How can Apple deal with the memory shortage? | Hacker News Reader

How can Apple deal with the memory shortage?

asymco.com

136 pointstambourine_man192 comments

Just a thought - stop bloody soldering the RAM and the SSDs. That partially transfers the burden on to the customers and makes their product more repairable ... Mac Minis (with un-soldered RAM and HDD / SSDs) used to be sold with minimum 4 GB RAM, if I remember right). This, way, you can still sell devices with lower RAM, and customers can upgrade in the future when supply increases.

Another problem is locked hardware. Newest Synology hardware has lifted the HDD locks but still doesn't allow third party SSD and RAM. Mac Mini storage upgrade is a DIY solution, why not use the standard M.2 2230 slot?

> Newest Synology hardware has lifted the HDD locks but still doesn't allow third party SSD and RAM.

True of NVMe SSDs, but SATA SSDs are no problem.

Yes. These units now come with dedicated NVMe slots and they don't accept third party drives.

It takes up more space and costs more (connectors are surprisingly expensive), as well as adding an electrical overhead, while most (yes, not all) customers don't take advantage of it.

Also somewhat less secure (cold boot attack: https://wikipedia.org/wiki/Cold_boot_attack#Preventing_physi...)

Not wrong, but some of these space/size/distance concerns were some reasons why the LPCAMM2 format was created:

* https://www.ifixit.com/News/95078/lpcamm2-memory-is-finally-...

* https://www.micron.com/products/memory/dram-components/lpddr...

* https://en.wikipedia.org/wiki/CAMM_(memory_module)

That's an acceptable compromise to get a system that can be upgraded and / or repaired. Apple already charges a "premium" so nobody (except their shareholders) are going to cry if they make a dollar or two less on each device because of this. Customers do care about such things when they decide to upgrade their device or, more importantly, when they have to repair it.

100% agree the problem is modular RAM can't handle the timings of the SoC.

There is a new modular RAM standard for precisely that but knowing Apple they will want to make their own.

SSD should be easy but since RAM does not last that much longer you still need to resolder that after 5-10 years!

you can have a mix- 4gb embedded ram + 1-2 slots of slower layer

Pretty difficult to code OS to take advantage of that. Basically need NUMA, which increases overall overhead.

Otherwise, you may end up filling up your fast memory with some cold data.

If apple cant support NUMA ill eat my hat

but since RAM does not last that much longer you still need to resolder that after 5-10 years!

Where are you getting this information?

Even though I don't necessarily like it, I understand why they solder the RAM on the SOC: Higher bandwidth/greater performance, better power efficiency, etc. But they have no excuse for the SSD.

The excuse for the SSD is that the controller is on the SoC

The shortage that connects to a modern Mac isn’t an SSD — it’s raw NAND.

Why does the controller need to be on the SoC when the market already has SSDs with built-in controllers that follow a standard and allow you to easily replace it?

It doesn’t have to be. It is an engineering decision that has design consequences. Integrating it likely has advantages in packaging, performance, power consumption, etc.

This makes everyone’s computer slower, more expensive, and less power efficient, and 95-99% of people will never open their computer anyway.

95-99% of first owners, maybe. But when you make devices that can be affordably repaired / expanded, they will be - and then they gain another 5-10 years of useful lifespan for a second owner.

If we ever want computers to be sustainably made - instead of scorching the earth with each new device - we need to stop thinking the way people treat their devices is some natural law of how things will always be.

> If we ever want computers to be sustainably made - instead of scorching the earth with each new device - we need to stop thinking the way people treat their devices is some natural law of how things will always be.

If this was solved by upgradable components, we would have "solved" e-waste in the 90s.

Component upgradability is not a sustainability solution, because it is architecturally bounded.

Hard disagree. First of all, it’s true that upgrades were not a solution to sustainability in the 90s, but the issue was that computers were gaining ground exponentially. In just a few years, everything you had was so outmoded it was hard to imagine hardware that lasted more than maybe 5 years. That has changed.

Second, sustainability isn’t a true false state. My previous computer, a 2015 macbook air, lasted me until this year. If it had upgradable ram it might have lasted me another 5 years. A computer that lasts forever is probably impossible, but 15 years is better than 10, and 10 is better than 5.

In the same vein, as hardware changes have slowed down since the 90s, so have the causes of obsolescence shifted away from CPU speed or RAM capacity. Lack of TPM/secure boot and end of Windows 10 support is what has recently pressured a lot of PC users to upgrade. End of support for x86 MacOS is what will cause a lot of Mac users to upgrade.

Anecdotally, I bought a HP laptop in 2014 with the basic 8 GB of internal RAM (4+4) and two 8 GB chips from a third party vendor. I opened the box and immediately threw away the 4+4 and replaced them with the 8+8. As you can guess it was much cheaper than the same 8+8 from HP. With soldered RAM one has to swallow whatever price the laptop manufacturer charges.

If you're generating e-waste right out of the box, I wouldn't say that's a plus.

You are concerned about 2 sticks of RAM as "e-waste" (even though it can be reused), but not about a whole Mac ending up as an e-waste because it can't be repaired or run other system softwares?

I think if you’re concerned about ewaste at the end of a product’s life you should be equally concerned about products that are ewaste at the beginning of their life. Most PCs are never upgraded, and their socketed components end up in the landfill in vein and at a higher rate.

PC's that can be upgraded form the bulk of the second-sales industry and don't end up in the landfill for a few more years because they can be upgraded / repaired and used for a few more years.

The thing is why do they have to do so? In terms of bargaining power, Apple has the best position comparing to other cell phone/ iPad/ PC vendors. Their software are way more memory efficient than the competitors (like Windows 11 with all the bloated Electron/ Webview2 wrappers).

Bloated Electron / WebView wrappers also tout the feature of running on macOS - and many are forced to use such apps on the mac too ... what then? Apple deciding how much RAM is "enough" for you is based on the use case it has decided for each category of the device. If that isn't sufficient for you, you are supposed to buy the more "premium" edition, even if your use case for that is rare. Imagine if you suddenly need more RAM for a popular app 2 years down the line but you can't really do anything on a Mac now a days because the RAM is soldered and fixed. Buying upgrade-able RAM is obviously much cheaper, and much better for the environment, than buying a whole new device.

I am just sadden to see this question constantly being floated. Someday Apple will make their own memory, own chip, own Fab. It makes zero sense for them to do so. People don't seems to realise how commodity works. When price is high everyone is saying it is a cartel. When they are losing money they say it is tough luck.

Yeah. Everyone seems to forget that making memory was considered a bad business just a decade ago.

The current memory crisis it’s just another/final nail in the coffin for Apple moving the design and engineering of memory in house they have the capability and they have the money to do so.

May be Apple should purchase Intel or TSMC first? So they could be their own Foundry? What about ASML so they could produce the machine that fabs requires?

And the current memory crisis? There is no crisis for Apple. There is only crisis for others buying at spot price.

Apple doesn’t need to buy them just need to continue to partner with TSMC, as far as Intel is concerned why would they buy or work with them now? Intel is absolutely low end these days and they were the company that said no to Apple when Apple wanted them to design the chip for the iPhone, Intel couldn’t imagine a future where computing was smaller better faster and hence they missed the mobile revolution take Intel at their word the future just isn’t for them.

>Apple doesn’t need to buy them just need to continue to partner with TSMC,

So why would Apple need to make their own memory? The theory is exactly the same. They just need to continue to partner with their NAND and DRAM vendors.

Got a shitty PC with 32gb ddr5 now the ram alone is almost worth as much as the purchase price of it all. It is playing up.. normally I'd return it to Amazon but...

I wonder if companies like Apple will eventually start making memory themselves.

This is probably the natural conclusion but it will take some time to get there

in a sense that's exactly what cartel wants - to lure out investments that will get squashed into uselessness by supply flood that will follow

Who is the cartel?

SK Hynix, Samsung and Micron?

They should be banned.

This AI memory crisis will be their last big payday. I think the Chinese will take advantage of this and take over the memory market worldwide, excluding the United States and some parts of Europe. The rest of the world will end up using Chinese memory.

The key is Apple can be their own customer and just not care anymore.

It’ll probably only be worth it if it enables something “new” like more bigger Ultra chips or something.

and that's part of risk management, on both sides

does Apple have enough of a design moat to overcome eventual overprice compared to competitors when "outisde" production is 2,4,10x cheaper

does Apple have enough of income/savings to maintain internal production capacity if it decides to switch back to outside sources

or can Apple acquire enough fab competence to negate internal/external price difference

we'll see how it plays out

long-term if you want to build new devices, which are smaller faster better memory is gonna have to come in house. How does the Apple Vision, Apple Watch get smaller and four or five times faster? How do you make those Apple Glasses in the future if you don’t have that capability memory, modem etc.. in house?

they blew it! They could have bought Intel for cheap and made memory AND CPUs!

Apple knows better than to buy a pile of incompetent smugs. Intel was rock bottom before Europe determined it was a “strategic move”[1] to buy factories in Europe from the only manufacturer that hasn’t innovated since 20 years, quickly followed by the US. In both cases, governments aren’t the most savvy spenders.

[1] A “strategic” expense is named like this when you can’t justify it by any rational means.

I shall keep this brief. Intel sucks. Intel and Microsoft rested on their laurels and now they’re playing catch-up Microsoft may hang in there, but overall worldwide Intel is in trouble.

I would suspect at Apple scale it makes sense.

Apple has started making a lot of different things in house, its only a matter of time imo.

I doubt they want to make a commodity.

But who knows. Their unified memory architecture across core types already puts them in a different design space. Maybe that design space leads them to further opportunities for memory architecture differentiation.

I could see them (1) taking the two processing chips that make up an Ultra in coming generations, (2) fabbed with logic on top, and power distributed on the back side, as Intel is going for, and (3) sandwiching the logic sides around a layer of unified RAM, with (4) massive optical linking distributed across the surfaces, resulting in (5) unbelievable bandwidths and parallelism we couldn't dream of today.

And then, (6) announcing it at WWDC 2029 and (7) taking my money 5 minutes after the midnight when pre-order's start.

(5.5) cool the whole thing in a way nobody else manages because of their vertical integration.

It memory almost certainly is coming in house all you need to do is look at Apple’s history Intel out, Broadcom out, Qualcomm on the way out, Nvidia has been out for a long time, AMD is also out. Apple has the capability and the money, the current market conditions have changed I believe Apple will make plans and move on.

> Apple has started making a lot of different things in house, its only a matter of time imo.

Define "making". Sure, they design a lot of stuff in-house (CPUs/SoCs, wireless chipsets, etc), but they do not manufacture these things in-house: they have no fabs themselves.

Definition: Apple is the architect and engineer, and someone else will do the fab probably TSMC who is also motivated to keep their engines running?

Given the current issue is with fab capacity, what would Apple gain by being the architect/engineer of RAM? They hand off the 'blueprints' to someone who is already at capacity?

And it's not about motivation, but capacity: everything is currently at 100%. To get more capacity means building more fabs, and given the historically cyclical nature of semiconductors not may people want to necessarily take that risk (potentially end up with over capacity). And it wouldn't help now, as it takes >2 years to build a fab.

Unless the suggestion is for Apple to design M- and A-series CPUs with the memory circuitry in-die instead of on-package? (Using up some 'transistor budget' for RAM.)

Once again, Apple is already given the bums rush to five other companies when they got in the way of what they wanted to build Apple isn’t going to sit around, based upon their recent history in the last 20 years and do nothing.

I don’t understand why that’s so hard to understand and I certainly don’t think Apple will hesitate with a new technocrat CEO coming in September 2026, if Apple had a “slug” MBA person coming in September, I might buy the argument that they may not do anything but they have shown that they are capable of playing the long game when necessary.

I wish I could’ve been a fly on the wall 5-6 years ago to see what the conversation was like on Hacker News when rumors of Apple replacing Intel came up. I bet it was much of the same. Oh no, don’t do anything.

Or they can go to existing manufacturers with bags of money and have the experts build them their own production lines, and secure the supply.

But memory is a quite a specialist manufacturing process, they couldn't just send a design to TSMC and get the same quality and cost. It would take years (decades) to create their own factories that might be able to produce competitive memory. If they use a third party to manufacture with existing skills (Samsung, SK Hynix, Micron etc) they might as well just use their designs too (and buy their chips)

"Decades" seems excessive

Static RAM needs 6 transistors per cell, an M2 Ultra has 134 billion transistors, do the division and you get... "it doesn't work that way".

Static RAM is not DRAM.

You certainly could do try a 20bn cell SRAM, in 155mm^2, if you could handle the routing, but the power consumption might surprise you.

My Guess

* Absorb the impact by some margin * Slash base models (which they are already doing) * Efficient software - So, end user experience is not affected. * Direct Price hike always be an option.

#3 may prove difficult besides Safari

Pausing non-foreground apps and compressing/swapping out their memory seems to work well on the iPhone.

You are correct. Though I’m talking more about never-before-solved in-foreground problems such as: Spotlight Search or SwiftUI List views.

This reads like Apple fanfiction to me.

> But then Apple can negotiate on another basis and say, well, if you don’t do us a favor here and give us a better rate, then maybe we won’t work with you when all this settles down. You know things are going to settle down. These things are always cyclical. There’s never been a semiconductor boom that’s not followed by a semiconductor bust. Never. And they know it.

I have to think that the RAM suppliers wouldn't be that easy to intimidate with threats, since they know perfectly well how few alternatives Apple has. And they are also perfectly aware that Apple will play hardball with them when the market turns, regardless of whether they were nice to Apple now.

Apple bought PA Semi as the starting point to getting off of Intel. Theoretically, memory seems like something Apple could figure out how to fab. And it's not like they don't have any capital reserves.

They bought P.A. Semi, but it was for their design capability; they never had fabs anyway, and Apple still depends on TSMC and others for manufacturing chips. Apple building fabs to ensure a guaranteed supply of memory (or logic) chips would be an unprecedented level of vertical integration, even for them.

SpaceX/xAI is investing in creating their own fab. If they can, Apple certainly can.

We don’t know if SpaceX’s plan will actually work, they announced it this year and it is a long-timeline Musk project. These have… mixed results.

yeah. right after tesla self-driving.

that's crazy. Looks like Intel is part of this https://www.reuters.com/business/autos-transportation/intel-...

Part of the trump bullshit that the USGov has taken 10% of Intel.

It's actual corruption and the standard fascist model of corporate takeover by the state.

They can’t.

Operating a FAB requires employing PhDs that are willing to work 8 hours shifts with no breaks (each removal of a bunnysuit is an expensive exercise), and there’s no reason to believe SpaceX is capable of hiring such people.

There was a point made recently by Musk that the whole clean room idea is outdated if you can just ensure the path the silicon takes from wafer to lidding is clean. Seems solvable to me, but leaves me wondering why it hasn’t been done before. I assume there is no human handling of raw/etched silicon now anyway, so why does the whole room need to be clean?

hmm yeah. its cool that musk knows more about this than the entire industry

You could probably apply that logic to any innovation in any industry no?

Reusable rockets likely got the same ridicule, as did fast satellite internet, self driving and fully electric vehicles.

I can understand that Musk does not have the most palatable personality, but floating ideas and at least attempting innovation regardless of outcome over a long time is a net positive for society and should not be discouraged.

> self driving

Aren't we still waiting for that?

Reusable rockets likely got the same ridicule, as did fast satellite internet, self driving and fully electric vehicles.

In those areas, Musk successfully leveraged government largesse to compete with fat, lazy incumbents who had either coasted for decades (rockets and satellite Internet) or who didn't bother to show up to the game (EVs, self-driving and otherwise.)

That does not describe the semiconductor industry.

Musk has never beaten anybody who actually put up a fight, as far as I'm aware. I guess Blue Origin technically counts, but again that's not exactly TSMC.

Intel and Microsoft are having that same problem right now when the playing field is level, they have trouble competing.

It wouldn’t be the first time an industry got bogged down by prior knowledge. Hell, it happens to all of us.

think different

The semiconductor fab process changes dynamically to manage yield. It is not a static environment, automating with robotics is fine when things are static like a automotive assembly line, but high end semiconductor fabs are a different beast (The analogy I heard was repairing a plane while in flight). Robots are not purely clean as well they shed contaminates as well, which must be managed too. Entropy is the reason why we still need humans in the loop.

So what? Maybe a hand full of full bunnies per shift, and another dozen or two half-bunnies. There aren't more. This can be seen/validated by some older yt-videos, where something went wrong in the fab, for instance a https://en.wikipedia.org/wiki/FOUP ejecting a wafer in wrong ways into a machine, then being ejected by that onto the floor, and shattering. Causing all systems to stop, and all the warning lights beginning to blink in an expanding cascade. At about 4:30AM. Maybe 20 seconds later two half bunnies with face masks appear, another 10 seconds later a full bunnie. Some gesticulating ensues, full bunnie opens his suit, gets his flip phone, half bunnies downing their masks. All looking very concerned and exasperated. Having a really bad day. No more bunnies appear over several minutes. Video ends.

You've got to wonder why. It makes no sense for xAI to make their own chips.

The "integration" of SpaceX/xAI is just standard Musk-move-losses-to-the-company-making-money-at-the-moment bullshit.

Apple actually have the runs on the board, xAI has Musk-BS.

>It makes no sense for xAI to make their own chips.

The initial investment in chip fabs is so big it can't be justified when the established players already make enough to satisfy demand, but right now they don't so there's an opportunity.

It's still risky for sure but it makes some sense that it happens now. Hyperscalers spend 100s of billions yearly, at some point the amount given to TSMC gets larger than starting your own fab.

If success was guaranteed (it's not, as AMD and several others have learned) I think many more co's would start their own fabs in the current market.

As for why xAI, well why not - many of the others who can afford a fabbing attempt can't risk getting on TSMC's bad side even for a year or two.

The Chinese will be the one slipping in because of this opportunity. The question is is whether or not you still want to be dependent on outside memory when the Chinese takeover a larger part of the worldwide market?

No RAM, no profits. Apple has vertically integrated in the past for less reason than this.

Moreover it's a massive economy of scale, while their consumer electronics competitors are busy fighting a losing battle against the server market for chips, Apple can undercut them, grow their market share and get even more service revenue.

RAM prices surging in the AI hype era does not mean they'll stay there for decades (see xAI already letting one data center go), and it would take a long time for Apple to become competitive.

Should they also start CPU fabs? Batteries? Lithium mines?

The risks are not symmetric. If the RAM crisis becomes the new normal it threatens Apple's business model which requires large quantities of RAM.

On the other hand, if Apple invests in RAM production and prices fall, it's not like the investment is wasted, RAM is a commodity. They lose at worst the opportunity cost of deploying the capital inefficiently, but they have so much that it hardly matters.

Apple should take this crisis as a warning that they aren't vertically integrated enough to protect their business model.

As for batteries, Apple is not even close to the largest consumer of batteries. If they were an electric car company then yes they should be making their own batteries.

That’s a decision for the new CEO (thank God he’s a technocrat), more than likely Tim Cook, and John Ternus) probably have already decided on what they’re gonna do long-term, from the outside looking in Apple has already replaced five companies? In recent times.

Memory is well within Apples design and Engineering capability. Long-term, Apple has to think about the Chinese getting a bigger part of the market in memory because they can undercut the three company cartel worldwide in time with this fake AI memory crisis.

Apple in the same timeframe also bought Intrinsity, and Anobit (a flash/SSD memory) company the Apple Silicon design group probably can do the design and engineering in house and we know they have the money the question is do they have the will their history says they do.

https://en.wikipedia.org/wiki/Anobit

In the Tim Cook era when Apple needs to lock down the supply of a commodity part, they have a history of buying a dedicated manufacturing line for a manufacturing partner.

It's crazy to think Apple would actually fab memory (or TSMC for that matter). It's an entirely different process than logic.

DRAM fabs are their own well-known specialized process which is covered by the DRAM companies. It doesn't make sense to start a competitor for it.

There's a bunch of chinese DRAM companies currently playing catchup to get closer to modern densities. Could Apple buy one of those? I'm guessing there would be regulatory hurdles to that on both sides of the pond.

Apple never bought a manufacturer, or built such capabilities.

They buy and build manufacturing capacity, and there’s also a huge shortage in that today.

China is playing a totally different game here.

Can a US company by a Chinese company... no. Number one, China won't let them. Number 2, China is building up these companies as a strategic reserve against the US/Korea for when they eventually go to war. So, yea, eventually the US will ban any imports of memory from those companies which would turn it into a toxic asset for Apple.

Which is funny, since until relatively recently DRAM was what you produced in fabs with processes that weren't competitive enough for CPUs anymore.

They probably could, but time is a big factor.

Yes, the author knows very little about the industry or how Apple operates. Fanfiction indeed.

They book manufacturing capacity often years in advance. Samsung is their majority RAM supplier and they reportedly agreed to doubling their price a few months ago.

https://finance.yahoo.com/news/samsung-100-ram-price-hike-12...

The original article is baseless speculation proven wrong by news announced in February.

> Yes, the author knows very little about the industry or how Apple operates.

Hardly. While it may be fan fiction, or speculation, Horace has been researching and writing about Apple's operations for decades. I tried listening to his podcast years ago and the discussion at the time of Apple's supply chain movements was extremely detailed to the point where it wasn't even listenable for me.

"Our team has over 25 years of daily research on Apple Inc"

https://asymco.com/about/

It's literally all they do

Ask Intel, Broadcom, AMD, Nvidia, Samsung chip division and soon to be replaced Qualcomm, Apples SOC designs, probably meant memory was coming in house at some point down the road anyway. The present market conditions will probably just hasten the inevitable move.

After all, how does one miniaturize future SOC devices if you don’t bring memory in the house eventually?

Ex-Apple kernel engineer here, Apple will deal with the memory shortage by making software more efficient in ram usage. Apple will just make every aspect of the system more and more memory efficient. They've done it before over and over and can do it again.

i wonder if this is the real reason behind the push for the snow leopard like release this year

Apple? Sure. What about other developers? Firefox, Chrome already use gigabytes of RAM.

Suddenly Safari can surge ahead again!

welcome to the rust community

It's the websites that use that RAM, not the browsers.

(Often the ads on the websites.)

I disagree, there is low-hanging fruit Firefox is leaving on the table. The main thing that comes to mind is tab unloading. They don't unload tabs automatically like chrome can.

I was pleasantly surprised at the tab unloading settings under "memory saver" in ungoogled-chromium.

Firefox has been unloading tabs for several months or so (at least on nightly).

I believe that Firefox does it but not as frequently as Chrome does, but don't quote me on it. However, I am using a "tab suspender" addon on firefox to control how fast the unloading happens on tabs that are not active.

Indeed, I am using something similar "auto tab discard".

Firefox unloads tabs under memory pressure since more than 4 years: https://hacks.mozilla.org/2021/10/tab-unloading-in-firefox-9...

It is a naive and suboptimal implementation, they even describe it in the link you posted

"We have now approached the problem again by refining our low-memory detection and tab selection algorithm and narrowing the action to the case where we are sure we’re providing a user benefit: if the browser is about to crash."

I would prefer FF to be more proactive in unloading tabs way before "its about to crash" to keep system level memory pressure lower. Firefox is the main memory hog on my M1 mac.

Chrome can do this, there is no reason we should be stuck with "manual tab unload" and "unload when the browser is about to crash".

I am using an extension, but that just reinforces the argument: they could be doing much more here.

Which extension are you using in Firefox for memory unloading?

And it’s the applications using web browsers as their UI kit that are the worst offenders in my experience.

Browsers still have a lot of memory usage on their own.

I am running Arch Linux here. When I boot my machine into a full desktop environment it uses 1.1 GB of memory total, for everything.

If I open Firefox, it in itself uses about 1.3 GB to have Firefox open with just HackerNews in 1 tab. I have no extensions except uBlock Origin.

Read about:memory then.

No, it's the browsers. Check how much memory they commit and how much is actually resident. Firefox often commits 2x more memory than it is actually using.

Fortunately, Apple devices only run approved software. Google will be forced to optimize memory or become unavailable on those devices.

Signed software. Not approved software. Mac apps can be installed after being downloaded from the web.

And as if Apple would ever block/pull/disapprove the world’s most popular browser.

Aren't they actually blocking alternative browser engines on IOS still?

Yes

not in the EU

also in the EU

They won’t be able to do that for AI models, because they suck at AI.

This is a great long term strategy despite what the share holders would want to believe. If you increase efficiency even on lower end devices, you will get people coming back for more. It isn't the sale today, it is the sale tomorrow that matters.

I read this as satire.

Because

I've been running an M1 Air w/ 8GB for a few years, and it's still working fine.

Me too but the latest macOS version has ruined it for me, I had to switch back to a previous version.

can you still go online with it? do you have access to security updates?

Apple provides security updates for the current and two prior versions of macOS. Occasionally, critical updates for older versions.

But how are Neo users dealing with this? It’s a new machine, surely it works with 8GB?

My Neo runs like a sowing machine with 8GBs. So well that I honestly cannot think of any instances where I have felt the system slowdown. I use the device appropriately though -- I'm not trying to run SOTA LLMs locally or anything. For web browsing, light programming, and an iPad replacement, the Neo has exceeded my expectations. Compared to my M4 Mini, I honestly cannot tell a difference.

Better performance will be achieved by adding more whitespaces and increasing the radius of rounded corners.

With no memory balooning device in sight for macOS virtual machines I don't really see Apple moving in that direction.

Apple also has all that fast flash to swap to. I never notice when I'm swapping. Even a Neo has fast enough flash to handle a little swap, as a treat.

You can thank Apple for buying Anobit many years ago…

"So much so that I heard Samsung’s making more money now with memory than Nvidia’s making with their processors."

I loved Asymco during the Apple 2010s run up, but this, inter alia things mentioned in other comments, should give the reader pause and evaluate how much of this is general knowledge x handwaving x vibes versus a practical ground floor understanding in 2026.

Apple can deal with the memory shortage in the same way they dealt with Intel saying no to building a processor for the iPhone the Apple Silicon design group can certainly do what’s necessary in house to design memory and SSD’s in house, and since Apple saved money on not participating in the AI data-center fiasco money won’t be a problem.

Apple can team up with TSMC to build some type of memory fab in the United States may take three years? Prineville, Oregon looks good close enough to water and Micron who won’t need many people once the Chinese use this memory crisis to take over the worldwide memory market…

https://en.wikipedia.org/wiki/TSMC_Arizona TSMC announced the Arizona Fab in 2020 they won’t have the final build out until 2029 for the 2nm (Build in house go around)

> since Apple saved money on not participating in the AI data-center fiasco

They have unused custom built AI servers sitting in warehouses: https://news.ycombinator.com/item?id=47221264

Don’t be obtuse, what Apple didn’t do and isn’t doing is spending billions upon billions and dollars trying to train AI models like Meta, Google or Microsoft, Oracle and OpenAI those five companies together have spent well over $1 trillion dollars? In their efforts to date.

Now I know Apple has given a measly one billion dollar refund to Google so far and they’re not paying OpenAI anything as far as we know and internally, maybe Apple spent a couple of billion dollars, which is less than what they spent on their rumored electric car so I think they have the money.

I think it's kind of inevitable for somebody to take up a RAM factory to sell for normal consumers and brands even if it means less profit the sales would be much bigger. The question is if somebody is kind enough to do so.

I read somewhere that there are other memory manufacturers in CN that are pivoting to consumer RAM to fill the gap.

Personally I think the AI boom will crash out when all of these datacentres get rationalized. RAM manufacturers will crash as is normal in the semiconductor market.

Memory optimization is incredibly low hanging fruit. We have been so spoiled with the abundance of memory.

There is just so much code out there that does not manage object life cycles well (over allocation, leaks, etc.), encodes data in text rather than binary representation.

The move to static binaries over dynamic libraries, applications that run web engines underneath rather than cleaner UIs.

I hope the memory shortage will encourage us to focus on efficiency again.

Can't even find a ddr2 sodimm that's not a ripoff.

This makes me wonder when we'll start trading memory on the commodities markets.

The author doesn’t seem to understand that Apple places RAM orders years in advance. I’m not sure if it’s even feasible or possible for Apple to fully integrate their supply chain and open up memory fabs, the cost of entry must be enormous.

And by "places orders" we mean "helps TSMC acquire plots of land on which their next facilities will be constructed" kind of level of scope, timing, and commitment.

TSMC doesn't make RAM do they?

Fair, and I meant it as illustrative of partner depth generally rather than as a specific example around RAM.

No, but they can and so can Apple if it becomes critical, I don’t know what is more critical to Apple than replacing Intel, Qualcomm, or Nvidia, but memory probably is number four on the list which means it probably is something that will be addressed?

Yes I believe that’s what being a manufacturer partner entails

In my experience, the corporate-speak "partnering with" can mean almost anything.

Apple gives TSMC a billion dollars to build a cutting edge fab dedicated to making Apple's chips, a deal they repeat several times over more than a decade? Partnership.

Youtuber takes $300 to read an ad, giving viewers a 10% discount code? Also a partnership.

There's partnering with Apple for several decades where they plan years in advance and pay billions without fail, and there's partnering with OpenAI where Sam Altman commits to giving you a Trillion dollars provided you can deliver all that ram up front and he can give you an IOU he got from Oracle who got it from Nvidia who got it from OpenAI. These are different things.

If Tim Apple can't beg China for more while in Beijing then I guess they need to port SoftRAM 95 to OSX.

Maybe even MagnaRAM after that. It had a crocodile on the advert.

Their best strategy is to buy Micron Semiconductor 12 months ago with cash equivalents on hand, for $106 billion.

No brainer. Best move they will ever did.

Our problem is lack of competition

High prices for RAM should attract competition.

In general, no.

It takes billions to tens of billions to setup a fab. It also takes years to get it working. Then when you add in the IP for memory, it pretty much ain't happening.

All the RAM monopoly has to do is wait 3 days before you're producing and drop the price and you're ruined. Meanwhile they've built up a battle chest of hundreds of billions in profits.

China might be the only competition we see come out of this, but only because they are playing the long game and have trillions of US dollars to play the game with.

There are a lot of companies that have billions in cash and are also prodigious buyers of RAM. Companies like Apple, Google, Meta, Nvidia...

Do they want to get into a commodity business like RAM production? Maybe not, but if prices stay high long enough that demand for their products falls off, they might think about it.

I know that I personally and my employer are cutting way back on new technology purchases and squeezing as much as we can out of old equipment due to the cost of RAM and storage now.

And none of these companies are operating their own fabs, that's the problem.

Fabs are a cutthroat business that's very hard to get into. It costs billions of continual investment to stay a float. That's why there's really only about 3 different companies with cutting edge fabs. TSMC, Micron, and Samsung. Even intel, who built a huge portion of their business on cutting edge fab tech, has struggled to keep funding it. AMD got out of the fab business almost a decade ago (spinning off global foundries) and that spin off is no longer cutting edge. AMD uses TSMC.

Fabs are some of the most expensive factories to operate on this planet due to a constant need for brand new equipment and cutting edge research. That's why there's not an Apple, Google, Meta, or Nvidia fab. That's why there's not an AMD fab. That's why Intel fabs are treading water.

Without the constant investment, you very quickly find yourself in the company of yet another cutthroat industry, the "not cutting edge" fabrication industry. And that, by and large, has already been locked up by about a dozen fab companies.

And a 64G DDR5 ECC DIMM costs $3K and is backordered. If ths isn't a bubble and demand persists, some new players are eventually going to want a cut of that.

We aren't talking about making new lug-nuts. A company can't just will a fab into existence.

For example, Micron is actively building a few new fabs. One of which has been in progress since Biden (pretty close to my home in fact). It's not going to be completed for another 5 years at a minimum. And this is a company that has the experience and partnerships for producing fabs.

Yes, a new company might decide they want to enter the market, but even if they decided, today, "Yes we'll do this" I'd expect a minimum of a decade before they start spinning out their first chips. That's also at least a $1T investment at this point to get started.

> And this is a company that has the experience and partnerships for producing fabs.

Not even they necessarily have the experience to do it! Intel has a policy called "Copy EXACTLY!" for fab construction where they make every irrelevant detail the same as their last fab, because they don't actually know which of the details matter.

Oh for sure. A story that gets told at the micron factory is that for a long time they'd experience chip failures at increased rates. During the night it was pretty good, but during the day these random failures would creep up.

After spending a lot of time studying the problem what they finally realized is they built the building too close to the interstate and vibrations from the interstate were ultimately making their way into the factory causing errors.

To combat this problem, they spent millions retrofitting shock absorbers onto the building.

It's not shocking that intel would do the same because even the slightest movements and vibrations can spoil the chips. Putting a restroom in the wrong spot might spoil a batch when someone flushes the toilet.

The thing is, pretty much everyone relevant assumes it is a bubble and that eventually large players will end up facing mob justice. That's why the hundreds of billions of $ IOUs are getting passed around like hot potatoes, and that's (in addition to ASML, the key part of anything EUV lithography, being booked out for years) why no one is planning to construct dozens of billions of dollars worth of fabs.

In addition, the know-how is concentrated in Taiwan. You literally can't train enough people in enough time to move everything out of there.

> concentrated in Taiwan

Where are SK Hynix and Samsung located again? Or 95% of Micron's facilities?

So yea, Samsung built a chip factory pretty close to where I live. Number one it is forking gigantic. You don't just slap one of these babies down. Next, the equipment that goes inside of that massive clean room building is a problem in itself. That takes years to get ordered, then years to ensure it works right, with employees that have a very particular skill set.

Again, people might want part of it, but they are also a bit smarter than you are and read history books to see exactly how this is going to play out and then they gladly walk away before they light their money on fire.

Absolutely. "If this isn't a bubble and demand persists" is doing a lot of heavy lifting there, though, and it may be some time before manufacturers decide that's the case.

I've made this same argument so let me make a counter-argument:

There are some ways to get this off the ground much quicker. One or more companies could buy an existing non-leading-edge fab like GlobalFoundaries. That buys a lot of expertise so you're not starting from zero.

DRAM also benefits from being very regular and relatively simple. It used to be what you bring up on a new process node to help prove things out.

It also isn't impossible to reduce reliance on ASML if you're willing to throw money at it. That's definitely a super-long-game move but it could be done.

I'm not going to argue that someone is going to do any of this but if demand is sustained it is possible.

It does help, but I have to wonder how many people are still working at glofo currently who are researching node shrinks. They stopped their research into the 7nm process in 2018 and all the indications are that they aren't really continuing it.

Meanwhile, I believe SOTA is at least 3 or 4 node shrinks beyond that 7nm process. It'll take years for them to catch up to where micron is currently.

That's fair. I had hoped GF was not entirely giving up but was going to wait until the tech for new nodes was proven and cheaper to obtain. It seems that is not really possible at the moment because anyone who is involved in that supply chain doesn't maintain the ability to stand up new processes on older nodes. Way too busy standing up new nodes.

So much for efficient free markets.

>a lot of companies that have billions in cash

They sit on billions because they avoid spending their money as much as possible.

The amount they spend on RAM in surrounding few years would represent almost nothing to the massive money hole that would happen if they tried to make their own fab.

Also, these problems tend to affect the entire market, which means if you're big, you're fine. It's when problems don't affect your competitors but affect you that the real issues for these companies crop up.

Real life is not SimCity, you can't just plonk more RAM factories like that. It takes an ungodly amount of capital investment, many years before you see a cent in return, plus there's only a couple firms worldwide that can do it in the first place.

Chinese DRAM production is already getting ready to ramp up.

https://www.tomshardware.com/pc-components/ram/hp-reportedly...

It might take a while but Chinese companies will figure out/steal/innovate into the right IP for different memory types and will cut out the American patents, ASML and East Asian fabs middle man.

I can’t wait for times when I can afford chips from less than 8 years ago.

The only thing that can actually introduce competition in RAM is some form of government backing around national security concerns. China has been doing this for some time though so there will probably be major Chinese supply coming in the medium term.

The problem is every twenty years or so DRAM makers get burned by building for demand that mostly disappears overnight. They've been through it enough times that they're going to be really reluctant to build new fabs. They'll certainly put some effort into getting the absolute most out of their existing installations, but I would be surprised if you see a lot of new fabs until they decide the demand is durable.

If we assume a spherical cow!

It will the Chinese, this will be the last big payday for the three company cartel and in time I think Apple will move in-house and design around.

No, it's the time, effort, and capital necessary to build cutting edge semiconductor fabs. Measured in tens of billions of dollars and decades.

Less APP, more LLM

What?

E=MC^2+AI

Reply on news.ycombinator.com