements for AI, but it needs lots and lots of parallel runners for scenarios. Those runners need RAM. So every bit of non-GPU capacity is also being used, and more installed.
c) With increasing capabilities comes increasing usage. Everyone is deploying "agents" and those need to run somewhere, and they take RAM (especially the badly coded ones). There've also been reports of labs / big players hoovering up every mac available, for "agentic computer control" / training / whatever. That's on top of people buying them to run "24/7" stuff a la claw/hermes/etc.
In all of this, RAM is the bottleneck. And for the skeptics, you don't have to take hynix's word for it. Just look at the sales for CPUs / Mobos. They're down, because building a computer now is bottlenecked at RAM.
I think it will be their last big payday. I think they’re forgetting that you can design around them. It just takes 2 to 4 years. It’s not good business in the long run to piss off your customers cause you may not get them back.
Who are those ram bottleneck skeptics you’re referring to? I cannot see them. I don’t know anyone who isn’t aware the AI rollout is responsible for years to come of bottleneck of HBM and other hardware
Just see the top of this thread. The theme of the comments is that the SK Hynix, Micron, etc CEOs are lying about this in order to collectively increase prices artificially.
I mean, they do. But that doesn’t mean there isn’t a bottleneck. They control the supply and are taking advantage of the situation to increase their margin as much as they can while limiting their expansion (they have some valid arguments though, given the memory cycles). And it is for sure to their advantage to have the industry accept their high margins for years to come.
I read the threads and don’t see unwarranted skepticism
They are they are taking future IOU from AI companies and presenting that as an actual sale and telling the other people/customers who need memory right away, you have to pay more.
Some of the backlogs apart from the "top level" one are not being viewed that positively from what I have heard. One of my football friends works at Vertiv's India office, soon transferring to Thane, they basically sell cooling infra for these hyperscalers. They have an insane amount of backlogs but because they are a "spoke" component, the backlog is super volatile. Like some power generation related delay or permit issue will randomly de-scale or cancel projects. And its really difficult to calculate risk of all that and maintain the book.
Yep. DCs are extremely capex heavy, so these backlogs can make or break entire projects. That's why tax holidays and subsidizes have become so common for these projects.
Those skeptical are anticipating a possible large drop in demand before that backlog clears. ie. a change in market conditions that means that many of the orders in that backlog would be cancelled or belong to bankrupt entities.
Many words have already been spilled as to whether or not that might happen, so I'll just shrug rather than adding to them.
I’m surprised by the skepticism too. You need luck no further the Chinese who certainly will do it and I think one company in America will do it too much is riding on getting the right products out in time.
The Chinese will use this stupidity by the memory cartel to get into the worldwide memory market it is a perfect entry point and they will sell to anyone who wants to buy no strings attached.
All you need to do is look at the way they do business across the world. They don’t want to be your mama and they don’t wanna be your daddy. They just want to close a business deal. The art of selling and trading is a hell of a lot older in China than the so-called communist party which only dates from 1921 in China. In short labeling them communist is very shortsighted The culture is thousands of years old.
Every time China visits, we get a hospital; every time the West visits, we get a lecture. - some African dude.
Honestly I don't mind the Chinese approach - at least you don't get bullied, unlike dealing with America. Of course, experiences vary if you're in their sphere of influence.
Clearly it’s the fabrication of RAM/HBM chips. But if the margin is so high, why can’t other fabs like CPU fabs be used? I assume they’re not as CPU prices are going down. Or are they radically different and not easily reconfigured?
Given the possibility of setting up new memory fab plants takes 10 years, given that data companies are spending money on power plants based on power consumption forecasts, but are not spending on memory fab plants, I take it that data companies are seeing the current memory makers as sufficient for their needs.
Meaning if the avg consumer is going to buy memory, they are going to have to fork a whole lot more money than they did before, because they will be paying the price of profit generating data companies.
The next phase of improvements will have to be in consumer software to be more memory efficient. Gone are the days of electron apps and other memory intensive systems of programming./
It takes $10-$20 billion dollars and 2 to 4 years to set up a foundry, TSMC took about four years to set up their location in Arizona and there is a hell of a lot more involved with what they do. It’s not impossible to do it in less time the Chinese certainly will.
If a company is motivated and has the money they can get it done. I think those three memory companies are counting on people to be stupid.
Intel, Nvidia, and Qualcomm certainly thought they had Apple over the barrel…
That's like 3 and a half years away? I can see that - fabs don't get built overnight and bubbles aren't under any obligation to pop on a timeline. Wild that 2030 is not that far away. Sometimes it feels like we never left 2020.
There's still a part of me that thinks 2020 will be the bright shiny future that will solve all our problems, but the other parts know that 2020 showed us that we live in the dark future where all our dreams died.
66 comments
[ 0.23 ms ] story [ 6.4 ms ] threadhttps://share.google/aimode/ewJpwCHUINOHzjBVg
https://www.polygon.com/ram-manufacturers-sued-supply-price-...
Can't they count on each other doing the profitable thing? Especially when they PR it:
> Furthermore, focused investment in HBM is creating a virtuous cycle that improves profitability in the general-purpose memory market. As resources are diverted to HBM, the supply-demand balance for general DRAM is improving.
January 5, 2026 - SK hynix market outlook.
https://news.skhynix.com/en/2026-market-outlook-focus-on-the...
If demand is projected continue for a long time, this may encourage more competitors to enter this space, so supply will increase.
On the other hand if everyone thought that the demand will end soon, nobody would invest in more capacity and shortage would last longer.
I think this is just for short term shareholder profit.
Competitors are going to come online anyway because of the threat of chip embargoes against China that mean they are investing in production regardless of what the outside world is doing.
There are no competitors that can muster the capital to make decent RAM... other than China but China is under a heavy sanctions and tariffs regime, when Apple mentioned using Chinese RAM to ease pricing they were quickly shushed by above.
The current monopolists aren't building new capacity either or are simply pretending to do it. It's a planned monopoly market for reasons I can only speculate about.
That's the problem with monopolies.
It's ram chips, the bottom of the barrel of semi manufacturing. If you get high prices for long enough, everyone is gonna jump into the market.
Whom exactly? Korea and Taiwan are already saturated. EU may take 5 yearrs just to approve plans formnew factory. US is interested in super fast super duper AIG, ddr5 is way too boring.
Only competitor is mainlaind china!
Exactly. China wil push for it as China factories will supply Chinese AI first.
Who is going to build a memory chip fab in Europe? Qimonda collapsed into insolvency 20+ years ago and Infineon has long since abandoned the market.
dude, DDR6 was planned for 2027 in 07/2025 [1] No one is going to build factories in 2026 to produce DDR5.
[1] https://www.techpowerup.com/339178/ddr6-memory-arrives-in-20...
Apple acquired Infineon from Intel, which was struggling, and developed an in-house modem. They certainly have the expertise to design memory and SSDs (PA Semi, Intrinsity, Anobit).
I believe Apple must implement this strategy if they intend to build their devices, and advance into the future in a timely manner, and Apple has demonstrated its ability to execute over time when necessary. This is undoubtedly one of those critical times.
I anticipate that the memory cartel will eventually experience its final significant financial gain, and those who can will design around them.
Yes, it will take time 2 to 4 years but that is nothing took 13 years to replace Intel and six years to replace Qualcomm which has already warned its investors on Apple, leaving them.
Fixed it for you.
Dell is making moves to introduce server grade hardware without ECC memory.
There is plenty of DDR5, its not even very expensive, its DDR5-ECC where the prices go stratospheric.
My 2x16GB of DDR5-6000 that I bought for 250€ is now 650€. 2x32GB that cost 200€ last July now goes for close to 1250€. Even my shitty ass DDR4-3000 sticks from over ten years ago would be selling for close to 300€.
DDR5 is horribly expensive, the only variants that are not as expensive are low frequencies or low amounts.
The rack assemblers like Dell, HPE, Supermicro, and Lenovo have 18 to 24 month backlogs on deliveries at the moment, and even the neoclouds have 6-9 month backlogs to onboard customers onto their SaaS platforms.
Also, 2030 is barely 3 years away from now. It's almost September 2026, so 2027 is around the corner. Yes we are ALL old now. Using priors from 2015-2019 is the equivalent of thinking your experience in 1995-1999 should inform decisions made in 2008-2013.
There are a bunch of things that happen in parallel to the AI boom:
a) everyone and their mother is building out capacity. For each GB of VRAM installed, you generally want at least 1 GB of RAM, if not more. So just take nvda's numbers and go from there. And that's just for the GPU servers. But you also need ancillary services around GPU farms. You need some compute for filtering, preprocessing, you need data storage, and so on. And they're all RAM hungry processes.
b) RL is giving lots of capabilities improvements for AI, but it needs lots and lots of parallel runners for scenarios. Those runners need RAM. So every bit of non-GPU capacity is also being used, and more installed.
c) With increasing capabilities comes increasing usage. Everyone is deploying "agents" and those need to run somewhere, and they take RAM (especially the badly coded ones). There've also been reports of labs / big players hoovering up every mac available, for "agentic computer control" / training / whatever. That's on top of people buying them to run "24/7" stuff a la claw/hermes/etc.
In all of this, RAM is the bottleneck. And for the skeptics, you don't have to take hynix's word for it. Just look at the sales for CPUs / Mobos. They're down, because building a computer now is bottlenecked at RAM.
I read the threads and don’t see unwarranted skepticism
There is a lot of SPV backed financial engineering upstream, but not here.
Many words have already been spilled as to whether or not that might happen, so I'll just shrug rather than adding to them.
FTFY. At the rate China's moving on this, it won't be long till they get a stranglehold on the consumer chip market.
All you need to do is look at the way they do business across the world. They don’t want to be your mama and they don’t wanna be your daddy. They just want to close a business deal. The art of selling and trading is a hell of a lot older in China than the so-called communist party which only dates from 1921 in China. In short labeling them communist is very shortsighted The culture is thousands of years old.
Honestly I don't mind the Chinese approach - at least you don't get bullied, unlike dealing with America. Of course, experiences vary if you're in their sphere of influence.
Clearly it’s the fabrication of RAM/HBM chips. But if the margin is so high, why can’t other fabs like CPU fabs be used? I assume they’re not as CPU prices are going down. Or are they radically different and not easily reconfigured?
Meaning if the avg consumer is going to buy memory, they are going to have to fork a whole lot more money than they did before, because they will be paying the price of profit generating data companies.
The next phase of improvements will have to be in consumer software to be more memory efficient. Gone are the days of electron apps and other memory intensive systems of programming./
If a company is motivated and has the money they can get it done. I think those three memory companies are counting on people to be stupid.
Intel, Nvidia, and Qualcomm certainly thought they had Apple over the barrel…
There's still a part of me that thinks 2020 will be the bright shiny future that will solve all our problems, but the other parts know that 2020 showed us that we live in the dark future where all our dreams died.