1. During the pandemic, consumer durables went from ~10-11% of consumption to ~13%. 3 percentage points of consumption is about half a trillion dollars annually. Many of these durables come with cheap microcontrollers or even high end SoCs now. New vehicles have dozens of cheap microcontrollers. https://fred.stlouisfed.org/graph/fredgraph.png?g=Ktkq

2. If you look at that chart, you will see that the decade previous to the pandemic was the worst ever for durable goods demand. Supply chains had adjusted to that.

3. Supply chains had also all gone to a JIT model to keep inventories lower. This was a huge source of efficiencies, but made them vulnerable to a demand shock.

4. There are new sources of chip demand in EVs.

5. There are new sources of chip demand for very cheap ARM and RISC-V chips. High volume, low margin stuff that has been underinvested for manufacturing, like the entire auto chip chain.

6. 2019 was a down cycle, and companies were idling capacity.

7. When the pandemic hit, companies projected lower demand, and idled more capacity. This put them in a huge hole from which they still have not extricated themselves.

8. COVID outbreaks in Asian factories complicate things

9. Transportation bottlenecks complicate things

Also, to answer your other specific questions…

1. Yes, everyone cancelled orders and idled capacity. Everyone was anticipating a deep and long recession until the fiscal/monetary bomb of March/April 2020. See #7 above.

2. The 120nm legacy nodes and similar are right at the center of the whole shortage. This has been underinvested for years because it has been low demand growth and low margin. Automakers are particularly vulnerable, because for years, every new feature has gotten its own cheap microcontroller to run it. There are now dozens in new cars.

3. Graphics cards have been affected by crypto miners sucking up supply at any cost, and everyone not named “Apple” competing for TSM’s top end processes.

4. iPhones don't just come with Apple A-Series chips made on TSM’s top-end node, but a host of others, some made on legacy processes where the shortage is. Likely we are talking about some cheap analog power chips or something like that. Apple has decided to push as much supply chain problems to iPad as they can.

5. For the vast majority of what automakers want, 7nm is overkill. Once there are more high end SoCs running multiple systems including self/assisted-driving, that will come into play. But that is not where the shortage is concentrated.

6. Apple was Intel’s 4th biggest customer before they left, so it has had a small impact, but only at TSM at their top node. Samsung has picked up Qualcomm and AMD as a result.

Crazy times in the semi world.

Disclosure: I work for GM, what follows is solely my own opinion and perception of reality.

On #1 also bear in mind that the supply chain is deep and complicated. OEM (GM, Toyota, Tesla, Ford etc) orders a number of COMPONENT X that contains a chip from a Tier 1 supplier. Tier 1 supplier order subcomponents from other suppliers. (Continue this until you have built a very confusing interdependent supply web).

At any level of this, anyone could have (and did) cancel their orders in 2020. As in OEMs cancelled some orders, Tier 1 suppliers cancelled some orders, Tier X suppliers cancelled some orders, etc.

Also known as the “bullwhip effect”:

https://en.m.wikipedia.org/wiki/Bullwhip_effect

And: https://en.wikipedia.org/wiki/Beer_distribution_game

I was very disappointed that the Beer distribution game, as described in the Wikipedia article anyway, is not a drinking game.

Not with that attitude it isn't!

How else do we simulate shrinkage in the supply chain?

> Continue this until you have built a very confusing interdependent supply web

I like to say that reality cannot be reduced to a single variable. It is a complex multivariate problem.

Most reporting out there on any subject reduces that being reported to a single variable. Example: Container shortage > bad > supply chain crunch. Etc.

The overwhelming majority of people have no practical understanding of business and, in particular, manufacturing. The complexities involved --the variables-- do not exist in their realities. What you explain about the supply chain is a revelation to almost everyone reading it. Even with this, reality is far more complex than what any of us can explain in a couple of paragraphs.

I have lived through a couple serious supply chain issues in the last several decades. This one is a bit different because almost everything in technology products is affected. It's one thing to re-spin a design because one chip becomes 60-week unobtainum, quite another when you can't get 75% of the BOM for 40 weeks.

> Example: Container shortage > bad > supply chain crunch. Etc.

One thing I learned working in manufacturing is to think about movement and management of empty carriers. If you fill something, it will be empty at some point and it doesn't just disappear. You have to get it back to where you fill it again.

> 2. The 120nm legacy nodes and similar are right at the center of the whole shortage. This has been underinvested for years because it has been low demand growth and low margin. Automakers are particularly vulnerable, because for years, every new feature has gotten its own cheap microcontroller to run it. There are now dozens in new cars.

> 5. For the vast majority of what automakers want, 7nm is overkill. Once there are more high end SoCs running multiple systems including self/assisted-driving, that will come into play. But that is not where the shortage is concentrated.

There was a discussion about this a few months ago. It's not just that newer nodes are overkill, it's that auto components apparently have very long and expensive validation requirements, which would be another explanation for why they are still depending on these older manufacturing processes.

https://news.ycombinator.com/item?id=28728965

Tesla, for the most part, is skipping that validation and just shipping cars. The warranty issues from Tesla have been notorious, but their business model is to ship cars as fast as possible and deal with warranty issues later. They were able to side-step much of the chip supply issues by either shipping cars without certain features (no adjustable seats, no USB-C ports, etc.) and also switching to newer 7nm+ parts, like the AMD Ryzen x86 chips that they've started shipping in the past few weeks. Their direct sales model allows them to easily change the price on the website, and so far, they still have unbelievable demand.

I think some of that is the difference in customer base.

Tesla purchasers, are, to some degree, buying into the idea of the brand. This works if your brand is nice and new and flashy, but I think that traditional automakers would probably suffer reputational damage if they tried this approach, and it is very hard for them to shake old reputations if they manage to get one of unreliability.

Also, traditional automakers have full lines (sometimes multiple) of different models. Each with its own refresh cadence. And each which then needs to be supported long into the future. (Yes, there are some common platforms)

I'd say Tesla ignored that future debt, but I think it's more nuanced. I'd hazard they made a wilfull decision that the relative simplicity of electric cars will allow them to avoid a maintenance nightmare down the road by simply replacing whole components (a la Apple).

We'll see how it turns out.

> It's not just that newer nodes are overkill, it's that auto components apparently have very long and expensive validation requirements, which would be another explanation for why they are still depending on these older manufacturing processes.

Also, if you're on a 7nm process, odds are there's not a decent way to e.g. drive the transistors for your power window actuator. Samsung's 7nm process has an 3 sigma intrinsic breakdown voltage around 4.8V, which precludes any kind of 5V tolerance and driving 2.5Vt power FETs.

Little 3V3, 5V tolerant microcontrollers are good for a lot of things, and they belong on bigger process. While 10nm might be able to do it, there's not really transistors for it in most libraries.

I know it's a nitpick, but I've never heard of 120nm... lots of 130 some 110 (half-node) and lots of 90nm.

Most of the 8inch wafer supply (mostly >90nm) was no where near capacity before the pandemic and supply restrictions. China did buy up a lot of it (actually moving equipment) and perhaps those were closed down?

Much of the legacy 12inch supply was fairly constrained even before the pandemic (especially specialty 55 and 40nm) due to shifts in display driver needs for OLED. The relatively new 10 & 7nm were pretty well dedicated to known customers.

Much of Automotive is in the larger 180nm to 90nm range with some 28nm for SOC due to cost, reliability, and legacy test qualifications. Except for entertainment only Tesla really uses much smaller than 22nm. It's interesting that they seem to have been willing to use commodity processors much more, and that seems to have made their supply chain more resilient (at the prices they were willing to pay).

It's interesting because a lot of mixed-signal analog really doesn't shrink well 28nm, and requires a lot of redesign to be cheaper than 180nm. There are ways to get down to 14nm and below, but the design costs require huge economies of scale to pay off.

The last couple of years in Semi really has been a game of how much are you willing to pay for the old backlog of parts that were hard to sell earlier. Only now are some new designs really making it out a year later. It's supply chain, but it's also price gouging and just taking profits while you can (why build more when margins will only plummet?). I do worry about the risks of concentration of this production in China.

Yeah, what I really meant by "120 and similar” was >14 and mostly >28 is where the problem is, especially anything with 3 digits.

What does this mean? "Apple has decided to push as much supply chain problems to iPad as they can." Starve iPad so everything else can thrive?

Well, all the devices share a lot of parts. Say there’s a particular power module that goes in iPhone, MacBook and iPad, and that’s the holdup. The choice is to put as many as they can get into iPhone and Mac, and let iPad be delayed. Just checked US Apple Store iPhone 13: Tomorrow M1 MacBook Air: Tomorrow iPad Air: 2 weeks

Indeed. I think iPads are their lowest margin product.

Honestly, the worst part is we find the short supplies are in these stupid, low-end, cents-a-pop chips that have nothing to do with high end processes. Especially in PSU applications.

Question about your 5. 7nm is overkill in performance but... isn't it better if they get more chips per wafer?

Cost is much higher than for the commoditized chips they use. Instead of developing high end SoCs that could run multiple systems, they went for these cheap microcontrollers that controlled a single function or just a few.

To give you an idea of what we are talking about, here’s Infineon’s product page. They are a big automaker supplier, and right in the middle of all this. https://www.infineon.com/cms/en/product/

Not if you want automobile rated durable chips with higher withstanding voltages etc. On many MCUs, the CPU itself is a minority with a large part being 5V tolerant IO etc.

I would add: Many chips were becoming difficult to source before the pandemic hit. The pandemic-related demand and supply shock only made it worse.

Source: Spent many weeks fighting supply chain issues and leading redesigns in 2018-2019 due to the beginning of the chip shortage.

I know the shortage is associated with the pandemic because that's when most people started hearing about it, but it was a growing problem even before COVID came along.

Curious, what were you trying to source, and in what volumes?

Even common parts like certain voltage regulators.

Volumes were 1-10 million annually.

Interesting. Some complete speculation about what you were mixed up in…

Demand in first half 2018 was super hot, and depleted inventories. Wholesalers and retailers built up inventories of finished goods in anticipation of The Greatest Christmas Ever. Manufacturers did not, and kept inventories thin.

But demand fell off a cliff in the back half of 2018, and the The Greatest Christmas Ever was a dud. 2019 saw canceled orders and capacity shutting down quickly in the first half of the year. By the holiday season 2019, I was driving around the ports of LA and Long Beach seeing containers stacked high in warehouse lots, and overflowing car lots, some with 2018 model year cars.

Anyway, 1-10m puts you in a weird pocket when things were moving so fast like that. In 2018, you are too big to fill the order, because bigger players are sucking all the air out of the room. In 2019, you are too small for them to turn capacity back on for.

I’ve been following this stuff since the 1980s, and 2018-now is the weirdest semi market ever, with two sharp upcycles and a sharp downcycle sandwiched in between, all in the space of 4 years.

> 3. Supply chains had also all gone to a JIT model to keep inventories lower. This was a huge source of efficiencies, but made them vulnerable to a demand shock.

I think it's worth emphasizing this - all of the other issues you describe are correct, but what's allowing individual disruptions to turn into a global crisis is a lack of buffering between components of the individual systems. This is a known issue with JIT, and in some ways it's factored into the models - assuming one bad year in ten, 9 very good years can still be more profitable than 10 pretty good years - but it widens the possible set of outcomes and allows for these kinds of cascading failures, which makes it harder for one institution to model or plan around.

It's also the sort of thing where if you're the first one using it, the costs of it are not nearly as large as when the whole economy is using it. Your buffer is, in effect, being stored at other people's factories. This is logically predictable, but was not in fact widely appreciated before the current problems, perhaps because this is the first shock to come after JIT was almost universally used.

Yeah, I think this is in part an outcropping of the Friedman/Chicago School notion of firms as purely self-interested - the notion that one can ignore the actions of the other actors in the space when planning one's own strategy is sort of silly on the face of it.

8b. Flooding in Malaysia where the factories are located. Some may attribute this to ordinary flooding, others may attribute it to global warming.

https://techwireasia.com/2021/12/malaysian-floods-devastate-...

Can confirm that this has had direct effect on car manufacturing in Asia.

Maybe its a stretch but I was doing research into North Korean hacks from APT38 and noticed that TSMC had been hit by WannaCry in mid/late 2018 shutting down some of their chip manufacturing facilities. Would it be possible to believe this snowballed into what we have today with a series of bad situations?

Re (1): It looks like (consumer durables / consumer expenditures) returned to the level it was at prior to the late-2000s recession. Do you have any explanations for what happened from 2010 to 2020? Seems like 12-15% is a more "normal" level than the 10-11% we saw during the 2010s.

Very good question and long discussion. But the short of it is that people favored experiences over things with their spare income. A trip to Disneyland instead of a new TV. Now Disney seems too dangerous for many, so they buy more TVs. This is the most dramatic shift in consumer preferences ever.

do you have a link to a source or more discussion on this?

This is a solid article, a bit old and less economically authoritative, but underscores what everyone has been seeing if you watch the economy (economic data lags, the best place to look is usually quarterly earnings reports and commentary i.e. transcripts) https://www.nytimes.com/2020/05/20/business/public-gathering...

> 4. There are new sources of chip demand in EVs.

I think there are also new sources of chip demands in other industries, such as perhaps companies utilizing ML having their own chip designs. But I'm not sure how much that plays into the issue.

Several factors:

1. There was real and relatively sudden increase in demand:

a. COVID-driven in part, all students and many additional employees had to buy computers to study/work from home,

b. a., driven by video, lead to increased demand on Internet & telecom infrastructure.

c. Electric vehicles consume an order of magnitude more chips than ICE cars. EV demand is also skyrocketing.

2. There were fabs (semiconductor factories) shut down due to COVID. This one in Malaysia [0] makes power chips and was particularly problematic for ICE auto makers.

3. The first indication of problems manifested in chip lead times pushing out. As buyers started to see more and more of this, they started increasing orders and soon adopted a hoarding mentality, similar to why many of us couldn't find toilet paper in stores early in the pandemic.

Result: As a personal anecdote, a chip that sells for $0.47 from that fab in Malaysia recently sold for $75 on the broker market.

Note that no, the state of the art fabs cannot be re-outfitted to make legacy chips. The old equipment they were developed on is generally no longer available and the profit from making those inexpensive chips would not justify the capital spend on new equipment.

[0] https://www.reuters.com/article/infineon-results-malaysia/in...

Regarding #3: We had an FPGA that is normally in stock or at most a few weeks lead time push out to a 56 weeks lead time. (Side note: This is for machine used to in chip manufacturing... uhm, oops.)

What happened then is that the vendor prioritized larger more expensive parts. They use more silicon per part but the over all profit per wafer is higher. It's not a terrible decision, it's possible to move designs to an over provisioned FPGA but usually not the other way. The lead time for higher end parts started coming down, not to normal, but less than a year. Meanwhile lead-times for the low end parts kept getting longer.

This scenario is very reminiscent of the leadup to the Innovator's Dilemma. The dinosaurs grow taller and fatter seeking and thriving on the juicy foliage higher up the tree, while the scrappy little mammals claw away at its soft underbelly.

Nobody's going to fabricate semiconductors in their garage. But couldn't someone make a modest profit without first investing $2.0e10, like Intel? Or is it just that the profitability scale is such that if you had an extra $2.0e7 to invest, you're better off buying Intel stock?

1) technically, someone is fabricating semiconductors in their garage, but as regards the world economy your point is still valid: https://hackaday.com/2021/06/29/garage-semiconductor-fab-get...

2) the equipment used in older fabs, is in many cases not even made any more. It's much like why the vinyl record manufacturing took a decade or more to ramp back up when the demand for vinyl records began going up instead of down. Nobody was making that equipment any more. The older fabs don't use the same equipment as the new fabs, except less of it; they use quite different equipment, and in some cases newer equipment actually wouldn't be able to make the older chips. I don't work in semiconductors, but I did from 1989 to 2004, and worked on several cases of trying to move old processes into newer fabs. It's not easy or quick, and in some cases you have to move the old fabs' equipment into the new fab.

Something is very weird with intel. Compared to fang pe around 30 or nvidia, amd around 100, intel has 10. Which combined with long term stable decently high profits is insane for an am tech company. Intel has been offered gigantic subsidies to build american fabs, and more recently just fabs period, but they arent, and at least one of these is approaching the point where they might get sued for not having even tried.

The answers I hear range from a deeply embarrassing and complete loss of competency on all levels, but primarily on the fab side, to an utterly incompetent and out of touch leadership. I favor the latter explanation as intel has been underperforming for nearly two decades now. My best guess is that intel has entered its cash cow phase, and will simply phase out product after product as they become obsolete, ending up as a half dead zombie barely supporting a few rare products in a few decades. Problem is, even if thats what is happening, their stock is surprisingly cheap. So it seems analysts seem to know something important I am completely missing.

Not to say they are the only company that fucked up, amd has been offered the same subsidies for building more fabs, and we arent talking 2020, we are talking for more than a decade, and they also havent for very unclear reasons. Amd also fucked up and completely missed machinelearning due to being unwilling to make a convenient interface or even just halfassed support for cuda. But they did that while nvidia went, ah fuck it, well let amd get every major console, and basically only make gpus with graphics ports as a side project.

Amd and nvidia is havent acted optimally and both have failed to exploit what seems like low hanging fruit, but they are highly valued in a way which matches market demand (which has been increasing quickly since 2014, not just 2020), and their improving technology. Intel however, does not even appear to have tried. Performance improvements have been incremental at best, and some "features" like the built in gpus are so terrible they mostly degrade performance.

> I favor the latter explanation as intel has been underperforming for nearly two decades now.

Remember "Only the paranoid survive"?

Well, people steeped in Intel took that to heart. So, when they retired or died, a lot of them had no heir apparent and their knowledge went with them.

No argument on the fab side, but that's easy to outsource. Samsung, TSMC, and Global Foundries will happily make Intel chips.

Alder lake is a good example, from what I can tell it's a market leading core, generally faster than the AMD competition, and a good basis to make future products. Future Intel GPUs seem likely to compete on price/perf, at least when any GPU from the last 3 generations seems to sell out instantly.

> Samsung, TSMC, and Global Foundries will happily make Intel chips.

And how long will it take to "port" those designs to another fab?

It's not a matter of "make install."

Yes, everything comes together with tools, but every one of those tools and its inputs is verging on full-custom and those designs were tweaked with the characteristics of those tools in mind.

The folks who know how to do that work are currently working on new products (either for Intel or someone else), so are you willing to slip those products?

Quite a bit faster than building a fab for a leading process and making a new product on that fab.

Alder lake seems plenty competitive (even market leading) for today, AMD has nothing big (like zen4) due anytime soon. Moving the next gen to some other fab in time to replace alder lake looks quite feasible.

Yes, Intel can do new designs for other fabs.

The question is when it makes sense to port old designs to other fabs.

FWIW, I'm not convinced that it takes less time to bring up a new design on a new fab than it takes to bring up an old design on a new fab. (The exception being a new fab that is identical to the old fab.)

I don't think that you can build a new semiconductor factory for $20 million (2e7), especially not when there is a chip shortage happening and you need factory equipment containing such chips.

There is the volume vs cost problem. You could put a fab together for under a million - but to do so your would trade labor for automation. A simple chip that a large fab can make for a profit (amortized) selling for 5 cents each would cost you more than $50,000 each to make. If you only need one chip that price is worth it, but most users of a chip are thinking a lot more and it typically turns out that getting a large fab to do your part is not only cheaper, but faster. Still you can do the highly manual process of making a chip if that is what you want to pay for.

It's a little weird to me. I see in the comments 50 cent chips going for 75 dollars. The old process chips _seem_ diy able - http://sam.zeloof.xyz/first-ic/ Obviously it would take me, like, a year or two to get anything to work. But I'd guess there are folks out there that could spin up far far faster.

I guess I don't understand why there isn't much, or any, labor intensive fabs combining on line, with a plan to automate away parts of the line as money comes in.

I suppose, the chemicals are so toxic, and the skills are so rare you can't get any production for months, even at the small scale. and perhaps even at the crazy high prices it's still not worth taking on the risk.

I know I don't know what I'm talking about. But golly it does _seem_ like I could make chips in a couple of years, and there's bound to be thousands of people out there that could do it in weeks. It seems like the kind of thing you could scale out with more labor and training, and improve process with automation as you go.

edit

Most trying to understand. Is it too toxic? are the skills so rare? Is the risk that high? Is that what's keeping low end chips from using a more expensive process even though they command higher prices - it's not enough to offset?

The $0.50 for $75 is a short term situation that will resolve once the fab's pipelines get filled again. We buy >1,000,000 per year and had no trouble getting them until the fab shut down. We'll be fine soon (looks like they've opened production again).

The primary costs of starting up a fab are 1. facilities, 2. equipment. Facility per square foot is likely the most expensive you'll find due to the appurtenant systems required for air flow and purity, a variety of gas handling systems, ridiculous electrical power systems, possibly gas plant systems, water purification and then treatment systems for both air and fluids exiting the plant. Some truly scary toxins are used and produced and have to be contained or captured and neutralized.

In terms of equipment, you'll need at least one of each of 40 types. Lowest cost is about $500,000. Highest is likely $25,000,000 (at the geometry you're talking about). Average in the area of $3,000,000 - $5,000,000. You'll need a cadre of engineers and scientists to keep it all running, and 2-3 years from breaking ground to producing product. If you did buy one of each tool type, you'd have most of them sitting around doing nothing most of the time, since the process times vary dramatically. In reality, you're looking at $500,000,000 minimum to build a fab that can produce $0.47 chips. Seriously, you don't want to do that.

Thank you. That makes a lot of sense.

I had some misconceptions around how versitile auto manufacturers are. I'd sort of assumed they'd be very good at turning a low volume, low quality process into a high volume, high quality process. I'd also sort of assumed they'd have a ton of resources on hand, like space and power, the and the ability to build a lot of the required infrastructure.

I suppose they are far far more specialized and reallocating resources like that is the kind of thing you'd have to do right away, make a big commitment, that's basically pointless.

They are experts in iron, and processes, but they still would have a lot to learn, and it is unlikely they could ever do better than TMSC who isn't stupid about manufacturing. Maybe they could find something, but so can the others.

The commitment just to play is very large, it probably isn't worth it. Even if you had proof of this would happen in 2015 and so enough time to build I doubt any would build s fab (though they would buy chips to store)

Yup. makes sense.

Wasn't thinking 7nm cutting edge, more along the lines of op amps or maybe a 70's era microcontroller - I had the impression lots of chips are used, and some are real fancy, but some are just old and, since they didn't order them, aren't being made.

I'd thought if someone was willing to roll back the clock 30 or 40 years - that tech might be cheaper and more realistic to bring in house. But yeah, there are already plenty of people that do that well, so it's really just an expensive boondoggle.

Definitely not something you can just put together in a few months.

Sure old process can be done,but the labor is high enough that you need to sell for more than 75. For simple stuff you still need well over 1000 each to make it worth it.. that is simple stuff. I know a few years back the government was paying around 60000 for Pentium level complexity chips (this was for a classified program so I doubt I was given full details)

State of the art fabs retail for US$10 Billion. TSMC, Samsung, Intel & China can afford to build them.

(for context, Freedom Tower cost US$2 Billion.)

I think in 2022 it’s more like 15B. 10 sounds very early-2020.

Perhaps you could do it at a 1000nm node, at very low volumes.

EBeam lithography will get you sub 10nm, but it's one at a time.

This is close to how they make the mask-sets, very slow.

Also, the Texas storms caused hard power cuts to multiple chip factories, erasing, in the best case, a full month of chip production as they slowly restart the factories [1]. These chips are, btw, generally built on such old tech that Intel is refusing to fab them [2]. So I would have thought a chip was a chip, but apparently car chips can only be sourced from a handful of specific fabs.

Car manufacturers are clearly going to start building huge inventories of them though:

> “Because of a 50-cent chip, we are unable to build a car that sells for $50,000,” said Murat Aksel, head of procurement for Volkswagen Group, during a press briefing in Munich last week. [2]

Ouch. Though honestly, this probably ought to result in some execs being fired... If you sell 100k cars a quarter, imagine telling the board that you saved the interest on $200k and the rent on a small room and that's why you're not selling a vehicle.

[1] https://www.reuters.com/business/autos-transportation/texan-...

[2] https://fortune.com/2021/09/17/chip-makers-carmakers-time-ge...

> Ouch. Though honestly, this probably ought to result in some execs being fired

It should, but it won't because everybody failed the same way.

It was just like the 2008 banking crisis. Lots of people saw it coming, but reasoned "If I predict doom and mistime it, I'm an outlier and get fired. If I simply keep my head down and cash out and it all goes pear-shaped, I'm simply one of the masses and nobody will notice me."

Can confirm. Anyone working in supply chain is now told to buy 1-4 year supply of as many of the critical chips needed as they can. It used to be all about how JIT could it be. When the whole chip supply chain shifts from JIT to “get all you can while you can” it causes the shortages. 2nd hand market is seeing astounding mark-ups.

If everyone buys 4 years of supplies, that exacerbates the problem. The supply chain bullwhip effect will be massive.

For the next 3 years, fabs will be swimming in profit. After that, I expect a major correction.

Everyone is suggested to place scheduled orders for the future. Nobody (that I’ve heard) is suggesting that you immediately purchase and carry four years of inventory. People in this market are generally very intelligent, and there have been a number of booms and busts in semiconductors in the last 30 years.

Will the fabs have the discipline to actually use the extra cash injection wisely?

They're the best positioned to do that, they're one of the last classic long tail, high capital investment portions of the industry. Each node is only competitive for a short while, but they have multiple nodes in the development pipeline because of how long it takes, and a misstep on a single node can devastate the whole company if you're not in the #1 spot originally.

So the answer is we don't know it, depends on the management currently at the fab, but they have the right external incentive structure to handle it well.

Does that mean most companies will be using 2020 generation chips until 2024?

Sounds they'll dump it in trash eventually and it'll be a terrible waste of resources

Companies in my industry have products they sell for 10-20 years. The chips will get used. Product updates often keep the same chip, and the same chip is used across multiple product lines.

EOL for these chips usually means a bunch of firmware engineers in sustaining need to scratch their head and read 10-15 year old documentation to find a viable replacement.

Where I worked that I participated in the board design process, we'd get "lifetime buys" of chips on a many year cycle anyway for design stability reasons. Who cares if the microcontroller is using a couple more watts than it could be from aggressively respinning if it's sitting next to a 500w motor?

For the vast majority of things I buy that use chips (obviously not consoles, phones and computers) they could easily be using chips from 2010 and the only difference might be a slightly higher electricity draw.

Chips are performing an astounding number of tasks these days. Microwaves, vacuum cleaners, toasters, etc. None of these applications require advanced chips and could settle for something from 20 years ago.

> c. Electric vehicles consume an order of magnitude more chips than ICE cars. EV demand is also skyrocketing

Source? My impression is just the opposite. Batteries and an electric motor is a relatively simple device compared to a gas motor. So you replace a fuel pump (and related sensors), cat converter (and sensors), transmission (and sensors), and a hugely complex (1000s of parts) engine (and sensors). Managing maximum efficiency, torque, and HP across the entire RPM range while minimizing emissions is quite an intensive process. Not to mention the interaction between the engine and the (compute controlled) transmission. Especially these days when the simple rules are gone and managing turbos, heat management, etc means that you might give maximum HP, but only for 5-10 seconds or even idling unneeded cyclinders, stopping the engine at redlights, etc.

Tesla in particular as moves many distributed chips into a highly integrated single board, various people (like Monroe on Youtube) claim the Tesla model3 has a market leading level of integration replacing dozens of distributed CPUs with a custom board/ASIC.

I used to work in EVs (and tangentially still do). They really only require some power electronics - gate drivers and IGBTs - in addition to the normal bunch of chips all over the vehicle. So even if demand was "skyrocketing" rather than being a tiny fraction of the market, they shouldn't impact the demand for chips on regular cars.

EVs require 2.3x the chips of ICE. (not orders of magnitude, apologies for the hyperbole)

https://www.idtechex.com/en/research-article/ev-power-electr...

Interesting, and weird. Although that's measured by $, not per chip, per per mm^2 of silicon. I suspect electric cars (being newer designs) tend to use much newer chips, where the traditional ICE manufacturers are often using many generation old processes and paying rock bottom prices. So during shortages the cheap chips don't get made as fabs migrate production to the most profitable/newer chips.

Based on the tear downs I've seen I believe the 2000+ chips in the model S/X, but would be surprised if a model 3/y even had as many as 1000 (the number in a ICE vehicle). The model S/X were crazy complex, CPUs for each door lock, each window motor, each windshield motor, etc.

As a result VW, Ford, GM, Chevy, etc have to halt production lines, while Tesla has a record year.

> The model S/X were crazy complex, CPUs for each door lock, each window motor, each windshield motor, etc.

A lot of ICE cars have similar complexity in these systems. It's much easier to put a module in each component and just send messages on the bus to control them. Most newer cars also detect objects in the path of the window and will stop/reverse the window to prevent pinching.

While EVs might have more $ in chips, I also suspect they have a lower chip count as more indicators and interactions are handled on screen. Meanwhile, most ICE vehicles still have analog gauges, analog gear shifters, physical gear selection illuminators, etc. and all of these need chips.

I compared a 2019 Model 3 to a BMW i3, Chevy Bolt, and a Nissan Leaf (as well as a fair number of ICE cars like subarus, bmw 335, VW Golf R). I did notice that the entertainment system/nav/music, etc where worlds apart. The Tesla felt like a current smartphone, pinch/zoom, fast UI, easy to use, and pretty good speech recognition. The others felt like 5-10 year old phones, similar in feeling to even my ancient nokia 770 with Maemo mapper.

Monroe (on youtube) did a breakdown and said that the only other place he say CPU/ASIC/Sensor integration like on the model 3 was in satellites. Apparently there's a large CPU with a ton of pins that handles quite a bit of cars needs. Even the usually Bosche supplied ABS/Stability management/traction control functionality has been merged into the Tesla chip.

> Even the usually Bosche supplied ABS/Stability management/traction control functionality has been merged into the Tesla chip.

That's interesting... that's usually it's own ECU on most cars.

I can see why they might do it this way. ABS/SC in ICE cars tend to focus mostly on the braking system. Since most ICE cars are two-wheel drive with an open differential, there's not much that can be done from a stability point of view besides cutting power. Even AWD cars with computer-controlled differentials have response times that are too long to really be of much use in an emergency situation.

EVs, OTOH, can apply or reduce torque to the wheels much faster than even the brakes can respond. And being able to not only apply stopping force to a wheel, but also apply torque to another wheel makes for a totally different approach to stability management.

Here's an interesting anecdote: Tesla cars have traction control that monitors the differential in rotation speed among the four tires. If you put new tires on the rear and leave worn tires on the front, the rotation differential between the inside rear and outside front on a tight curve is enough to fool the traction control into thinking the worn front tire is slipping and will pulse the brakes on that wheel only. It's pretty unnerving.

Put new tires on the front and the problem is solved.

This happens on my BMW too (2011 3 series). I’ve also had traction/stability control activate when hitting a large bump due to the slight change in wheel speed as each wheel goes over it.

I suspect it's more the case that electric vehicles ship with more "smart" features (touchscreen consoles, mobile phone connectivity, etc.) which in turn leads to larger chip requirements.

Dunno, I don't think so.

Streaming music, maps for nav, top down views that integrate numerous camera/utlrasound etc sensors, onstar (and similar), traffic updates for maps, bluetooth connections/app for phone, etc seem common on pretty much all new cars. At least they were in the subaru, honda, bmw, and VW cars I was test driving. A center console mounted screen for entertainment/maps/music is becoming very common on pretty much all new cars I looked at.

Also 4: There's been underinvestment in the sector for years now. Lots of money got plowed into stuff like social media, crypto/blockchain & ML but semiconductors just weren't sexy to the investors.

I don’t really buy this idea that chip fabs were being starved of money because of blockchain. Do you have any evidence to back your claim up?

It's not so much that chip fabs were starved by blockchain as much as the fact that many investors generally chase high short-term low-risk yields, which primarily results in major investments being in software-based ventures. Hardware can yield quite large gains too, but the upfront risk is much higher and. the growth rate is slow (if at all nonzero) for a long time before actually growing.

The chip fabs of Intel, Samsung, TSMC, and a couple other companies were not being starved, no. But, nobody was starting new semiconductor fab companies like they did in the early days of Silicon Valley (hence the "silicon" part of the name). That hasn't been happening for a couple decades now.

> The chip fabs of Intel... were not being starved,

Maybe, but consider that Intel stock was pretty flat for a long time in the 2000s/2010s especially compared to the amount of money that was flowing into the FAANG stocks and other social media companies during that time. If Intel's stock had rising at something like those rates they probably would have been able to invest quite a bit more into fabs.

Companies don't invest off their stock price, they invest off their balance sheet, or by borrowing money. All these companies have plenty of cash at hand and can borrow at historically low rates.

Venture capital investment is flashy and gets the press, but there's lots of other money that wants more predicable returns. There was around $600B of VC invested in 2021, but Intel alone has $23B cash on hand, and US pension funds alone (which traditionally invest in safe sectors) has over $35T in management.

This isn't true.

Most semiconductor companies are public companies, with deep pockets and plenty of cash on hand. They invest based on forecast demand.

> Note that no, the state of the art fabs cannot be re-outfitted to make legacy chips. The old equipment they were developed on is generally no longer available and the profit from making those inexpensive chips would not justify the capital spend on new equipment.

Interesting. I had assumed that the difference between state of the art fabs and older fabs was that state of the art fabs can make smaller things, and perhaps more geometrically complicated things, but that they don't have to. If they wanted to make a design that only uses bigger things and simpler geometry they could.

Where does this assumption break down?

State of the art machines are used to make state of the art high-value chips. FPGAs, GPUs, CPUs, RAM, SRAM

As those chips run through their life-cycle, their value drops. After 15 years, the price the chips can sell for will drop 90% or more.

If there is a spike in demand at that point, there is no ROI on purchasing brand new equipment.

The demand spike is seen as temporary. Hence, the chip makers are telling the buyers to design out the old chips and upgrade to new. Many companies can do this. But some are constrained by industry regulation (aerospace) or industry practice (semiconductors).

I've mentioned before the irony that companies that make semiconductor equipment are struggling to ship more tools because they can't get enough (of certain chips) to complete building them.

While this is mostly correct, it is worth noting that FPGAs are mostly 28nm and 20nm processes, and have just started shipping on 16nm[1]. That doesn't require state of the art. The FPGA market is comparatively small though so doesn't affect the market dynamics much.

[1] https://www.xilinx.com/about/generation-ahead-16nm.html

>Interesting. I had assumed that the difference between state of the art fabs and older fabs was that state of the art fabs can make smaller things, and perhaps more geometrically complicated things, but that they don't have to. If they wanted to make a design that only uses bigger things and simpler geometry they could.

>Where does this assumption break down?

Why would they want to? There is heavy and overbooked demand for the 'smaller and complicated things' that make more money, and add the extra cost of the design to make the 'bigger and smaller' things, and it's a no brainer.

They wouldn't want to. What I'm curious about is if they can without retooling or purchasing equipment. Say due to chip shortages affecting the US military the government invoked the Defense Production Act to require some company with only state of the art fabs to make an older chip.

From what I've read, the way a layer on a chip is made is a coating is deposited on the surface that is sensitive to light at certain wavelengths. They then use light of wavelengths that the coating is sensitive to in order to transfer a pattern from a photomask to the surface, making some change to the coating at those places where the light is not blocked by the photomask. Then they can treat the surface to remove the coating at those places were it was changed by the light.

They then can deposit thing onto or remove things from those parts of the surface not still protected by the coating. Then they can use something that can remove the coating. The result is a layer on the chip with things deposited or removed right where they wanted them to, as determined by the pattern on the photomask.

Repeat for more layers.

The major improvements from generation to generation are the ability to use shorter and shorter wavelengths of light for the photolithography which allows for transferring patterns that have smaller features, and maybe the ability to make thinner layers so you can make chips with more layers.

I don't see anything in the photolithography process that requires using small feature size. If for some reason you wanted to build an older chip that only used much larger features, I'd expect that it would just be a matter of using the photomasks for that older chip, and setting the layer thickness to what the old design called for.

Only the most critical layers use the most advanced litho tools. The other layers use n-1 and n-2 litho tools and print larger geometries. You are correct that it is not a problem to print larger geometries.

However, there is more to it than just litho. Many thin films, implants, etching, polishing. A product is designed to a specific fabrication process flow. The real showstopper is that the process flow guarantees the electrical parameters of the transistors and passive devices will be with a certain range -- the product circuits are designed based on these parameters' values. A different process flow has all different electrical characteristics and the product won't work without redesign.

I think the deal breaker is "without retooling".

For one thing there's a lot of steps involved in photomask -> repeat. If your masks are all built for a smaller wafer size, or a different focal length setup or physical process diameter, then they literally won't fit in the die, or won't focus correctly.

You'll run into similar issues all up and down the chain. If your doping and prep all assumes a smaller wafer or different geometry, you're out of luck either physically because things don't fit, or because your chemistry needs to be redone. Your various deposition, bake and cure steps will necessarily also assume a process geometry, because some of the processes diffuse out from the center, or at different angles relative to some source that's bombarding them.

It's sort of akin to saying "well if we can build a vw bug they can build a model-T". Generations of equipment changes make things incompatible.

As someone with no experience in this area, I think you're ignoring the option of making the equivalent parts on the newer equipment.

The design costs would be higher (based on what I read, the new methods are more complex and costly), but if you're legally required to do so ... this would possibly be faster and cheaper then redesigning the factory. Larger wafers, larger yields, etc ... once working they could probably meet the required output quickly(month long delay?), provided the demand / legal punishments are great enough to justify the costs.

CMOS devices scale in all 3 dimensions, including gate oxide.

It is quite unlike, say, PCB fabrication, which is way more flexible with regards to layer stack-up.

Good list. Also, Taiwan drought and Texas freezing impacted some fabs last year.

And the fire in Ajinomoto's factory that makes Ajinomoto Buildup Film, used in high-end chips, briefly cut production of those.

I vaguely recall a shortage of epoxy resins as well. (For packaging chips, and for printed-circuit boards). It seems like more went wrong than usually does in 2020-21, even leaving aside the pandemic.

Non EVs also need a lot of silicon for infotainment, driver assist, etc. it’s not just EVs demand that is skyrocketing. All cars are.

True, but I believe there is also a missing factor: When people wasn't able to get outta home, demand for cars plummeted. And whoever made ICs as for cars just had the opportunity to fill the gap from automakers with more lucrative parts.

One thing I heard from a guy working on chip for a big German car company was that they used to have a decades old contract with a major Taiwanese chip manufacturer which had extremely good terms for the Germans. In particular with terms to delivery times, and fees for late deliveries, and price. This deal had apparently been used as a framework for the purchase of not just the then used chips, but pretty much all latter ones too. In retrospect the deal seems to have been at the very least partially politically motivated by Taiwan, but the chip manufacturer could not cancel the framework agreement without significant penalties. There was however an option for the car manufacturer to cancel quite freely, and when the news of the pandemic hit, they did. It seems like no one in the department quite understood just how extremely good the conditions had become over the decades, and after they canceled it, and decided they wanted the same deal back... Well lets just say they are at the back of the line despite paying substantially more than they used too.

Ironically enough demand for cars went up during COVID. I can't tell you why, but I know it's a fact. Lack of supply is certainly a factor but I think there's probably more going on. Go check how insane the market is right now. It's nearly impossible to buy a new car, and used cars are selling at way above the notional blue book value.

Demand did go up, after the market adjusted to projections that with everyone in lock-down, demand for cars would go down. Demand for in home entertainment was projected to go up. Auto makers sold their capacity to people who were projected to use it in lock-down friendly ways. Once reality set in it was to late to get the capacity back.

We sold two cars and bought one new one. Remote work and a move to a rural area meant we had more cars than we needed, and that we needed to replace our city cars with something more rugged/utilitarian.

I became 3400 dollars richer buy selling few old unused stm32 reels

I'm sure there are a lot of significant factors, especially for different types of chip, but also pretty confident the shortages related to GPUs and similar are due in no small part to the massive buildouts of digital Chuck E. Cheese token factories, e.g.:

https://twitter.com/wired/status/1335445557788233728

https://twitter.com/BitcoinMagazine/status/14149943888208977...

https://twitter.com/dmgblockchain/status/1438547780646019086

AFAIK, the cost to add significant additional manufacturing capability for most products means that increased demand has to be forecast for a really long time to be worthwhile. Hence why it was still hard to find quality nitrile gloves, isopropyl alcohol, and Clorox wipes during 2020 even though everyone knew there was a bigger market for them for potentially quite awhile. I'm sure the chip and hardware manufacturers don't want to invest millions or billions in assembly line machines and then have buying levels drop back to pre-cryptocurrency-mania levels.

The market has decided that entertaining meatbags was worth less than evading taxes and stifling regulations.

It seems its invisible hand is a bit self-serving, go figure...

This does not follow at all; the GPUs were already iterating basically as fast as they can go - after all, why would it be, it's not sped up by putting more on the production line or more engineers on the design side.

NVIDIA at least have been trying anti-cryptomining measures in their firmware. A form of defensive innovation, I suppose.

The crypto advocates really do not want to hear it, but their proof-of-waste scheme has big negative externalities for other people.

>it's not sped up by putting more on the production line or more engineers on the design side

Of course it is.

If revenue fell by 90%, nvidia would spend much less on R&D and progress would be much slower.

If revenue doubled, nvidia would spend much more on R&D and progress would speed up.

If revenue fell to $0, nvidia would go out of business, spend $0 on R&D, and progress would halt.

There's a direct, although not linear, relationship. The more money there is, the faster progress is (as long as you trust nvidia to spend it wisely and execute)

I don't mine crypto, but I like fast tech, so I say let the money keep flooding in so we all benefit.

The demand from gaming & workstation & datacenter is absolutely huge, revenue would not fall without buttcoiners because the GPUs would still fly off the shelves.

In fact buttminers create a new risk for the vendors: whenever buttcoins seriously crash, the market will suddenly be flooded with used GPUs.

Proof-of-work is not the only workable consensus mechanism, as Ethereum moves from a PoW to a PoS model Soon(*) that should improve.

(*) {currentDate} + 18 months*

Highly doubt that this will reduce the amount of mining in the world. Eth miners will mostly just switch to a different shitcoin. As long as BTC is dominant, mining will keep wasting gigatons of power. And how exactly would anyone dethrone BTC? The first mover advantage is everything, and nothing else has it.

How? By not being massively expensive but still being equally secure (in addition to doing more than just moving ledger entries from one wallet to another) Crypto 'mining' is not a fundamental or necessary feature of the universe - it is a very strange, very short duration thing that will show up in the history books as one of those wildly irresponsible and completely ridiculous things people did for a little while until they figured out a better way to do it. It will show up on the same chapter as biplanes and trench warfare 100 years from now.

Even without covid, the change to EUV is a massive bottleneck. You see , there is but one company in the entire world who is making the EUV machines, ASML in the Netherlands. EUV was on the "edge" of roadmaps since, I dunno, mid-90s or something. ASML persevered and now reaping the rewards. EUV fabs need a ton of capital and investing this much without good institutional knowledge in chipmaking is too risky. And TSMC and ASML had a very good relationship for a very, very long time including TSMC buying 5% of ASML in 2012. So a very very long history culminated in TSMC being the sole top dog in the industry. Intel tried to make chips with parts similarly small without EUV and it took them an awfully long time to ramp it up -- it was supposed to ship in 2015 https://www.techspot.com/news/48577-intel-rd-envisions-10nm-... and the first desktop processors on this process shipped like two months ago, barely making 2021.

So whether automakers could use a 7nm fab is a theoretical question -- there is but one 7nm fab, the one at TSMC and they already sold every wafer they can start. They would be able to sell more if they could start more but there are only so many EUV machines ASML can make so there are only so many wafers TSMC can make. All the gaming consoles? yeah, that's 7nm chips. With AMD and Apple moving to the next line, there'll be some capacity freeing up but the demand is still very, very strong.

Also, another effect I do not see here is the container disruption. Read https://www.vox.com/recode/22832884/shipping-containers-amaz...

The big shortage seems to be in microcontrollers and power ICs, not the kind of stuff you'd want to fab on TSMC.

Putting an STM32 onto N7 would be nuts.

OP asked whether automakers could use N7. I answered. Also, in theory, if there would be ample N7 capacity, nuts or not nuts, there's no theoretical problem with it and you could get a real metric shit ton of STM32 out of a single wafer using N7.

There's also the confounding issue of the AKM factory fire. AKM made the majority of the DAC chips used for consumer electronics and automobiles, and their factory burned to the ground in 2020. Then another company that tried to step up production to fill the gap (somewhat mysteriously) also caught fire.

https://www.prosoundnetwork.com/business/akm-factory-fire-sh...

There are other companies that make similar chips, but in much smaller quantities, and they are really in no better position than AKM to fill the void.

A recent "update" : https://evertiq.com/design/51031

Older generation fabs for major manufacturers (Samsung, TMSC, others) were heavily impacted by staffing shortages due to covid/lockdown/etc. When staffing is low, the new fabs run 100% and the oldest ones just shutdown.

This caused a bunch of chips that are old and cheap to be complete unattainable, due to a lack of safety stock across the industry. Cars, power supplies, and other "large" things are highly dependent on these older chips, and have very long redesign timelines.

Newer generation is actually simpler, its just demand growth outstripping supply. I'm sure 2020/2021 also impacted the ability to bring more capacity online but probably only to a small extent, these are long term investments that aren't responsive to instantaneous demand.

This is correct. Shortages of new generation chips can generally be resolved by new capacity. We are seeing record levels of capacity spend by the chip companies.

However, older chips are difficult to increase capacity for. The tools are not available, and the profit from legacy chips from them is many times insufficient to make the business case for.

> Can automakers not use a 7nm fab?

Intel offered car manufacturers 16nm chips "as many as they wanted", but they'd have to be re-engineered from, from what I read, 150nm and certified for harsher conditions (temperature, humidity, acceleration, probably noxious fumes if installed under the hood) than consumer electronics, which adds to the cost.

In addition to that, with the upcoming switch to electric cars they probably wouldn't produce enough of them to recoup the investment.

according to Intel's CEO, chip demand in 2020/2021 was 25% higher than it was in 2019. The typical annual growth for the industry over the last thirty years is 4 or 5 percent a year.

This massive jump in demand has caused backlogs. Covid supply chain issues didn't help.

>Covid supply chain issues didn't help

It would be nice to get to the bottom of these issues I keep hearing about. Yeah, COVID was/is a big thing but when it comes to feet on the ground, what's the impact? Not this much surely.

Not to be conspiratorial but is this a cover story we're being given? Why does it feel like there's suddenly this massive void in productivity?

> what's the impact? Not this much surely.

Yes, this much. I saw someone explain this well enough on reddit, but I can't find a link so I'll try and paraphrase it.

Let's say you're a lumbermill and you run at 100 boards a day in normal times, with enough capacity to do 110 boards a day if you stretch. COVID shut you down for 3 months. You now have a 9000 board deficit from your normal production, with people lined up to buy the boards before you've produced them. Once you get started again, you crank up capacity to 110/day - it takes you 900 days - almost 2.5 years - to recover that original shutdown deficit, without accounting for any of the increased demand for your product.

This is roughly what other industries are experiencing. "Why don't you just make a new factory to increase your production?" well, it would take a long time to do that, would take a lot of capital, the shortages impact you too, and there's no guarantee that once COVID passes, demand won't go back down to 100 boards/day leaving you with a huge unnecessary capacity.

A simple example, of course, but this kind of thing is happening in some form or another in most industries, and will just take some time to work itself out.

Great example! Easy to understand, and clearly away from any "conspiratorial" answer.

Just to tack onto it...

> Once you get started again, you crank up capacity to 110/day - it takes you 900 days - almost 2.5 years - to recover that original shutdown deficit, without accounting for any of the increased demand for your product.

This is best-case scenario. Likely is that a few % of your workforce will have been out sick with covid (or even passed away), and a few % of your workforce may have quit to care for family, or quit for better pay elsewhere, etc. So your "work extra hard to hit 110/day" is more likely "work extra hard to stay at 100% or even 90%". Especially limiting if you need to buy parts for your output that may be delayed due to others who are behind.

For sure - my example is just a very basic one to show how easily most industries can slide into a hole that takes time to dig out of even under good conditions. The reality is even more complex.

In 2020 I was working for a consumer electronics company and went to some presentations on how our supply chain and operations teams were dealing with the pandemic. I'm far from an expert on this, but this is what I heard from people who were working on this.

First, you had a massive lockdown in China in January 2020, which shut factories temporarily. Then when those factories came back online, the production lines had to be reconfigured so that stations were 2 meters apart. At the same time, a lot of the collaboration that typically happens by people from a company flying out to China to meet with the supplier face-to-face were shifting to video, which frequently met that at least one side of the call was early in the morning or late at night. And as all of that was happening, there were also large shifts in demand as some products saw spikes in orders. That pattern then repeated itself across the world - a company might move production from China to, say, Indonesia, only to see the new factory have to temporarily shut down.

I can imagine that two years of this would create major backlogs and disruptions.

We have just-in-time inventory at all levels of the supply chain. China was shutting down highways between major cities and forbidding people to go out on the street in the earlier parts of the pandemic. I'm surprised the ripple effects haven't been worse.

To turn silicon crystals into a PS5 requires multiple steps in the supply chain which, when running smoothly, have multiple steps where people may be planning years in advance.

Sony builds their products out of dozens of components from different suppliers, which themselves may be made out of finished materials provided by other manufacturers, which in turn may be made out of chemicals or raw ingredients from other companies. The dependency tree is quite big, and if any of them are messed up, you can't finish your product.

Normally all of these dependencies are planned well in advance (quarters or years) so that everything is ready when the time comes to ship. Mess it up one level deep, and you're a quarter to a year behind. Mess it up two levels deep and you're two quarters to two years behind.

On top of this, semiconductor production was already stressed to its limits before COVID.

There has been 5,443,453 deaths from COVID-19 so far, with no end in sight. Just counting the deaths, that's a pretty large impact from something that just appeared two years ago. Not to mention the psychological impact the pandemic has on humans, which is even more than just counting the deaths. This is why productivity is falling off a cliff. Everything changed underneath our feet and people are starting to feel the long-term effects of having a virus lurking around every corner.

To add to this: Even if you're not killed by Covid, infections will usually result in you being quarantined for two weeks (speaking of 2020 numbers) which created HUGE problems with DHL for example. There were simply no drivers for all their trucks to get stuff moving around (in Germany e.g.). This had a huge impact on logistics, for example.

Accounted deaths are a fraction of actual. The Economist used a pretty good model to estimate 15.3mm excess deaths globally back around November

Perhaps… but has nothing to do with the chip shortage.

I'm sure no one in the chip industry died (or got sick). And no one in an industry that supplies chips. And no one in an industry that buys chips. And no one in a service industry that services the employees of the above industries.

/s.

Its global. Its everywhere. Every industry will have impact simply due to huge disruptions to everyone. Tons of people got died, and many more got sick, and subject to lockdowns, and saw their purchase patterns change (eg. toilet paper). Directly, the IC industry changed when people had to purchase laptops and monitors and webcams for their at-home employees. That is a huge surge in demand that was a 1-time thing, but all those office supply orders are still depressed, so anyone making things has to re-tool their factory.

How can you be so sure about that? COVID, either by first-order effects or by second-order effects, is affecting almost everything in the world right now, why not the chip shortage too?

> what's the impact? Not this much surely.

A few months ago there were seventy billion dollars of goods anchored off of LAs docks alone. Some of those were finished goods, but some are intermediate goods.

It's so bad that it's cheaper to make new shipping containers in China than spend the time to return them once they've been used once. We have a glut of practically new containers clogging up our infrastructure.

Even as COVID stopped causing the direct backups, backups propagate all on their own. Like a giant traffic jam and you get to the front and the accident was cleared hours ago.

There was a large spike in unemployment in the US/worldwide. Combining that with drastic shifts in consumption patterns, causing a ripple effect in delayed production changes. And perhaps also shortages are just getting more attention right now, simply because people are looking at them more?

A lot of it has been that COVID exposed how many industries were relying on Just In Time fulfillment so then when the demand spiked because people were home isolating the supply wasn't there. I recall a particular story about the toilet paper industry explaining that.

imagine a quickly moving but congested highway. anyone even tapping the breaks will cause a massive slowdown. I think the chip supply chain is similar.

On the contrary. It's a cover story being blown. The system was already operating at the edge of its capacity, and now it's toppling over.

My guess: It is lowered expectations, and people accept the excuse.

So, what would you tell your customers, if you can't deliver on time?

Same thing with Brexit really. An easy excuse to reach for.

Think about what happens if half of the factory personal is in home office or can only work at reduced capacity due to Covid concern => production capacity will be reduced. Then consider this happens transitively. Each supplier in the chain will work only at reduced capacity, and therefore the overall shortage can be even stronger.

Too many redundancies and buffers have been removed in the last 20 years chasing "just in time" and "lean" and other various ways of outsourcing your responsibilities for input materiel to improve (local) efficiency. The system became more fragile, and then a large shock occurred.

The good news is that we'll learn a lesson for the next 20+ years.

The bad news is that we'll forget it again.

There has been a colossal shift to home working and home entertainment. So many people have bought themselves better gear to do this with, especially if they got more stimulus cheque than they needed.

Meanwhile the cryptocurrency boom provided a means to turn graphics cards into virtual money, so those all got bought out.

All while staffing, mobility (can't fly troubleshooters out so easily!) and other aspects of working have got harder, and a related physical shipping crisis is causing an increase in inventory both at factories and in transit.

The drought in Taiwan is also a factor: https://www.taipeitimes.com/News/feat/archives/2021/04/22/20...

I thought fabs were mostly closed loop at this point, and occasionally just need a small sip of water after the initial fill.

The drought was a factor for a very short while, but it is completely over after rains that caused flooding a few months later.

> “TSMC has always maintained contingency plans for each stage of water restrictions... So far there’s no impact on production,” it said in a statement.

In addition to the answers you’ll receive about this specific shortage, it’s good to remember that chip shortages are common. Consumers aren’t usually this aware, but they happened commonly prior to Covid. So it’s worth asking, why did chip shortages happen historically, and how did the current situation affect those previous issues?

Linus Tech Tips, a reputed YouTuber made a video about this a while ago, where he explains it really well. If you're interested, check this out: https://www.youtube.com/watch?v=3A4yk-P5ukY

Basically demand is increasing by orders of magnitude, supply was barely keeping up but there was a symbiosis. Any disruption created a massive backlog.

Its just a bottlenecked queue, that then became closed off, which amplified the size of the queue so much faster.

Graphics cards and gaming consoles are heavily affected by availability of many other things than their headline processors. There's a ton of other parts in there, power management ICs, level shifters, muxes, switches, all sorts of tiny cheap parts that are out of stock. Suppose you're making a game console. You're missing one power management chip, and suddenly you can't produce any units unless you redesign your board for another part, assuming you can get it. It's enough for a single part in your bill of materials to be out of stock and you can't manufacture anything. This dependency means you will tend to buy as much as you can of all the parts you might need, draining the market of any excess capacity. Apple needs lots of chips other than their SoC - they have chips for power management, radios, digital microphones, sensors, antenna amplifiers and switches. Even if supply for their SoC is guaranteed for years, they can't build a single device unless they have every single part that device needs.

Automakers caused this. The issue is not just that they fucked up and ruined the market that one time in mid/late-2020. The issue is that that one fuckup destroyed the confidence in the market. So you can no longer rely on stock being there when you need it, meaning you need to buy as much as you will ever be likely to need, because you don't know when you'll be able to buy again. So you buy as much stock as you can afford to, and if you can't afford to you borrow money and buy it anyway. This is a big problem, because the signaling mechanism of the market is broken. Any stock that shows up now will be gone instantly, but the aggregate demand for parts hasn't really moved - people don't need significantly more parts, but they do need to be certain they'll have them when they need them, which is why they buy more than they need now. At some point you have enough stock and don't need to buy again for years. This is why it doesn't make sense for semiconductor manufacturers to build out additional capacity - by the time it's up and running 14-18 months from now, there's no guarantee anyone will be buying parts in anywhere near the volumes that are needed right now. So this demand shock at the moment is temporary and nobody can predict how long it will last, but you need to know demand way out if you're going to set up a fab expansion. This is why the bullshit move that automakers pulled is having so much outsized impact on the entire industry.

To answer your other questions:

All complex devices with large number of parts in them are super vulnerable to one tiny part halting the production of the entire device. This is why complex devices are poorly available. For GPUs and game consoles in particular there's an extra market-distorting factor of bitcoin fuckers and scalpers, respectively. Apple builds complex devices with large numbers of parts in each.

You can't use a 7nm fab to build things you would normally build on 120nm. Not only is there way too little 7nm capacity, and it's much more expensive per chip, but the smaller the node size the higher the defect rate is going to be, and the lower voltages your parts are going to be able to handle. Automotive and power electronics need to handle high voltages. You will never see 7nm process parts being used for switching mains voltages, for example, or driving 600V motors. In addition, designing parts for newer processes is a lot more work and has very long setup times.

None of the chip shortage can be attributed to AMD or Apple or Intel. They are a totally different category of part/process than the ones that are currently most affected. Intel is still operating near their capacity and selling countless server and laptop CPUs. I don't think this is a factor at all here.

Was about to link to your older comment I have bookmarked[1], glad to see you sharing your experience once again, Kliment!

1: https://news.ycombinator.com/item?id=26931498

Couldn’t agree more.

For a board I just designed, I need 22 uH inductors. I ended up ordering two substitute parts, then just today was able to order the desired parts. Those substitute parts will likely be scrapped. Hell, we now order parts before the design is finished, then switch parts mid-design. The board is not ordered until all parts are in hand. It’s a massive waste.

My opinion: dramatic increase of demand for stuff vs services due to covid - as most services involve face-to-face contact and are either unavailable due to lockdowns, or people are afraid of using them for the fear of getting infected. Because money is saved on using services, people pour it into stock market (hence bubble) and buying stuff, hence shortages.

People who are more likely to be concerned with covid (more educated) are also more likely to consume electronics, so demand on electronics is particularly increased.

Supply side issues are secondary and by now they aren't a big deal at all.

I am curios to know right reason(s) for this because I hear different answers from different people.

> Can automakers not use a 7nm fab?

You cannot take a, say, 150nm or 45nm design and just rescale it in Photoshop to make a 7nm design.

Not only does electrical characteristics change which might require changes, such as leakage currents, but in order to reach the densities of "7nm" things such as transistor construction has changed. For example older, larger nodes used planar transistors while a modern node likely uses non-planar transistors like FinFET[1]. A nice write-up can be found here[2].

The point is, moving from one node to another, especially if they're far between, might require a huge redesign, along with all the validation and testing that entails. A new set of masks for the latest nodes is also very expensive[3].

[1]: https://en.wikipedia.org/wiki/FinFET

[2]: https://semiengineering.com/knowledge_centers/manufacturing/...

[3]: https://anysilicon.com/semiconductor-wafer-mask-costs/ (not current but illustrative)

I'm curious, was any of the shortage due to Chinese companies preparing for sanctions?

I am sure we have been through this many many times on HN. Kliment's comment here [1] with an Auto industry perspective.

Consumer electronics including GPU and consoles demand have increased dramatically due to pandemics. And they use components that uses bigger nodes which are in demand.

>Can automakers not use a 7nm fab?

They could but even 7nm are in short supply.

>How much of the chip shortage can be attributed to AMD and Apple taking the laptop market away from Intel? And if this is true, does this mean that Intel now has unused capacity

Both AMD and Intel are at full capacity. Intel still have server capacity to fill.

Read the Bullwhip effect. [2].

All of that is excluding crypto, and investment / trading ( buying and hoarding ) in semi- components.

[1] https://news.ycombinator.com/item?id=26931498

[2] https://en.wikipedia.org/wiki/Bullwhip_effect

consistently manufacturing some of the most complicated machinery (themselves composed of many thousands of dependencies) on earth is hard during normal times and these are not normal times

The shortest explanation is that COVID changed customer demand, in both directions. And switching from one fab to another, even if they are the same process size, requires significant work.

This caused a scramble where customers found alternate paths to satisfy requirements. Buying up stock of other, similar chips (which now creates a shortage of chip "b"). Reserving time with a new fab. Or deciding not to have navigation in certain car models. Or halting sales for a while. All of which drives another cycle of churn into the schedule.

This repeats itself over and over, like an earthquake with aftershocks, until things smooth out.

My understanding is too much demand in a short period of time. Demand outruns supply. Factories can pump as much chips as they can. Overall there is huge demand for electronics, cars and all other stuff that's using chips. Lockdowns contributed to this because majority of demand shifted from equilibrium(balance of goods and services) to goods and no services or very little services. Or another version of the story is people were saving money in order to buy goods when lockdowns were over.

Something I haven't seen anyone mention yet is demand-pull inflation. I think the reality is a lot of manufacturers and distributors have been reluctant to raise prices, feeling that the demand surge is temporary. In truth, I think it's likely that prices just haven't caught up yet. That means demand has outpaced supply for that whole period.

I think there are other reasons for the shortage, but they've been exacerbated by the inflation issue.

Is there anything to the theory that this situation is, at least in part, artificially created by businesses looking for an excuse to raise prices and increase margins? I've heard that, but I've never seen any actual supporting data. I'm not even sure what data would/could support that.

I guess I'm asking if this theory has any credibility outside of the various conspiracy theory blogs/sub-reddits?

Business can't make more money by not delivering product. If demand and capacity were both the same, but prices were simply raised, things wouldn't be out of stock. So either demand increased, or supply decreased, or both - but simply saying "businesses wanted to increase prices" can't explain the shortages for things that are hard to find.

Even for most of the hard-to-find things, many of the hardest-hit businesses haven't significantly raised their own prices. E.g. in-demand new cars have dealer markups but largely haven't seen huge bumps in MSRP even when sold out like the Ford Maverick or Kia Telluride. PS5/Xbox/Switch prices have been flat. Apple device prices haven't seen any unusual spikes due to shortages.

There's the interesting question for the conspiracy folks: why aren't many of these businesses raising prices? Why is Sony letting resellers sell PS5s for over $100 over MSRP still (e.g. https://stockx.com/sony-ps5-playstation-5-blu-ray-edition-co... ) instead of meeting demand where it is?

Probably those businesses see the PR hit of big MSRP bumps as worse than the short term boost in revenue from capturing more $$$ from the high current demand.

If you mean, would companies take advantage of the situation to raise prices and increase margins, then yes of course they do.

If you mean, this whole problem wouldn't even exist if the businesses weren't trying to create it, no that is not the case. They were fully as greedy and deceptive before 2020.

It should also be pointed out that some businesses involved had, in fact, too low of a profit margin to invest in new capacity, so trying to raise prices and increase margins, while self-interested, isn't always bad from the perspective of the entire economy. If you want new capacity, it needs to have money put into it to make that happen.

Very short version - new plants for making more chips aren't ready yet (but they are being built) and demand still outnumbers supply.

Automobiles are a special case.

Automobile manufacturers are used to ordering suppliers around wrt price, lead times, and so on. Chip makers used to put up with that, but when things got tight and other folks started waving dollar bills around, the chip makers said "hold it - we can get paid more AND not have to deal with those jerks?"

from an eagle eye point of view. A shortage always means, Demand is higher than supply and prices and or supply cant adjust quick enough.

Price cant adjust: Long term contracts, regulation etc.

Supply cant adjust: Move to newer chips (so essentially misplaning), shortage of people, congested supply chains etc.

Intel doesn’t seem to have enough capacity, there are no in stock intel fpgas under $2000 at Digi-Key. Try and get any cyclone v or max 10 chips.

In addition to the manufacturing and logistics mess the pandemic has created globally for everyone, the semiconductor shortage is even more of a special case. Just looking at graphics cards, we're in a perfect storm from a demand standpoint:

1) The pandemic resulted in a significant increase in demand by businesses buying laptops/PCs for employees so they could work from home. Lots of GPUs there and they are still buying.

2) Lots of people as individuals buying gaming PCs and consoles since they were spending much more time at home and wanted something to do. They too are still buying in record numbers.

3) With AI being a hot field right now, lots of GPUs being purchased for deep learning by researchers and businesses. Maybe this is a permanent increase in demand, maybe not... time will tell.

4) With cryptocurrencies being white hot right now, lots of GPUs being purchased for mining.

So you might say 'well why not just build more fabs/factories/whatever and they can all make lots of money?' Well sure, for now. Fabs cost billions of dollars and take a couple of years to come online. The companies that have to invest in them likely spent much of 2020 and 2021 doing everything they could to increase production in the short term while asking themselves 'how much of this demand is long term?' as all of the items listed above are relatively recent changes, rather than things proven to be structural changes. It sounds like by mid- to late-2021 companies like TSMC decided to take the plunge after securing some up front payments and volume commitments from major customers.

In addition to concerns by companies like AMD and nVidia about how long term items 1 and 2 are, they know for a fact that 4 isn't likely to be sustainable given that there was a major crypto crash in the last 5 years and likely to be another one soon. When this happens, this will not only decrease demand as crypto miners stop ordering, but it will likely also increase supply as miners dump their GPUs on the used market (Ebay etc.) to cut their losses. If you follow the GPU news, you'll notice that AMD and nVidia are being very strategic in releasing new mining-specific GPUs. This is partially to offer products that specifically target the needs of miners, but I suspect more to minimize how much of this product can be dumped on the used market hurting GPU demand from consumers down the road.

In addition to all this going on specific to their industry, they're dealing with all of the other manufacturing and logistical issues everyone else is. So they're also having to figure out how to get that PITA power/memory/whatever controller that they need manufactured in quantity and in time to build their own products and so on.

So these companies are responding: they probably did as much as it financially made sense for them to in the short term while also adjusting their longer term plans. I'm expecting them to come online with significant new supply (for the new products released in 2022) just in time to see it dry up soon after (at least in the short term) since that's typically how these things go.

Occam's razor: the simplest explanation is that during global disasters (even slow-moving multi-year disasters) supply chains have trouble coping.

I think, there is a possible grain of truth there.

>> If there were classified orders taking up fab space, would we ever know?

The shortages could have led to larger back-orders, and potential "double booking". Those orders is probably a trade secret, so it probably won't be easy to find out.

Just before the pandemic, everyone was optimising for zero inventory, and probably now you have a feedback loop which probably overshoots a bit in the other direction too.

You need something in half a year or a year, you likely order it right away instead of doing it just-in-time.

It's called the bullwhip effect.

Everyone at every level in the supply chain is buying six months of buffer stock. Then when normality resumes, everyone will consume their buffer stock while cancelling orders, leading to layoffs, idling, and potentially bankruptcy of component manufacturers and another induced shortage.

Because of just in time fanatisicm.

Two people couldn't even cover up Bill Clinton and Monica Lewinsky. You want tens of thousands of people to cover up massive chip orders?

Classified orders from the US government depending on Taiwanese fab space?

No

ASML had a fire today. They provide companies with equipment to bake the chips. It's the hugest one and also kind of the only one with its capabilities. It's not the reason there's a chip shortage today, but also sort of, it is.

Processors are a commodity and everything but there just aren't enough suppliers with good enough product at a low enough price. The process is toxic and complex and requires heavy investment.

It's almost a miracle we get to see the type of efficiency in development that we do today (although others may call this miracle a spiraling debt crisis..)

Trump is also partially to blame.

When he placed that trade embargo on China, many companies were forced to change their orders to factories in other locations, such as Taiwan. These factories were already running at capacity.

Edited for source: https://www.theverge.com/2020/9/26/21457350/us-tightens-trad...

From your source it seems to be an export ban, not an import ban.

From what I remember of various new articles over the past three years (I'm not going to go find all these articles over such a broad time span, I'll leave verification as an exercise to the reader) there are a large number of contributing factors.

- Long time push for "just in time" supply chains. A lot of manufacturers stopped keeping a local stock of chips they needed for the next 6 months of manufacturing to save on storage costs. This is fine as long as there are no hiccups in any earlier part of the supply chain (this transition has been happening over the last 15-20 years).

- Crypto-coin mining and the sudden extreme demand of graphics cards. This has been going on for a long time. A lot of low-nanometer demand goes to a large number of graphic cards and ASICs for mining crypto-currency. This led to the graphics card bottleneck that existed for the last 3 or so generations of graphics cards.

- pre-COVID there were several (I want to say Taiwanese) fab plants that were destroyed due to Typhoons. This led to reduced production capacity.

- There was contamination of shipments of silicon wafers which disrupted production in several fabs. This was a pretty short delay more than anything but it was a contributing factor.

- At least one US based Fab was closed. I believe some other international fabs also closed down either temporarily due to COVID or permanently for other reasons. Some fabs were acquired by major companies and likely switched their production pushing more pressure on other fabs.

- Global politics. There was a lot of aggressive changes to global policy during this time frame that forced a redistribution of where certain chips were being manufactured adding more work for all fabs (retooling, reworking designs, getting a design working in a new fab).

- The canceling of car chips actually relieved the pressure on the fabs as they were able to take other fab orders. I don't think this slowed down fabrication in general, but when the car manufacturers realized they actually did need that capacity and more, the capacity had already been sold to other customers. This is entirely on the car manufacturers and is only kind of related to the overall chip shortage.

- Global distribution issues. This isn't just in the final stage of consumer products that has been affected but also raw supplies affecting the rate some chips can be manufactured.

- Increased demand overall. Consumer consumption of chips has increased pretty dramatically over the last decade pretty consistently as everything wants to be connected to the internet now from your toaster, to your butt plug.

- Complexity and difficulty in building up a new fab. These are incredibly specialized facilities costing in the 9 figures to setup the facility. The people that work in these facilities are highly specialized not just in the profession but to the intricacies of that specific facility and their pipeline. Training up these extremely skilled labor tasks can't be done quickly.

> - The canceling of car chips actually relieved the pressure on the fabs as they were able to take other fab orders. I don't think this slowed down fabrication in general, but when the car manufacturers realized they actually did need that capacity and more, the capacity had already been sold to other customers. This is entirely on the car manufacturers and is only kind of related to the overall chip shortage.

iirc this was the root of it. i think the whole chip shortage began in the car industry.

This is far from a complete answer, and not unique to the chip shortage, but some context about some markets and industries in general for those wondering about "how could this be related to COVID?"

It's important to understand that many manufacturing industries operate on a "JIT (just-in-time) manufacturing" model[0], where goods used in a manufacturing process are only ordered and received as needed, minimizing inventory and storage costs. (e.g., instead of having a warehouse full of screws sitting around waiting to go into products, you order X screws from another manufacturer every month to fulfill Y number of effective orders) This means that many industries (especially the automotive industry) operate on a constant stream of supplied products, in many interdependent chains.

There are two downsides to this approach:

1. The JIT model has to be at least partially predictive. You can't necessarily order components right as you receive orders, because it takes time to produce those components (remember, no one has warehouses full of those components ready to ship, because it's often JIT all the way down) and you don't want to keep customers waiting as components get shipped to you. Instead, you need to at least somewhat predict how many orders you expect to fulfill in order to balance between keeping customers waiting ("shortage") and having excess stock laying around which might never sell (waste)

2. The other main downside is that JIT production is highly interdependent on other industries and can be fragile. If you expect to be able to order X screws this month from your supplier, but some external event causes them to only be able to supply you with Y screws (Y << X), you're shit out of luck. You literally cannot produce your product because the underlying components simply do not exist, which leads to a shortage of your product

Both of these downsides come to a head in the presence of some global event that causes uncertainty and inefficiency. If a low-level JIT process depended upon by many industries suffers from an inefficiency (e.g. workers can't come in because they're sick), many higher-level processes start to slow down (e.g., no one has screws to build their products with). This effect is multiplicative across industries, since many manufacturers specialize in specific products and downtime can affect many "downstream" clients.

Along with this, such global events can dramatically change customer demands, and when you've only predicted needing to produce X products but you now have Y demand (Y >> X), customers are out of luck because the products literally do not exist to be sold (shortage).

Considering that computer chips are now present in almost all major consumer products these days, it doesn't take much uncertainty and inefficiency at the "base" for a domino effect to significantly slow down a lot of industries.

[0]: https://en.wikipedia.org/wiki/Lean_manufacturing

Surely this would explain the initial supply chain shock.

But it has been two years, why isn't the backlog from the initial shock diminishing. I understand that supply is not elastic but two years should have been enough for the producers to ramp up the production.

Do they think that this is a temporary spike and building capacity would be unprofitable?

> Do they think that this is a temporary spike and building capacity would be unprofitable?

Ayup.

A lot of the previous decades of chip manufacturing is all about how superior it is to be a "fabless" company. You don't have those nasty capital expenditures and your upside is basically unlimited--while owning a fab means you have to employ a ton of people and your profit margin is effectively capped.

The fab folks know that as soon as this shock is over, owning a fab goes back to being crap. So, you make hay while the sun shines, and you don't even think about doing capital outlays.

Even if they could find the equipment and people (not at all an easy thing), there is ZERO incentive for the fabs to expand. To first order, it's actually more profitable for them to maintain this shortage. Nobody is going to pass up a sale just to keep a shortage going, however.

Eventually, things will equilibrate. However, I suspect that there's going to be a lot less choice in semiconductors for quite a while afterward.

The only short-term thing that I'm interested in tracking is how many companies were hoarding and are going to panic now that inventory tax is coming due. That's going to move some commodity microcontrollers back into the system from the hoarders. It will be interesting to see how much comes back into the system in the next quarter or two.

> two years should have been enough for the producers to ramp up the production

Along with what Kliment said about not investing in building capacity, I don't think this specific statement is necessarily true. Many of the original factors that caused these setbacks in the first place are still ongoing: everyone is still trying to catch up, and demand is still outstripping supply almost everywhere, for many, many products; worker mobility is still extremely limited (across countries and regions) because of the ongoing pandemic, so it's not trivial to ramp up.

Short of investing in that capacity and expanding into new regions, I'm sure it hasn't been easy to actually ramp up production again, and changes in customer demand and patterns are still shaking out. (And even for those companies which have been investing into ramping production, two years can be the low end to start seeing returns on investments: plants take time to build, resources need to be allocated and transported, new employees may need to be trained, etc.)

See my other comment in this thread - it makes no sense to ramp up production because the current demand increase is almost entirely due to lack of market confidence and there's no guarantee it will last long enough for new capacity to go online. That's why ~nobody is building new capacity for the affected nodes.

It can take time for supply to increase. Chip fabs can't be built instantly no matter how much money you're willing to throw it the problem, any more than babies can be made in five months if the mother really tries hard.

Trust me, if you pay enough for that chip, it will be there first thing tomorrow morning

I don’t understand. Are you saying there’s no shortage because _technically_ you can get whatever you want, given infinite cash? Is that particularly insightful?

If it’s due to inflation, why did chip prices skyrocket a few orders of magnitude more than inflation?

It's a spectrum, not black and white. But "supply" is being used as an excuse to mask the much bigger problem: inflation. Everything is more expensive, and unless people are willing to pay more for it, they won't get it.

A lot of people don't want you to understand that.

When the chip demands (by the downstream companies) went down on the early phase of global pandemic, Chinese companies gobbled up the capacity because they were recovering first and also feared that the Trump would trash more Chinese companies.

supply chain causing raw material lag altogether ?

Covid

because of the ASML fire?

crypto

> Ask HN: Why is there a chip shortage?

Most shortage todays is on 200mm.

Why? Trump's sanctions on SMIC brought down like half of world's 200mm capacity.

> Can automakers not use a 7nm fab?

No, process, and cell libraries are incompatible. Chips will have to be redesigned.

95% of automotive chips are being manufactured on 90's era i-line and KrF steppers that China can manufacture locally.

Only missing piece domestically were the excimer lasers, but they've had that solved (probably through industrial espionage) for three or so years.