upvote
> I still remember the huge DRAM shortage in late 80s forcing my startup at the time to delay launching our new product for a year.

That one was caused by manipulation by politicians, not market forces. Micron started a price war with Japanese memory manufacturers, the Japanese cut prices to compete, Micron sued them for "dumping". The saga ended with the 1986 U.S.–Japan Semiconductor Agreement, which, among other things, created production controls that limited the total dram supply. The level was set based on then current demand, and due to the rapid growth of demand at the time it almost instantly caused a massive global supply deficit.

The agreement also caused the rise of the South Korean memory industry, because the Japanese companies offloaded their now surplus equipment for cheap.

reply
> That one was caused by manipulation by politicians, not market forces.

Funny how you then go on to explain it was market _actors_ that drove the politicians to act, but that this is nevertheless totally not an example of market _forces_.

reply
Market forces are usually meant in the spherical cow sense, not that the cow has politicians on their pockets.

Here it's useful to point out that free markets can't exist without an external force keeping them in check.

reply
This has become like opec's best years. Only there's no cartel at least not openly.

But there's another key bottleneck. Even with all the money in the world, getting those machines that etch the RAM could be a multi year ration shop queue. And they're not making those companies every day!

reply
> This has become like opec's best years

Yes, except no DRAM maker is taking this as a signal to go deep into debt to double or triple capacity. It takes a minimum of 3+ years to dramatically increase capacity and no one expects this inflated demand bubble to last that long. Because they've seen this cycle before, they'll bump next gen capacity up a bit more than planned, maybe 15-25% because the current windfall profits are enough to build a buffer to absorb the hit if that excess capacity comes online in the next DRAM demand crash.

In the meantime, they'll just apologize to everyone for the "market conditions beyond our control" while banking as much of these crazy profits as they possibly can while it lasts. But deep down they remember they were starving just yesterday and know they'll be starving again tomorrow.

I understand why everyone's pissed at the "evil DRAM makers" but I also remember boasting about how I scored some crazy cheap RAM sticks not so long ago. None of us were shedding tears when they were selling at a loss just to survive.

reply
> Yes, except no DRAM maker is taking this as a signal to go deep...

Incorrect. CXMT is taking this as a signal to open new fabs, create new products, develop new silicon manufacturing techniques, and enter the consumer market worldwide (except USA).

reply
> Incorrect.

Okay, you got me. I was reflecting the prudent way long-term players generally interpret the market. But from time to time external macro-economic and/or geopolitical factors may shift. Those changes may persist over time or eventually swing back like tides. Other times someone thinks they've got a new angle or a new technology (the RDRAM format was a notable attempt). Occasionally, the people who always say "But this time it's different" are lucky enough to actually say it one of the rare times it IS different.

Time will tell if CXMT's gambit will pay off. Even if it doesn't maybe their state sponsors will absorb the losses for other strategic reasons.

reply
CXMT is a company with heavy state backing and control that started its current trajectory years and years ago. It has infinity money and isn’t allowed to go bankrupt. CXMT is reading political signals (Beijing wants indigenous AI up and down the vertical) as much as market.
reply
> CXMT is a company with heavy state backing and control that started its current trajectory years and years ago. It has infinity money and isn’t allowed to go bankrupt.

This also describes the U.S.A. steel industry. "Heavy state backing" and "not allowed to go bankrupt" often result in particularly terribly-performing industries. If anything, it's notable that China seems to buck that particular trend.

reply
Not sure "only the Chinese model can solve the normies pain" is a great look.

Lately consumers are told capitalism is cheap TVs, phones, and computers while housing, healthcare, education, energy, and food climbed further out of reach. The "bread" part of bread and circuses has shrunk dramatically, "but at least you're getting cheaper circuses" is what they've been saying.

Now you're saying only a heavily state-directed Chinese model is still willing to aggressively finance new fabs and meet demand to provide that? "Capitalism demands we don't expand capacity and don't meet demand' is kinda a tough sell after capitalism has been sold as expanding access. "Consolidation, maximal extraction, shareholder preservation, and AI" isn't going to be a winner. I get it makes business sense and wouldn't be a big deal in the past, but the rapidly changing dynamics hitting nearly every part of daily life across the entire economy feels destabilizing.

We're down to wearing dystopian sci-fi level cheap clothes, can't afford chips/sodas, candy/sweet treats have been enshitified to no longer be fulfilling, we can't afford to call an ambulance. Media is polarizing instead of mass media calming. Heck thrift stores are becoming unaffordable to the low end as the middle now resorts to them.

"You don't have the foods you are used to, decent clothing, shelter, health, or social cohesion but you have a sweet 6 year old phone and flock cameras keeping you safe". Jesus it feels... not good. Even cyberpunk dystopia understood you needed to at least jack people into something.

reply
Isn’t that the point?

Everyone knows the commodity market (outside HBM) is going to be margin crushed within a few years… by CXMT, demand swings, and so on.

So nobody cares to fully meet the demand today. They are perfectly happy telling customers (those not willing to sign a 10 year contract paying up front at least) to go pound sand.

They don’t even need an OPEC like arrangement. They are effectively perfectly coordinated already in dismissing customers.

reply
Yes, there are a bunch of bizarre comments in here. "Why aren't they rushing to make more fabs? Don't they know China is using this moment to the enter the market in a big way?" Yes, they do! And building more fabs would make that worse. The last thing you want to do before a state-backed player enters the market in a boom cycle is load up on debt to produce more capacity just before the bust.
reply
It's a bit surprising many here are referencing the memory supply cycles without mentioning this revolutionary new application for memory people call AI. It'd be like talking about weather cycles and climate without mentioning global warming. Just like the planet is not going to get colder on average than it was for the foreseeable future, we're not going to need less memory for the foreseeable future.

China is trying to be vertically integrated, completely independent of outside influences and own the future supply/means of production. Memory and chips are piece of their larger plan. In any event, whatever China brings online will be absorbed - imagine a future where people's home computer has 100 to 1000GB of RAM in one variety or another. Folks are going to want better chips and more memory for years to come, supply will be absorbed.

reply
> this revolutionary new application for memory people call AI.

Most analysts think LLMs will elevate the long-term base RAM demand level. I mentioned 15-25% above prior projections without AI, which I think everyone agrees is highly likely. That's actually a lot because it's an overall market number and RAM goes lots of places other than PCs, servers and high-end mobile (depending on how you segment, 25% overall could be in the neighborhood of doubling PC, server, high-end mobile demand).

Above that range analyst estimates diverge. Some are more bullish, and a few are much more bullish. But everyone's error bars get much wider when the numbers go over 30% overall. It's hard to tease out exactly how much of the current demand bubble will persist in the long-run. Clearly, the current market is distorted by short-term dynamics but which part is base demand and which distortion?

How much consumer AI compute will be on-device vs aggregated in load-balanced clouds? How much RAM will that kind of compute require? Will the market find it's more efficient to consolidate around two or three mega-datacenters or will each frontier lab (and geopolitical block) continue drag racing each other to tie-up future RAM (as much to keep it away from competitors as for their own needs). I don't know. I've been watching this game as an interested bystander for several decades and I wouldn't bet too much of my own money on the most bullish estimates.

reply
If this future is so certain… where are the customers (non HBM) willing to sign a 10 year contract paying up front?

Even taking bank loans at a 10% rate to pay (maybe even 15%) upfront would still make sense.

reply
To reframe your great comment:

Is it fair to conclude that DRAM is basically a commodity that can be specified well enough by a set of parameters?

If so it won’t allow you to get any competitive advantage in your products and thus wouldn’t be a business you want to be in as Apple.

reply
Yes, that's what I meant by "Mainstream DRAM". The basic parameters are specified by standards like JEDEC, so the most common types can be used in many systems thus reach high volumes. This can be less true for specialized RAM like HBM or the highest performance grades.
reply
The competitive advantage is having access to it.
reply
The only reason I can see Apple do this is if it enables them to sell entry level devices with vastly more ram than the competition can afford. Say entry level MacBook Pro with 256GiB ram to facilitate running frontier level local models. If that is an edge they want to have.
reply
In what world would 256GB of RAM be “entry level”?
reply
In the same world 8GB used to be unfathomably huge not that long ago?

Hell, an 8GB hard drive was unfathomable when I was a kid in the 90s. I remember getting a 30 megabyte drive for our Mac LC.

reply
The first computer I ever used had 512K, which was a great deal back then, and by the time I was old enough to learn to type it had 10MB disk too.

My childhood best friend and neighbour had the same kind of computer except they only had something like 384K of memory and I tried to convince them their computer was broken when it didn't count up all the way.

reply
> The first computer I ever used had 512K

Mine had 4K of RAM and an 800 KHz CPU - and I was living in the future, man. No way I could use that much memory. After all I had to type in whatever program I wanted to run every time I turned it on. Then I got a manual audio cassette recorder and thought "Woah, I don't think it gets better than this!"

reply
And that Mac LC only had a max of 10MB of ram.
reply
The same world where a MacBook Pro G4 12" came base level with 256MB of RAM.

The same world where a classic Mac came with 128KB of RAM.

We might someday live in a world where entry-level is 256TB of RAM.

reply
That is the timeline I want to live in. Probably.
reply
A world where Apple has invested in their own fabs, so they can sell devices with drastically more RAM than their competition at entry-level prices.
reply
The point of my GP post was that due to being one of the world's biggest and longest-term buyers, Apple is already paying very close to actual manufacturing costs + amortized capex because RAM is an undifferentiated commodity. Owning the factory themselves doesn't reduce the actual manufacturing cost + amortized capex that Apple would have to pay their own factory. Apple is already buying RAM at the lowest possible margins. It's similar math to deciding whether to spend your own cash or get a loan. If the loan's interest rate is low enough, it's better take the loan and put your cash to work where it can return a higher margin. And at the incredibly low margins Apple pays for RAM, keeping that cash in long-term investments will actually earn more money than putting it into building RAM factories.

If Apple could go back in time 3.5 years and decide to build their own factory, that would put them in a great position today. But deciding to do it now won't increase their supply 3.5 years from now more than just increasing their long-term orders with existing suppliers. Those suppliers will start building new factories based on Apple's increased orders and they'll do it faster and cheaper than Apple can because they don't have to build some factories in the U.S. for political reasons or worry as much about environmental regulation, permitting and ensuring Apple employees in Penang get benefits similar to employees in Cupertino.

reply
Isn't one of the points of the article that memory manufacturers leave demand unmet for their own financial safety? In which case, nobody (including Apple) is paying close to manufacturing costs. There isn't enough memory to go around and prices are extremely inflated.

You're talking about the "best" things Apple could do with their money, in terms of investment returns, but I think that misses the point that Apple literally can't buy enough memory at any price.

reply
I'm no business expert and Apple is of course in a unique position, but owning your own fabs has rarely worked out long term. They require eye watering amounts of CAPEX that needs to be amortized over a timeframe that's longer than apple's products. Today's bleeding edge fabs become tomorrow's "cheap" fabs that pump out chips that don't need to be bleeding edge for the components that go into everyday products like microwaves, cars, etc.

One of the reasons Intel fell behind is that they couldn't give access to their competitors for business reasons, and therefore could never scale as high as TSMC could.

There are many other reasons, but accounting is a huge one. Unless there is a huge ROI or something else we don't otherwise know, I don't see Apple adding such expensive deprecating assets onto their books as chip fabs.

reply
Why would they sell a device with 256GB of RAM as the lowest-spec device rather than making 8 32GB or 16 16GB machines as their entry-level?

Apple’s not exactly famous for their low pricing on spec upgrades nor competing based on being the price leader…

reply
If 256GB of RAM enables them to run on-device AI models that (for reasons) are a key feature differentiator?

Personally, I think there's no way memory heavy inference moves on-device (vs cloud) due to the economics, but it's not impossible technology + platforms go that way for currently unforeseeable reasons.

reply
I think there’s a realistic chance consumer inference moves on-device. I think it really depends on marketing.

My non-tech friends and family would probably be served perfectly fine by local models today, if they had a working web search tool. Their queries are often “soft” and don’t have an exact answer. My mom and aunt used it to pick a hairstyle, my mom used it to get an image of what a room would look like with particular drapes in it, etc. Stuff I think mid-sized local models like Gemma or smaller Qwens could do without issue. They just don’t have a device that will run them.

Businesses won’t move. They need a huge context so they can stuff a bunch of Confluence pages in it and 300 tools and it needs to read an entire codebase and yada yada. The hardware depreciation and electricity will probably make it a net zero or even cost more than paying for API access.

reply
The economic argument in favor of cloud inference: higher utilization is always going to have a ROI for inference hardware.

But maybe that hardware becomes so commoditized that it's not difficult to obtain / stuff in a box.

reply
My argument is predicated on the assumption that mainstream hardware manufacturers will copy the way Apple and Framework have made system memory usable for inference.

In that world, a) we are already at or close to having enough memory in local devices to do inference locally, and b) that memory isn't inference-specific and can be utilized for other things. Most devices come with enough memory to do some level of inference, and some come with plenty (eg a gaming desktop probably has 32GB+ of RAM in it).

You aren't going to run Kimi on it, but I think the reality for a lot of consumer inference is that it doesn't need to be. It's going to be a lot of things that are soft, and easily answered by a search API, so the LLM really just needs to be able to skim and summarize. Going a step further, we may even see some kind of hybrid approach where a local OpenRouter kind of thing decides whether the task is soft enough to do locally with models that fit in RAM or if it needs to be farmed out to a PaaS provider.

reply
Right. I’m not arguing that Apple wouldn’t offer a 256GB model if they could make money doing it; I’m puzzled as to why they wouldn’t offer several lower-spec models as the entry-level into and then progressive upgrades within that line, since only some people need that 256GB feature differentiator of running frontier-level models on their MacBook Pro.
reply
And I'm saying, if 256GB of memory is a requirement for running customer-expected local models (and local models are preferred for some reason).
reply
Think past on-device inference... imagine what on-device training could do. And that would need a lot of RAM.
reply
And yet, Apple had to drop base configs from their lineup. They weren't selling $599 Minis at cost. They could take someone over and inflict damage on competition.
reply