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.
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!
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.
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).
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.
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.
Even taking bank loans at a 10% rate to pay (maybe even 15%) upfront would still make sense.
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.
I'm concerned that the regular Joe/Jane won't be getting a GPU or a system upgrade for some time. And then you have people in Africa and elsewhere who can't afford new phones.
There probably is some distortion in that the top spenders aren't terribly against cornering the RAM market and forcing everyone to pay more for local AI due to higher RAM, GPU and CPU costs while squeezing people through their APIs to access AI services, but that's an expensive gambit that depends on local models not getting good enough and frontier models to continue to get better while requiring large swaths of memory and compute, keeping consumers demanding the latest models from the APIs. But if the supply increases and people have more access to RAM, GPUs and CPUs, we'll likely see more local model usage and people doing all sorts of creative work, even training their own models.
I'm bullish, not so much of an overzealousness as having witnessed so many of these computing explosions to not doubt it much anymore. We'll collectively find all sorts of creative uses for these wafers and tech. If I were to bet on it, I'd simply buy a basket of stocks and hold, riding out the noise - nothing too surprising there.
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.
Who says this? The latter are heavily regulated and not exposed to market forces, which is why they didn’t get cheaper, unlike TVs etc.
Also, from the article:
> In other words: you can buy a computer thousands of times more powerful than the best consumer device from 40 years ago, for something like 0.3 percent of the price. No other good in history has experienced a decline in cost on that scale: poor people can now carry around in their pockets computers many orders of magnitude more powerful than what the richest slice of the world’s population could afford a few decades ago.
That extreme level of efficiency increase is something very specific to computers and not something translatable to other areas of the economy.
Cato https://www.cato.org/publications/trade-buys-goods-services-...
Cato “One of the big reasons Americans’ inflation-adjusted wages have climbed in recent decades is that the exorbitant prices of things such as housing, health care, and education have been offset by significant declines for tradable goods such as toys, clothing, and consumer electronics.” https://www.cato.org/commentary/cheap-talk-cheap-stuff
The Adam Smith types: https://www.adamsmith.org/blog/economics/economic-nonsense-2... "Many consumer goods follow the same trajectory; most recently smartphones have done so. Far from being left behind, the poor are pulled along by the progress that initially caters for the rich."
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.
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.
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.
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_.
Here it's useful to point out that free markets can't exist without an external force keeping them in check.
256K bytes of RAM that way made us $128 in sales. I earned $3.35 an hour. Any inventory I accumulated went out the door next morning.
It was a brief moment, but it makes me wonder if we'll see refurbished memory in fashion next year.
BTW the company, OEM Parts, was one of the last great surplus shops, surviving until just recently. Probably there were still cardboard bins with my handwriting for all the TTL chips that weren't in demand.
We are not launching our product this year. We can scrape together enough memory for shareholder and show floor demo units, but we have tens of thousands of people on the waiting list and we simply can't get that much memory at any price. What we can get would mean a $400 price bump.
What the hell do you even do?
Apple’s not exactly famous for their low pricing on spec upgrades nor competing based on being the price leader…
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.
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.
But maybe that hardware becomes so commoditized that it's not difficult to obtain / stuff in a box.
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.
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.
They make a strategic investment in the DRAM maker. It's the same idea as buying an equity stake or entering a JV with a flat-panel maker. Or deploying cash to provide the latest EUV tooling to TSMC.
Large famous-brand computer maker has no expertise in running a fab.
I could punch holes in this idea right away. But it seems better than "tilting up your own fab".
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.
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.
They can't buy enough today. I think I already explained this as well as I can in my second paragraph above. You seem to not be appreciating that every decision in this business has a 3.5+ year latency. And 3.5 years ago when Apple made the decisions they're living with now, Micron was selling their furniture (euphemistically speaking). Sinking billions into building RAM fabs would not only have gotten you laughed out of the boardroom, Apple shareholders would have rioted (for all the logical reasons I laid out in my prior two posts). The only scenario where the prudently balanced decision Apple made to not vertically integrate RAM manufacturing looks bad is the very unlikely scenario that actually happened this one time - but that 'bad look' is only temporary because the situation will probably look different 12 months from now (which is less than a third-of-a-fab-decision away)).
Back when I first became a strategist for a public F500 tech company whose products you probably use often, a very wise man took me to dinner and counseled me to never post mortem my decisions on the outcomes, but only on the decision process. In other words, knowing what I knew when I made the decision, did I choose the option most statistically likely to get the desired result - regardless if it worked or failed that time? In highly uncertain games, some bets don't pay off. In my work, if 60% of my decisions came out positively, I'd have been the greatest of all time (they weren't that quite that high). If strategists don't think this way we end up leading our companies to doom by "chasing black swans" (aka unlikely outlier outcomes).
> memory manufacturers leave demand unmet
That's not their plan or desire - it's just the inevitable result of trying to predict both future demand and future capacities of unbuilt fabs on new nodes (every new node is an adventure). Both are rough estimates with bell-curve probabilities and wide error bars. You might imagine the Chief RAM Strategist drags the center of the bell curve so the big bump in the middle is right over their best guess of what the fab yields and market demand will actually be three years from now - perfectly slicing the odds in the probabilistic middle. But they don't. Instead, being a smart, experienced strategist, they slide the bell curve a little bit to the low-side - because very profitably selling everything you can possibly make and regretting the 5 or 10% of demand you left unmet is a far, FAR better (and more survivable) problem to have than sitting on 5 or 10% unsold excess capacity from expensive fab space you paid to build but can't monetize. That extra capacity will often force you to take lower margins on ALL your capacity to ensure you sell through the extra 10%. This is how you speed run becoming an ex-strategist, sometimes of an ex-company.
In short, being wrong on the low side is the best kind of wrong you can be. But being wrong on the high side is the worst kind of wrong, and it's not linear. Every extra percent you go over costs significantly more than the last. If they do their job perfectly, it looks to consumers like "manufacturers leave demand unmet." If they do their job less than perfectly but pretty well, it looks like "RAM's too damn expensive and the idiots created a shortage - what were those morons thinking?" If they do their job catastrophically it looks to consumers like "Woohoo, RAMs dirt cheap. Look at me! I've got so many sticks I'm bathing in them like Scrooge McDuck!"
What I did say is that in a hypothetical world where Apple did invest in their own fabs, they'd have much better access to memory and could be selling machines with vastly more memory than their competitors today. Should they be doing this? I'm making no statement either way.
> a very wise man took me to dinner and counseled me to never post mortem my decisions on the outcomes, but only on the decision process
I appreciate this wisdom on one hand, but on the other it's comically misguided. You have no way to evaluate the decision process if the outcomes are not considered. Don't post mortem your decisions in a way that leads to self-blame or paralysis, but certainly post mortem them to see whether the process was suboptimal.
In this case, any board laughing at the idea of investment in memory 3.5 years ago should instead have been looking more closely at likely memory needs over the then-near-future, given the quite obvious upcoming demand for GPU memory. There's no magic bullet and they wouldn't have been able to make perfect decisions regardless, but clearly the process they followed at that time was suboptimal.
Okay, I understand your point. I've always taken it as given that vertically integrating DRAM fabs was one of many options available to any very large manufacturer with sufficient cash flow. However, I think it's inaccurate. If Apple invested in their own fabs, they would not have better access to memory today, unless they ALSO did something quite irrational and overbuilt their fab's capacity by a factor of 2X or more. In terms of greater capacity, at Apple's massive scale of consuming more than a "one fab" amount of output, owning their own fabs vs ordering from fabs owned by others still requires pre-ordering their expected RAM needs ~3.5 years in advance. Regardless who owns the fab, no fab can make significantly more chips without that long lead time. The difference in lead time, total capacity and margin between a self-owned fab you started building 3.5 years ago and "pre-buying most of a fab's output 3.5 years in advance" is minimal enough to be virtually immaterial.
> Should they be doing this? I'm making no statement either way.
Given the dynamics, I assumed you weren't just stating that acting irrationally is always an option, but that doing so was a rational option strategically. But it's not. Even in your scenario it wouldn't have changed anything UNLESS you imagine Apple of 3.5 yrs ago was psychic and built two fabs for every one they actually expected to need - but that would have been insane. And if they were that insane, they could have accomplished very nearly the same thing by pre-buying 2X their expected RAM needs from their 3rd party suppliers. At Apple's scale, 3rd party suppliers are what we call 'virtually captive'. You're such a large customer, you're practically funding the supplier's capex. This is a serious strategic problem for such suppliers (but great for Apple).
> I appreciate this wisdom on one hand, but on the other it's comically misguided. You have no way to evaluate the decision process if the outcomes are not considered.
Yes, of course you consider the outcome as one data point. But I said you don't assess by the outcome, meaning we don't judge the decision process's ultimate validity only (or even mostly) by the outcome. It's just one element. The reason is that for highly uncertain decision spaces with unknowable key externals, you can have NOT just: 1. Optimal decision-making / successful outcome, and 2. Sub-optimal decision-making / failed outcome but ALSO: 3. Optimal decision-making / failed outcome, and 4. Sub-optimal decision-making / successful outcome. Understanding the ramifications of this matrix is crucial - especially the last two.
A key example, is the hypothetical where a crazy strategist consults an astrologer, pre-commits to 2X the projected need for fab output (whether by building/owning fabs directly or just indirectly by pre-buying one or more fab's entire output in advance). If the astrologer says "build 2X what you think you'll need" and there's an unforeseeable confluence of events 3 years later that causes a worldwide shortage of RAM, then if you overweight the outcome in assessing if your strategic planning process is optimal, you'd always just do what the astrologer says. In short, in high-risk domains, sometimes a sub-optimal decision-making process can randomly result in a successful outcome - but that's a spurious signal.
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.
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.
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!"
And when I upgraded from tapes to the 1541 floppy drive (5.25" floppies), it was like "thats it - I have INFINITE possibilities now"
Simpler times :)
Ah, the legendary 1541. We used to joke it was the only floppy drive on any 8-bit micro that was actually slower than some dial-up modems. It wasn't actually slower, but it was really slow compared to most other 8-bit computer's drives (unless you added after-market accelerators).
Who knew :)
That's not how I remember it. By the time we got data cassettes, we knew very well that floppies were the future.
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.
While Apple et al certainly have the money to tilt up their own fab, they're savvy enough to understand the memory market's long history of constant boom/bust cycles. 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.
People assume Apple cares about vertically integrating cost but they're actually focused on integrating margin. Apple has billions in cash on hand and when they think about what to do with it, a key metric is Return on Capital, especially the margin that capital will generate. Since a core metric public companies are judged on is blended margin, they are looking for ways their bags o' cash can be put to work generating revenue at margins that will pull their current average margin up vs down.
Averaged over time, mainstream memory devices are historically one of the worst margin areas of the semi market. It's super expensive to tilt up a fab on a new node but once you do, turning the crank faster to make a lot more chips isn't too hard because mainstream DRAM tends to be quite uniform. So when a fab on a new node and/or RAM generation first opens, the margins tend to be pretty great. But as the node matures and/or the RAM generation goes from 'new' to 'commodity', competition heats up as everyone gets better at making more faster. Then they're tempted to maximize revenue by cutting prices until their mature fab is at 101% utilization. And that eventually drives margins down until someone's selling near cost to sustain their low-price-enabling volume - with occasional dips below cost when they get stuck holding excess inventory. That's why cash-rich companies with high margins like Apple are delighted to buy DRAM built with Other People's Money. As long as the DRAM market is under competitive pressure, Apple gets to shop their huge orders around to get the absolute lowest price on RAM that was built with other investor's low margin dollars.