> CXMT is in the process of converting wafer capacity equivalent to about 20 percent of its total DRAM output — some 60,000 wafers per month — at its Shanghai plant to the fourth-generation HBM3 chip production
As a outside observer, NAND and DRAM prices have skyrocket ed with the AI infrastructure boom just as the China-based fabs are coming online.
It is wise for these Chinese fabs to eventually use a very aggressive dumping strategy to price well below cost push out other players forever, especially in DRAM.
But right now it seems they can max out their supply capacity without selling below cost.
Appears to me like China's endless state led (often unproductive) investment in semiconductor manufacturing subsidies (for decades) is about to pay off with some industry dominance soon.
> It is wise for these Chinese fabs to eventually use a very aggressive dumping strategy to price well below cost push out other players forever, especially in DRAM.
Crucial's departure from the consumer market left such a gaping hole, that CXMT doesn't even need to push other players out to gain a footing.
The downside in general is that other countries lose production capacity in steel, heavy industry, semiconductors, machine tools etc - industries that took decades to build and can't be easily replaced.
Also they gradually lose the ability to meaningfully innovate in those sectors because there's no grounding against production reality anymore.
This has geopolitical consequences further down the line.
> The downside in general is that other countries lose production capacity in steel, heavy industry, semiconductors, machine tools etc - industries that took decades to build and can't be easily replaced.
That's not really what happens though. You don't actually "lose" capacity, you just move to higher-valued special niches within the overall industry because (1) you can afford to, while low-cost competitors can't and (2) you can no longer expect to be the lowest-cost supplier for the bulk of the market. That's a win-win development and something to be encouraged.
> You don't actually "lose" capacity, you just move to higher-valued special niches within the overall industry
That's not what people mean by "lose" capacity.
Suppose DRAM companies expand capacity because prices are high, then demand levels off, the price crashes, and they all go out of business except for the one in China which gets a government bailout. That's fine, right? We're not interested in making DRAM, that's a fungible commodity, we want to make iPhones or something. (They make those too anymore, but never mind that.)
What happens now if China restricts what you can buy to give an advantage to their own companies who are trying to displace you in the higher-valued special niches? Or just raises the price for you and not them? What if there's a trade war? Or a conventional war?
When you still have a domestic industry, you go to them and have a source for the commodity. If only one country becomes the sole global supplier and that country isn't even particularly friendly, that's bad.
The domestic industry is still there, only instead of mass-market DRAM it has started making higher-valued varieties of the same stuff. If there's a trade war, they can easily reconvert to making the mass-market stuff, just at much higher cost. You can't expect more than that, since they never really were as big or as low-cost as the lowest cost suppliers can be in normal times. That's not "losing" capacity, it's just acknowledging that you can't create capacity out of thin air.
Different types of DRAM can literally be made from the same already-etched wafer. The DRAM bits themselves don't change at all. What's different between DDR4, DDR5, and HBM is the IO interface to the chip. Changing this does not require significant retooling or relearning.
As though moving production to China wasn’t something the West did intentionally.
And now continues to push manufacturing out of Western countries by, for example in the UK and Germany, and Australia too, making electricity and gas so expensive it becomes cost prohibitive to manufacture much at all.
> steel, heavy industry, semiconductors, machine tools
the question is if single country can carry all these industries at loss for prolonged period of time.
Another approach is to rely on international supply chain and speed of innovation, we can't produce steel domestically profitably today, fine, we may buy it from diversified international supplier network, and rebuild it fast tomorrow if needed using new tech, and focus on many other high margin verticals, instead of putting many billions of resources into infra which could be obsolete tomorrow.
Agree, worth analysing what is genuinely commodity.
There are more elements to it though which can be sort of hard to explain.
There are whole cultures and ways of thinking built around production. The children of engineers who worked on xyz v1.0 have a genuine advantage when its time to work on xyz v2.0. There is a lot of tacit knowledge in these engineering fields and you have a huge advantage in knowledge retention if you can maintain unbroken chains of succession.
You can't achieve the top levels of ability (decades of experience, generational knowledge) if you are whip-sawing production to and fro across the globe every 10 years.
There are also cross pollination effects. Being in the same community with as many related fields as possible (co-located) is what drives cross-pollination and mobility of ideas and people between industries.
Think how many countries have tried to copy "silicon valley" and failed, and _why_ they failed.
What I'm saying is that technology is built by _people_ and there are human reasons why having local capacity is beneficial for all the related industries in the area.
> The children of engineers who worked on xyz v1.0 have a genuine advantage when its time to work on xyz v2.0.
my point is that other children with no extremely heavy investments into perl v1.0, will have some skills in c++ v1.0 and python v1.0, and will have advantage in adapting Tensorflow v1.0, which is more valuable than skills in perl v2.0. Heavily investing in one industry you sacrifice some flexibility.
So, this is multifactor analysis, lets say wise American people will elect me as next president, I would create list of industries, assign metrics (national security importance, potential revenue in 5y from now, impact on other industries, potential margin, risks of failure, etc), then build some formula which aggregate those metrics into single, and base on final metric allocate weighted funds to support N top industries.
First, they're not selling at a loss; the huge price increases have allowed them to push aggrssively in the legacy markets. They're making "slightly smaller" profits than other manufacturers (of which there are now very few).
Second, they can drive out all competition and then have a captive audience for whatever prices they want, as the barriers to entry in these markets are very high. This is essentially what's happened with all higher-end manufacturing in the west over the past 30+ years.
I don't know if it's still a thing, but China was getting a lot of heat about a decade ago for purposefully devaluing their currency to make their exports more attractive.
They kind of had to do this, because their large amount of exports were pushing the value of it up compared to others.
It's funny that you call this an "very aggressive dumping strategy" while AI vendors are doing the same but with even greater losses and on a much larger scale.
It's all simply a fight for market share.
The original sin is the existing DRAM vendors selling their entire (spare) capacity to the likes of OpenAI.
That explains the cheap DDR4 DIMMs on AliExpress. Can get 2x 16GB DDR4-3200 DIMMs for A$252 delivered to Australia. A local PC store has same spec “name brand” RAM for around $380-$400.
I was wondering when people would find out about CXMT. I wish them luck and hope the US doesn't sabotage them. We need diversity and competition right now.
DDR4 going from $1.35 to $11.50 in a year shows this market was already distorted before CXMT showed up.
Legacy DRAM is still over half of Samsung and SK hynix's production capacity. That's where the volume pain actually lands while they're betting everything on HBM4.
A near 10 times price rise of legacy chips and surely a lot higher profit increase couldn't motivate SKamsunix to increase their capacity for that segment, even more, they proudly informed the market about their decision to stop making DDR4 at the end of this year... to focus on even higher margin products.
But they aren't going to stop whining about China, no matter how much pain the market experiences.
This feels like a classic business blunder. Focus hard on a single business segment, leaving an opening in the market for your competitors. Not because it wasn't profitable, but because it wasn't profitable enough for you, right now. Only downside is that now you've created an opening for a new player in the market.
This feels like a short coming of western business/stock market thinking. Focusing on profit within the next few quarters, and not caring about the longer term consequences. For all it's flaws and shady business practises at least China can think beyond a single fiscal year.
Ok but this is how the market is supposed to work. If the incumbents aren't doing what their customers want, then competitors can rise and fill the gap and compete.
This isn't a shortcoming, it's a competitive market working as intended.
Who said I thought businesses would always make the right move?
Businesses blunder. "The market" is just a set of observations, including that competitors will tend to take advantage of those blunders. It is not a failure of the market that businesses have blundered, nor surprising that it will happen occasionally, and neither I nor mrweasel implied otherwise.
Lets not forget there is no competitors. There is one competitor- the chinese state, one huge company willing to subsidize any endeavor that will help it fmgain more marketshare with already captured markets.
NO you see, we have to hate Chinese companies because they are unfair competitors since they get state funding from the Chinese government, unlike Intel, Micron, TSMC, ASML, Samsung who don't get state funding from the US, EU, Taiwan, ROK ... oh wait.
Scratch that, we have to hate Chinese companies because they do business with the Chinese military, unlike Intel, Nvidia, Samsung who don't do business with the US and ROK military ... oh wait.
I know you are being sarcastic but the reason why we have to hate Chinese is simply because the standard of living of Americans depends on China not succeeding, simple as that.
Does it, though? If anything it seems like the opposite: China's success had directly enabled my standard of living as an American to be as high as it is.
I suppose it'd be true that the standard of living of some Americans depends on China not succeeding — specifically, those Americans who own corporations competing with Chinese firms — but I think they'll survive just fine with only 10 yachts instead of 15.
Well then hey can just say THAT, instead of coming up with hypocritical BS that doesn't pass the smell test. People internationally have enough IQ to see through the double standards BS, especially since youtube is a thing.
And the standard of living of working class Americans has been on a steady decline since Reagan by the hand of US administrations, not by the hand of CHina.
This isn't defending anyone's standard of living, it's defending profits of domestic monopolies like Micron, who indulge in state subsidies from US taxpayers and then fuck then over on prices.
CXMT sells the vast majority of their bits at the prevailing market rate, just like everyone else. They are adding capacity as quickly as they can, with a 5-10 year planning horizon, just like everyone else. It’s really not that deep!
PRC asked them to curtail DDR4 production so they didn't bottom out the market a year or two ago, and to focus on latest gen development, like HBM. They were the world leader in cost efficient DDR4 production at the time.
Yes of course their messaging to customers and the investment community is that they will be rational and measured in their investments. In reality, they are adding capacity as quickly as possible as margins are too high. However, capacity addition leading edge semiconductor manufacturing has a multi-year lead time.
> In reality, they are adding capacity as quickly as possible [...]
Even if we ignore the fact that you can't build out factories in secret, this would be securities fraud by publicly listed companies.
> ...as margins are too high
This is not the first RAM boom-bust cycle, and memory makers are an actual cartel convicted of coordinating in the past - none if them are going to break rank when they can invest the minimum and reap outsized benefits. Also, no one wants to invest in additional capacity when the bottom can fall out at anytime, and shareholders demand your head- not even the AI companies want to bear that risk, which is saying something.
It's not really a blunder though. Given that total capacity is tightly constrained, Samsung and SK Hynix are happy to focus on what they do at their best and with the highest margins. Why shouldn't they supply the HBM market?
It's putting all eggs in one basket. If/when the higher-margin category collapses, they'll have no fallback. Imagine Chevrolet had discontinued Impalas and other low-margin cars, and switched to Corvettes during the pandemic
Even if there's a collapse in the high-margin market (I don't think anyone is expecting this right now), it will be slow and telegraphed in advance, giving them plenty of time to refocus on lower-margin products.
There is really nothing about the stock market that means only thinking about the mext few quarters. See all the losses on the profit and loss statements of AI tech giants, or, say, game console companies? Why are their stocks still valued so highly during these periods? The answer: investors are thinking long term.
It is really impossible to have quality long term thinking without capitalization accounting and similar instruments that come out of the "wester" system of business that chinese free enterprise gladly and speedily copied when it was made free.
I think the sentiment here is about management's tie of bonuses to near-term stock performance. Maybe not about the market itself, I agree with your view on investors want long term gains over short term fluctuations mostly.
Is there a reason GPU's don't use insane "blocks" of sdcard slots (for massively parallel io) so the model weights don't need to pass through a limited PCI bus?
Yes. Let's do the math. The fastest sd cards can read at around 300 MB/s (https://havecamerawilltravel.com/fastest-sd-cards/). Modern GPUs use 16 lanes of PCIe gen 5, which is 16x32Gb/s = 512Gb/s = 64 GB/s. Meaning you'd need over 200 of the fastest SD cards. So what you're asking is: is there a reason GPUs don't use 200 SD cards? And I can't think of any way that would work
SD is obviously the wrong interface for this but "High Bandwidth Flash" (stacked flash akin to HBM) is in development for exactly this kind of problem. AMD actually made a GPU with onboard flash maybe a decade ago but I think it was a bit early. Today I would love to have a pool of 50GB/s storage attached to the GPU.
It might be very effective marketing. The big non-Chinese OEMs trust and use Korean and Japanese DRAM, and they might have been unwilling to put DRAM from CXMT into their products. (CXMT is newish, does not have access to ASML gear, which ASML would like you to believe makes it harder to make high-quality DRAM, DRAM is historically not a very large fraction of the cost of most non-huge-memory machines, and a bad DIMM is an expensive mistake for a company like Dell or HPE that is on the hook for repairs.)
But now CXMT seems to have gotten at least Dell, HP (I wonder if the article meant HPE), Acer and Asus to buy and attempt to qualify samples. If CXMT lands some serious purchasing agreements while still selling well above cost, that’s a win for them.
Does the last part of your comment explain it? They need revenue to expand capacity and the market has opened up a window to become a bigger supplier while still being profitable.
Because market rate is a 400%+ markup right now and not everyone is a greedy American kleptocrat with a diagnosable addiction to extracting every possible cent of wealth on the planet within a single fiscal quarter.
Has DDR5 caught up to DDR4 latency yet? I remember it was worse at least in the beginning. There's more bandwidth per channel but a hw design can always add more channels for the desired BW. Not so for latency.
and unfortunately increase latency even more with registered DIMMs. Comparing bandwidth increase (50 GB/s) to the stagnated latency (~80..120 ns total, less than ~0.1 GB/s) over last decades, I'm wondering, whether one still can call today's RAM random memory (though sure it can be accessed randomly). Similar to hard disk drives. Up to 300 MB/s sequentially but only up to less than 1 MB/s 4KB random (read).
As far as I understand, the "entity list" you are referring to is part of the "Export Administration Regulations", so it restricts sales from the US to restricted entities, not the other way around.
OTOH, now I read small Taiwanese manufacturers who are left out of the Nvidia supply chain are reverting to DDR4 motherboards because of the DDR5 shortage. Strange times.
It ain't the price raise that's the mistake (even if that's what's currently painful for those of us looking to buy RAM). It's the willingness to only raise prices, and not meaningfully expand production, that's the mistake.
But since when? There are public announcements about new energy deals since summer 2024. But I'm missing any information about similar RAM/NAND/HDD deals back then, so that corresponding shortages could be only for short time until, say, summer 2026.
The pertinent question, then, is who's gonna eat China's lunch?
My guess would be south/southeast Asia (India and Vietnam seem especially promising), but if the US was smart it'd put its efforts fully toward as many infrastructure investments and trade agreements and immigration agreements as possible to create a pan-American economic union. We have the resources and technology to turn every country in North and South America into an industrial and technological powerhouse. We have the resources and technology to finally conquer the Darién Gap and connect North and South America with highways and high-speed rail. We have the resources and technology to go on the offense against drug cartels (while also eliminating the failed border controls and drug prohibitions that keep those cartels in business in the first place).
If we're gonna be imperialists, then by golly let's at least be productive about it.
If that's the case, then why are the cheapest options I can find online multiple times that much?
> CXMT is in the process of converting wafer capacity equivalent to about 20 percent of its total DRAM output — some 60,000 wafers per month — at its Shanghai plant to the fourth-generation HBM3 chip production
It is wise for these Chinese fabs to eventually use a very aggressive dumping strategy to price well below cost push out other players forever, especially in DRAM.
But right now it seems they can max out their supply capacity without selling below cost.
Appears to me like China's endless state led (often unproductive) investment in semiconductor manufacturing subsidies (for decades) is about to pay off with some industry dominance soon.
Like the electric vehicle sector.
Crucial's departure from the consumer market left such a gaping hole, that CXMT doesn't even need to push other players out to gain a footing.
The downside in general is that other countries lose production capacity in steel, heavy industry, semiconductors, machine tools etc - industries that took decades to build and can't be easily replaced.
Also they gradually lose the ability to meaningfully innovate in those sectors because there's no grounding against production reality anymore.
This has geopolitical consequences further down the line.
That's not really what happens though. You don't actually "lose" capacity, you just move to higher-valued special niches within the overall industry because (1) you can afford to, while low-cost competitors can't and (2) you can no longer expect to be the lowest-cost supplier for the bulk of the market. That's a win-win development and something to be encouraged.
That's not what people mean by "lose" capacity.
Suppose DRAM companies expand capacity because prices are high, then demand levels off, the price crashes, and they all go out of business except for the one in China which gets a government bailout. That's fine, right? We're not interested in making DRAM, that's a fungible commodity, we want to make iPhones or something. (They make those too anymore, but never mind that.)
What happens now if China restricts what you can buy to give an advantage to their own companies who are trying to displace you in the higher-valued special niches? Or just raises the price for you and not them? What if there's a trade war? Or a conventional war?
When you still have a domestic industry, you go to them and have a source for the commodity. If only one country becomes the sole global supplier and that country isn't even particularly friendly, that's bad.
You can't just spin up a 2nm wafer fab when the latest you've been running is a 300nm process.
Compare: US shipbuilding industry to China or SK.
"easily" is doing a lot of work in that sentence. Depending on the good and what they switch to making, this may neither be easy nor quick.
Factories, tooling, supply chains, and engineering knowledge aren't fungible in the way they would need to be for your statement to be true.
As though moving production to China wasn’t something the West did intentionally.
And now continues to push manufacturing out of Western countries by, for example in the UK and Germany, and Australia too, making electricity and gas so expensive it becomes cost prohibitive to manufacture much at all.
the question is if single country can carry all these industries at loss for prolonged period of time.
Another approach is to rely on international supply chain and speed of innovation, we can't produce steel domestically profitably today, fine, we may buy it from diversified international supplier network, and rebuild it fast tomorrow if needed using new tech, and focus on many other high margin verticals, instead of putting many billions of resources into infra which could be obsolete tomorrow.
There are more elements to it though which can be sort of hard to explain.
There are whole cultures and ways of thinking built around production. The children of engineers who worked on xyz v1.0 have a genuine advantage when its time to work on xyz v2.0. There is a lot of tacit knowledge in these engineering fields and you have a huge advantage in knowledge retention if you can maintain unbroken chains of succession.
You can't achieve the top levels of ability (decades of experience, generational knowledge) if you are whip-sawing production to and fro across the globe every 10 years.
There are also cross pollination effects. Being in the same community with as many related fields as possible (co-located) is what drives cross-pollination and mobility of ideas and people between industries.
Think how many countries have tried to copy "silicon valley" and failed, and _why_ they failed.
What I'm saying is that technology is built by _people_ and there are human reasons why having local capacity is beneficial for all the related industries in the area.
my point is that other children with no extremely heavy investments into perl v1.0, will have some skills in c++ v1.0 and python v1.0, and will have advantage in adapting Tensorflow v1.0, which is more valuable than skills in perl v2.0. Heavily investing in one industry you sacrifice some flexibility.
So, this is multifactor analysis, lets say wise American people will elect me as next president, I would create list of industries, assign metrics (national security importance, potential revenue in 5y from now, impact on other industries, potential margin, risks of failure, etc), then build some formula which aggregate those metrics into single, and base on final metric allocate weighted funds to support N top industries.
Second, they can drive out all competition and then have a captive audience for whatever prices they want, as the barriers to entry in these markets are very high. This is essentially what's happened with all higher-end manufacturing in the west over the past 30+ years.
Do you think they are just stupid?
They kind of had to do this, because their large amount of exports were pushing the value of it up compared to others.
It's all simply a fight for market share.
The original sin is the existing DRAM vendors selling their entire (spare) capacity to the likes of OpenAI.
https://globalcio.com/news/16062/
You're maybe talking about the spot market, but companies are free to make any sort of supply contract.
https://a.aliexpress.com/_mssanU1
Legacy DRAM is still over half of Samsung and SK hynix's production capacity. That's where the volume pain actually lands while they're betting everything on HBM4.
But they aren't going to stop whining about China, no matter how much pain the market experiences.
This feels like a short coming of western business/stock market thinking. Focusing on profit within the next few quarters, and not caring about the longer term consequences. For all it's flaws and shady business practises at least China can think beyond a single fiscal year.
This isn't a shortcoming, it's a competitive market working as intended.
They'll blunder. They'll do it even harder in the absence of competition.
Businesses blunder. "The market" is just a set of observations, including that competitors will tend to take advantage of those blunders. It is not a failure of the market that businesses have blundered, nor surprising that it will happen occasionally, and neither I nor mrweasel implied otherwise.
Yes, identifying a problem is easy. But solving shortages in all cases requires perfect knowledge of future demand. So, good luck.
Scratch that, we have to hate Chinese companies because they do business with the Chinese military, unlike Intel, Nvidia, Samsung who don't do business with the US and ROK military ... oh wait.
I suppose it'd be true that the standard of living of some Americans depends on China not succeeding — specifically, those Americans who own corporations competing with Chinese firms — but I think they'll survive just fine with only 10 yachts instead of 15.
And the standard of living of working class Americans has been on a steady decline since Reagan by the hand of US administrations, not by the hand of CHina.
This isn't defending anyone's standard of living, it's defending profits of domestic monopolies like Micron, who indulge in state subsidies from US taxpayers and then fuck then over on prices.
https://www.tomshardware.com/pc-components/dram/memory-maker...
Even if we ignore the fact that you can't build out factories in secret, this would be securities fraud by publicly listed companies.
> ...as margins are too high
This is not the first RAM boom-bust cycle, and memory makers are an actual cartel convicted of coordinating in the past - none if them are going to break rank when they can invest the minimum and reap outsized benefits. Also, no one wants to invest in additional capacity when the bottom can fall out at anytime, and shareholders demand your head- not even the AI companies want to bear that risk, which is saying something.
It is really impossible to have quality long term thinking without capitalization accounting and similar instruments that come out of the "wester" system of business that chinese free enterprise gladly and speedily copied when it was made free.
But now CXMT seems to have gotten at least Dell, HP (I wonder if the article meant HPE), Acer and Asus to buy and attempt to qualify samples. If CXMT lands some serious purchasing agreements while still selling well above cost, that’s a win for them.
Because market rate is a 400%+ markup right now and not everyone is a greedy American kleptocrat with a diagnosable addiction to extracting every possible cent of wealth on the planet within a single fiscal quarter.
and unfortunately increase latency even more with registered DIMMs. Comparing bandwidth increase (50 GB/s) to the stagnated latency (~80..120 ns total, less than ~0.1 GB/s) over last decades, I'm wondering, whether one still can call today's RAM random memory (though sure it can be accessed randomly). Similar to hard disk drives. Up to 300 MB/s sequentially but only up to less than 1 MB/s 4KB random (read).
Search aliexpress for X99 dual socket motherboards.
The big Taiwanese manufacturers are chasing the AI dragon.
https://www.digitimes.com/news/a20251021PD219/ai-server-asro... (Oct 2025)
OTOH, now I read small Taiwanese manufacturers who are left out of the Nvidia supply chain are reverting to DDR4 motherboards because of the DDR5 shortage. Strange times.
Once established, the Chinese vendors will retain most the market share if the quality is ok. The SK/JP vendors are making a big mistake.
They will compete on price if they are forced to, but they aren't forced to right now
And if so, how?
USA got dominant, got arrogant, letting China eat their lunch.
China is indeed getting dominant. They will get arrogant one day. Meanwhile, Western Europe and the USA are still very good places to live.
My guess would be south/southeast Asia (India and Vietnam seem especially promising), but if the US was smart it'd put its efforts fully toward as many infrastructure investments and trade agreements and immigration agreements as possible to create a pan-American economic union. We have the resources and technology to turn every country in North and South America into an industrial and technological powerhouse. We have the resources and technology to finally conquer the Darién Gap and connect North and South America with highways and high-speed rail. We have the resources and technology to go on the offense against drug cartels (while also eliminating the failed border controls and drug prohibitions that keep those cartels in business in the first place).
If we're gonna be imperialists, then by golly let's at least be productive about it.