Micron Suggests Apple Helped Cause Memory Price Crisis

[High IQ Person]

I don't live in the US, so I could not care less where Apple sources their components from, and what it means for those respective countries. Ram is quite readily interchangeable, there isn't really anything special about the ram being produced by Micron, and If anything, the entire issue is precisely because the majority of the world's ram is controlled by a small cartel of 3 companies who were unwilling to invest in expanding capacity when they had the opportunity. It's not that they didn't have the money to, but that these companies did not want to be stuck with a glut of ram, which would in turn lead to a drop in prices and further impact their profitability.

I will also go on to argue that the current climate in the US is not conducive to scaling the production of ram either. Investment is expensive and a long-term endeavour (and long-term planning is something China seems to excel at in part because their leadership is so stable), and US lacks the skilled labour willing to work in such fields (people tend to outsource what they consider as manual labour overseas).

If the US wants to maintain their lead in this area, then the current status quo cannot continue. The current US economy is being propped up by the same few AI companies passing imaginary money back and forth, and its people would be better off worrying about what this means for them if / when this bubble does (eventually) pop, and less about other countries' rise to power.

I am not insane, just pragmatic.

Ram in consumer products is interchangeable as long as the memory was manufactured on the same/similar node.

However, the Ram used to run AI models and data centers are not as interchangeable, with a lot more focus placed on the Ram's heat generation, power consumption and yield.

A yield of sub-80% leads to significantly higher manufacturing costs when you stack them into an HBM. Data centers also generate massive amounts of heat and use a ton of electricity, so the dram's heat and energy profile makes a difference in whether a manufacturer's HBM gets qualified or not. The emphasis on yield, energy and heat profile will continue to gain more importance as newer HBM generations continue to stack more and more dram modules. China's RAM companies are approximately 3 years behind on 1c dram node and approximately 5 years behind on HBM. Giving them business so they have enough capital to expand and, more importantly, gain experience in dram manufacturing will let them catch up to us and eventually surpass us.

This isn't about Apple selling iToys with a slightly higher profit margin. It's about securing our lead in AI.

I have no idea what country you're from, but any country not in China's sphere of influence should not want Chinese memory manufacturers to win.

 

[Smartuser]

Ram is quite readily interchangeable, there isn't really anything special about the ram being produced by Micron

The moat of these companies is only the investment to build the facility.

and If anything, the entire issue is precisely because the majority of the world's ram is controlled by a small cartel of 3 companies who were unwilling to invest in expanding capacity when they had the opportunity. It's not that they didn't have the money to, but that these companies did not want to be stuck with a glut of ram, which would in turn lead to a drop in prices and further impact their profitability.

That reasoning is common, it's just weird to me that many companies and industries can manage this without any drama.

 

[High IQ Person]

The moat of these companies is only the investment to build the facility.

That reasoning is common, it's just weird to me that many companies and industries can manage this without any drama.

Completely wrong. The moat is the manufacturing process. The complexity of memory design and fabrication is what causes the cost to be so high.

There is no proof that throwing money at semiconductor fabrication is all you need to run a state-of-the-art semi fab. If that were the case, China would be the leader of semiconductors today. In fact, every country would be throwing money and starting a business making $300+B in profit annually if it were just a matter of investment.

You're missing the human capital part of the equation. You can be the richest country in the world but if you don't have smart people, your investment will go down the toilet.

The work that Micron and the other memory manufacturers do is far more difficult intellectually than anything Apple does.

 

[Smartuser]

It makes sense.

Sure, no need to address how the cartel is treating non-Apple companies.

Apple is big enough to be a monopsonist

Do you have any numbers to back up your story?

We need very aggressive enforcement of antitrust laws.

But the AI explosion from the US government's inflationary ways with tech and executives of firms buying AI systems getting easily pilled with fads they know very little about and the stock market bubble is a bigger factor still.

I thought politics were off-limits here by the forum rules?

 

[High IQ Person]

Perhaps you should read their 2023 Form 10-K for fiscal year ending August 31 which is available on SEC EDGAR.

Speculation as to how much Apple could be to blame if guilty. It is not known how much of Micron's $15.5 billion*** 2023 revenue was from Apple. Neither company provides relevant data, but Apple is believed to be Micron's largest customer with high end estimates of 15%. We do know that 10 companies represent 50%.****

That makes a high end estimate of $2.4 bliion Apple paid Micron. What sort of premium do you think Apple should have paid? 10%-$240,000 million? 20%-$480,000 million?

Accounting 101: cash flow and profit/loss are two very different things. For example, expenses that had zero impact on cash flow included over $7.7 billion depreciation and amortization.* And guess what...Micron also paid over $1 Billion in stock-based compensation to its executives charged as an expense for profit/loss but also have zero impact on cash.

Micron ended the year with ~$12 BILLION cash, short term investments, and receivables (with an additional ~$8 billion of unsold inventory net of markdown to market below cost the cost to manufacture it).** Had they not paid a dividend of $504 million and not bought back $425 million in their own stock, they could have had ~$13 BILLION available to start (just start, not total cost to complete) another FAB.

Had they believed in themselves rather than blaming their customer, they would be 3 years into the construction with plenty of cash to finish it.

*Consolidated Statements of Cash Flows, p61
**Consoldated Balance Sheets, p59
*** Consolidated Statements of Operations, p57
**** A significant portion of our revenue is concentrated with a select number of customers, p29

Yeah, Micron should risk total bankruptcy when the memory market in 2023 was trash and they were posting $6B in losses. That $13B was needed to keep the business afloat for the next few years if the market was still in shambles.

Blaming Micron for not predicting the growth of AI data centers is an unreasonable take. Almost trollish. If it were that easy to predict where the market was heading, why didn't Apple lock up their memory supply with a LTA far below current rates?

I hope you are not a CFO or in any position that deals with a lot of money, because this is perhaps the most stupid thing I've heard this year. If I proposed what you just said to my execs, they would've fired me on the spot. It's so astoundingly dumb.

 

[Abazigal]

This isn't about Apple selling iToys with a slightly higher profit margin. It's about securing our lead in AI.

This reply sounds suspiciously like it was written by AI, but at any rate, I am not particularly a fan of AI at the moment, regardless of which country it is from. Sure, it has its useful moments for me, but I dislike how it is responsible for devaluing and deskilling so many people, the environmental impact, the blatant disregard for intellectual property rights, and honestly, I don't see how these companies will ever be profitable.

The Ai bubble could pop tomorrow, these AI companies could go bankrupt overnight, and I wouldn't shed a tear.

The moat of these companies is only the investment to build the facility.

Which is apparently an even greater moat than people give it credit for, given that nobody else is doing it.

You're missing the human capital part of the equation. You can be the richest country in the world but if you don't have smart people, your investment will go down the toilet.

I feel that specifically, it's not about having smart people, but having the right people willing to work in such conditions for what I imagine is not very fantastic pay.

For the longest time, the US has outsourced whatever they could to other countries. Everybody wants cheap stuff (many of my friends are increasing ordering their furniture from Taobao), but nobody would ever work in the conditions under which they were made. You expect me to believe that thousands of Americans can magically be convinced to work in ram factories when the time comes, and long after the demand and prices of ram falls back to normal?

People are losing their jobs (and finding it harder to secure employment), wealth disparity is greater than it has ever been, and now prices of electronics are increasing everywhere. You wonder just what the final straw that breaks the proverbial camel's back will be with regards to AI sentiment. 🤔

 

[High IQ Person]

I feel that specifically, it's not about having smart people, but having the right people willing to work in such conditions for what I imagine is not very fantastic pay.

For the longest time, the US has outsourced whatever they could to other countries. Everybody wants cheap stuff (many of my friends are increasing ordering their furniture from Taobao), but nobody would ever work in the conditions under which they were made. You expect me to believe that thousands of Americans can magically be convinced to work in ram factories when the time comes, and long after the demand and prices of ram falls back to normal?

People are losing their jobs (and finding it harder to secure employment), wealth disparity is greater than it has ever been, and now prices of electronics are increasing everywhere. You wonder just what the final straw that breaks the proverbial camel's back will be with regards to AI sentiment. 🤔

There is nothing more cognitively difficult that state-of-the-art semiconductor manufacturing.

Also, Hynix and Samsung semiconductor workers are making $400K-1M annual bonuses, so they're not underpaid.

It's an industry with high revenues and high margins. It has to do with human capital. A $20B investment is nothing to most governments, and every single one of them would gladly fork over $20B every few years if it meant they had an industry generating $300+B in annual profits. In fact, if projectioins for 2027 and 2028 are right, Hynix and Samsung are projected to make $600+B in combined profits. Is it really about pay? With an industry generating that much profit, any country could pay their workers millions but they don't because they don't have the human capital.

 

[theorist9]

The memory market was strongly over-capacity at the time Apple negotiated those contracts with Micron. It thus doesn't seem plausible that Micron would have broken ground on a new plant at that time, regardless of what pricing they negotiated with Apple.

The WSJ reporter who interviewed the Micron CEO should really have challenged his assertion on that basis. And maybe the reporter did. But, if so, that portion was not shown in the MR summary.

 

[Fallforward]

You first. What was their net income?

Accounting 101? Done. Accounting major and CPA exam as well. For a start, understand the difference between cash and accrual accounting.

Read their 2023 Form 10-K? Done and provided page numbers for you to understand why 2023 net income wasn't a limitation in their decision to not start more fabs - not even close. Also provided an educated guess of the little impact of a sweetheart deal with Apple, anyway. Micron simply didn't see this boom coming.

 

[High IQ Person]

Accounting 101? Done. Accounting major and CPA exam as well. For a start, understand the difference between cash and accrual accounting.

Read their 2023 Form 10-K? Done and provided page numbers for you to understand why 2023 net income wasn't a limitation in their decision to not start more fabs - not even close. Also provided an educated guess of the little impact of a sweetheart deal with Apple, anyway. Micron simply didn't see this boom coming.

An Accounting major and CPA show proof that you can read a financial statement, not that you make smart or reasonable investments. In fact, your plan is the dumbest thing I've read this year and would get you laughed out of the board room of any company. Even Micron's gross margins was negative (-10% IIRC). I don't think I need to tell you what that implies (Hint: There's no amortization and depreciation with gross profit). That was how dire their financials were in 2023. Or is your view that Micron should operate at an indefinite loss just so Apple gets cheap memory forever?

Stick to your debit/credit tables. Let the smarter people do the critical thinking and make the investments.

Accounting 101: cash flow and profit/loss are two very different things. For example, expenses that had zero impact on cash flow included over $7.7 billion depreciation and amortization.* And guess what...Micron also paid over $1 Billion in stock-based compensation to its executives charged as an expense for profit/loss but also have zero impact on cash.

There was a $7.7B expenditure for Capex ("Property, Plant and Equipment") that does impact the cash flow.
Their cash flow only ended up being positive because they took out $6.7B in debt.

You're only focusing on the lines that don't affect cash flow but ignoring the ones that do. What kind of accountant are you?

 

[loby]

Not me because I put it in quotes.

Though a serious comment, your choice of the word “Pleistocene” was both appropriate and actually “funny”.

I stand corrected.

 

[Smartuser]

This reply sounds suspiciously like it was written by AI, but at any rate, I am not particularly a fan of AI at the moment, regardless of which country it is from. Sure, it has its useful moments for me, but I dislike how it is responsible for devaluing and deskilling so many people, the environmental impact, the blatant disregard for intellectual property rights, and honestly, I don't see how these companies will ever be profitable.

The Ai bubble could pop tomorrow, these AI companies could go bankrupt overnight, and I wouldn't shed a tear.

The problem with the term "bubble" is that when a bubble pops, nothing remains. After the Dotcom bubble, internet businesses kept existing, and many have thrived. Housing has survived housing bubbles. AI isn't going to go away, but the business model is likely to change.

This isn't going to be done primarily via data centers forever. Those have issues of latency, trust, scaling and capacities (and therefore cost). It'll be hybrid, and move more and more to local hardware. The transformation will take a few years, of course.

Which is apparently an even greater moat than people give it credit for, given that nobody else is doing it.

I'd rather say that this is because it's been efficient up to recently, and that has ground to a halt now.

I'm not an economist, but it looks to me that one way many monopolies develop is that something is the most cost-effective option for a while, so everyone goes with it and builds up the silent monopoly until it isn't so silent and cost-effective anymore. For a comparison, Apple used to make everything in China but is now slowly diversifying its production to reduce its dependency. Some of my recent Apple products came from Vietnam.

There are some other hardly known near-monopolies like Luxottica, but for now, they're not calling the shots and their pricing is acceptable to most people apparently.

For the longest time, the US has outsourced whatever they could to other countries. Everybody wants cheap stuff (many of my friends are increasing ordering their furniture from Taobao), but nobody would ever work in the conditions under which they were made. You expect me to believe that thousands of Americans can magically be convinced to work in ram factories when the time comes, and long after the demand and prices of ram falls back to normal?

I haven't heard anyone ask for these to be built in the US.

People are losing their jobs (and finding it harder to secure employment), wealth disparity is greater than it has ever been, and now prices of electronics are increasing everywhere. You wonder just what the final straw that breaks the proverbial camel's back will be with regards to AI sentiment. 🤔

Are we allowed to discuss politics here? I have answers, but I don't want this post to be deleted.

 

[Abazigal]

Are we allowed to discuss politics here? I have answers, but I don't want this post to be deleted.

Truth be told, I am still not very clear on what constitutes politics here or not, so yeah, I will probably stop here as well.

 

[vantelimus]

When the AI investment bubble pops and financing for hyper scale data centers halves, what will these memory companies do? They will have overproduced GDDR7 and the insane profit margin will be pulled out from under them. I suppose they will start trying to readdress their DDR5 market, but Micron has closed its consumer production plant. They will be at a deficit wrt to other memory producers. The CBO will probably find a way to blame Ternus for not putting Micron's expensive VRAM into Apple products–actually it will be his successor, because the current CBO will have been fired for missing yet another big market movement.

 

[Abazigal]

When the AI investment bubble pops and financing for hyper scale data centers halves, what will these memory companies do? They will have overproduced GDDR7 and the insane profit margin will be pulled out from under them. I suppose they will start trying to readdress their DDR5 market, but Micron has closed its consumer production plant. They will be at a deficit wrt to other memory producers. The CBO will probably find a way to blame Ternus for not putting Micron's expensive VRAM into Apple products–actually it will be his successor, because the current CBO will have been fired for missing yet a big market movement.

It would be funny if by then, Apple has already managed to find alternative sources of RAM and leave Micron to choke on their own oversupply. 😏

 

[boswald]

"What? No, it's not our fault! It's the customer's fault!"

Blaming others is something kids do. If you screwed up, own it.

 

[amartinez1660]

It’s like a reverse complain reading statement: “we are making more money than god currently, because we can and chose to (explained by other commenters). Difficult situation to be in thanks to Apple some *wink* companies”

In any case, what’s the excuse now? Now they have the money to invest, they have had it since, like, the first data center future memory deal several years ago.

 

[maflynn]

Blaming others is something kids do. If you screwed up, own it.

New to the planet earth I see. 😛

 

[Fallforward]

Yeah, Micron should risk total bankruptcy when the memory market in 2023 was trash and they were posting $6B in losses. That $13B was needed to keep the business afloat for the next few years if the market was still in shambles.

Blaming Micron for not predicting the growth of AI data centers is an unreasonable take. Almost trollish. If it were that easy to predict where the market was heading, why didn't Apple lock up their memory supply with a LTA far below current rates?

I hope you are not a CFO or in any position that deals with a lot of money, because this is perhaps the most stupid thing I've heard this year. If I proposed what you just said to my execs, they would've fired me on the spot. It's so astoundingly dumb.

Learn to be mature and civil.

Learn how accrual accounting differs from cash accounting, particularly how net income significantly changes. Companies like Amazon and Tesla were able to invest with negative net income to up to 10 years.

Learn that high level executives tend to be mature and civil people with open minds. They are more likely to appreciate "dumb ideas" as out-of-the-box thinking. In 20-20 hindsight, that astoundingly dumb idea proved to be brilliant. Open question whether the companies driving the chip boom like OpenAI are astoundingly dumb or brilliant.

Learn that no company helps suppliers by just "paying more." They make strategic investments. For Apple that includes Corning, TSMC, Dialog, Finisar (financed a plant in Texas) among others. During 2023 it was Globalstar, financing a contract to pay SpaceX to launch satellites and investing an equity stake.

Apparently Micron never engaged with strategic win-win investement ideas like all those other companies because all he does is whine about price.

An astoundingly dumb idea would be for Apple to watch Micron pay out nearly a billion cash dollars in dividends/buybacks and then, out of the goodness of their heart, pay an extra billion dollars for Micron product.

 

[BSDnostalgia]

Apple has repeatedly demonstrated they don’t care about anti-trust

I don't think very many companies of any kind are worried about anti-trust these days. But, Apple doesn't present a barrier to entry for any of Apple's product areas.

Unless, of course, you consider that Apple has a virtual monopoly on "privacy". But again, any company could start doing modest things to improve customer privacy if they chose to, at any time. No barrier. But, most companies are actively working to undermine privacy as part of their business models. Apple does have a level of integration in its "ecosystem" that is moderately privacy-preserving across the ecosystem. If only every company would do the same.

That doesn't change the fact that they lowballed everyone. There were plenty of articles about it over the years.

Well, I've certainly read that Apple is a very tough negotiator. But, some companies will and have signed contracts with Apple because it keeps the lights on. Different companies in different positions sign different kinds of contracts. That is "market capitalism".

Pure market capitalism doesn't work that well in some kinds of markets, like housing, and utilities. But, it seems to me that it worked very well in the cell phone industry. (Other than the side effects of gluing everyone on the planet to their phones.) Apple created the "modern" cell phone with the iPhone. It seems odd that so much angst surrounds Apple. If you don't want an iPhone, buy a -- so many hundreds of different competing models to choose from -- a Google Pixel 10 Pro XL, say.

 

[theorist9]

It seems odd that so much angst surrounds Apple. If you don't want an iPhone, buy a -- so many hundreds of different competing models to choose from -- a Google Pixel 10 Pro XL, say.

I don't know how much this applies to phones (since I don't know how Android compares to iOS), but it's entirely understandable that so much angst applies on the Mac side.

And the reason is that most Mac users are Mac users because they strongly prefer MacOS over Windows (or Linux).

If you are a Windows/Linux user, what a specific PC maker does isn't that critical, because you can simply switch to another.

But if you're a Mac user, that option isn't available to you (if you want to continue to use MacOS). Hence any changes to Macs that the user base doesn't like cause angst because that user base has nowhere else to go.

 

[Smartuser]

Learn to be mature and civil.

Great answer to many posts here.

Learn that high level executives tend to be mature and civil people with open minds. They are more likely to appreciate "dumb ideas" as out-of-the-box thinking. In 20-20 hindsight, that astoundingly dumb idea proved to be brilliant. Open question whether the companies driving the chip boom like OpenAI are astoundingly dumb or brilliant.

Every great idea started as something that in the equivalent of the social networks was derided as a "dumb idea". (The converse isn't true though, unfortunately.)

Apparently Micron never engaged with strategic win-win investement ideas like all those other companies because all he does is whine about price.

This is a poorly run, vision-less company making a product that others are making as well with the exact same specs. They got lucky with the current market, but their product is as innovative and cultured as crude oil. They act accordingly. They hire generic businesspeople to do generic things and issue generic statements.

I don't know how much this applies to phones (since I don't know how Android compares to iOS), but it's entirely understandable that so much angst applies on the Mac side.

And the reason is that most Mac users are Mac users because they strongly prefer MacOS over Windows (or Linux).

If you are a Windows/Linux user, what a specific PC maker does isn't that critical, because you can simply switch to another.

But if you're a Mac user, that option isn't available to you (if you want to continue to use MacOS). Hence any changes to Macs that the user based doesn't like cause angst because that user base has nowhere else to go.

The problem here in this forum is that there's a lot of people who aren't mature or civil, and certainly don't display any strong preference for anything Apple does. If I would hold these opinions, I would change my suppliers immediately. I've changed brands several times for several types of products because I was unhappy.

What I don't do (and what I'd think most reasonable people wouldn't do) is hanging out in forums that focus on my formerly used brand and expend energy to tell everyone how horrible I think it is. It comes down to being mature and civil.

 

[bgillander]

Every great idea started as something that in the equivalent of the social networks was derided as a "dumb idea". (The converse isn't true though, unfortunately.)

Honestly, I would expect that there were actually more "great ideas" that were originally pitched and believed to be world changing yet eventually turned out to be dumb ideas, versus those unicorn "dumb ideas" that eventually turned out to be brilliant and world-changing.

And I am trying to account for the extra "random average person in forum" bias in that, though I will admit that specific and very random unpredictability is so very hard to measure.

 

[visualseed]

Micron blaming Apple for the memory shortage is classic corporate deflection. Apple just scored a great deal during a market slump. It’s not like Apple cornered all of Micron’s supply because Micron was still serving the rest of the industry and selling straight to consumers through its Crucial brand. Claiming Apple’s pricing stopped Micron from expanding is ridiculous. Chipmakers build new fabs based on global demand forecasts, not what one client is willing to pay. Even if Apple had shelled out more cash, Micron wouldn’t have used it to boost capacity because overall market demand wasn’t there. They’d have just kept the money to ride out the downturn instead of building imaginary production lines for customers who didn’t exist.

 

[Jeeve Stubbs]

It's a humorous comparison to your tone-deaf way of mandating what is a luxury and what people should do or like.

There are lots of products that people buy that I think are in atrociously bad taste or generally pointless, but I don't venture into their forums to tell them how great it would be if they couldn't afford them anymore because I think they should by different things.

I don't need a lecture from you. This isn't "their" forum, but a forum for all members.
Apple products in general, and its top-tier items in particular, are pretty much luxury items if only because of the margins Apple makes on them. I certainly won't shed a tear because Jimmy's annual Ultra Pro impulse, which he buys on credit, will cost him $2,000 instead of $1,700 the next time around, and he might decide to pass as a result.