Any bets on how long before the component shortage bubble explodes?

[peter-h]

It is just a bubble. The fact that most passives are easily available confirms what is already obvious: the demand for the end product is not significantly larger.

What's happened is that a lot of firms converted cash into a stockpile of single sourced chips. The crazy JIT process (operated by most big firms) has just made everything worse.

I've just had a delivery of 100 STM 32F4 chips, from a normal disti. Order was placed in July. At one point the delivery date was revised to mid 2022...

Also there are the usual cowboy distis selling the chips for say $100; the key being that they have the chips to sell.

I would bet on a bloodbath in a few months' time.

[Kleinstein]

With computers there are not that many passives that go with the chips - mainly the decoupling caps and there we had a supply shortage some years ago and may have reserve capacity as a consequene. For the very fine structured part there seem to be some shortage and limits in the capacity. Even now not many put an Intel CPU in the shelf hoping the price will go up. Still these chips see some shortage.

For the more common chips there is some stockpiling - not just for speculation, but from those who need the chips and now buy before and not just in time. There may be an increasing number of NOS parts coming up, but this may take some time to happen. Many stock will be just used and than maybe lower sales in 2023, when the inventory goes down to a more normal level.

[Stray Electron]

  What is your interpretation of "Explodes"?   Companies like Ford have parking lots full of new trucks that are missing the engine CPUs and the deliveries of new vehicles are running months behind and it's not uncommon for dealers to be charging $20,000 above list price but still business continues so what would you consider an "explosion"?

  Frankly I don't see any impending explosion, all I predict is that some industries, such as new car production, will simply grind to a halt due to a lack of parts. Even so, there is still a huge number of used, vehicles, used computers, etc so it's doubtful that anyone or any industry is going crippled due to a lack of NEW products.

   Keep in mind that most countries have encountered this repeatedly in their histories; consider the lack of NEW cars and hundreds of other NEW products during WW-I and WW-II.

  Sorry, Chicken Little but the sky isn't falling!

[jmelson]

I was looking for some Xilinx chips several weeks ago, Digi-Key said they would have no stock until December.
I just placed the order, to get into the waiting line, they shipped them 2 days ago!

Jon

[bdunham7]

It is just a bubble. The fact that most passives are easily available confirms what is already obvious: the demand for the end product is not significantly larger.
I would bet on a bloodbath in a few months' time.

The manufacturing and marketing of simple passives is probably quite different from semis.  The backlog of container ships at ports would indicate that demand for end products is larger at the moment.

I agree that there is likely to be an implosion of some sort, seeing as China seems to be having a second Cultural Revolution of sorts, but I wouldn't bet on that implosion turning out in a particularly good way.  If you think it will result in universally plentiful semis at bargain prices, you're more optimistic than I am. 

[Cerebus]

  What is your interpretation of "Explodes"?   Companies like Ford have parking lots full of new trucks that are missing the engine CPUs and the deliveries of new vehicles are running months behind and it's not uncommon for dealers to be charging $20,000 above list price but still business continues so what would you consider an "explosion"?

  Frankly I don't see any impending explosion, all I predict is that some industries, such as new car production, will simply grind to a halt due to a lack of parts. Even so, there is still a huge number of used, vehicles, used computers, etc so it's doubtful that anyone or any industry is going crippled due to a lack of NEW products.

   Keep in mind that most countries have encountered this repeatedly in their histories; consider the lack of NEW cars and hundreds of other NEW products during WW-I and WW-II.

  Sorry, Chicken Little but the sky isn't falling!

You've clearly missed the sense that "bubble" was used in, stemming from the "south sea bubble" of 1720, meaning an unsustainable overpricing of, and consequent demand for, something so that prices inflate, and inflate, and inflate until the "bubble bursts".

[TimFox]

The "South Sea Bubble" of 1720 was actually securities fraud.
More recent "bubbles" have involved inflated real estate values.
A good historical summary is "A Short History of Financial Euphoria" by J K Galbraith.
https://www.penguinrandomhouse.com/books/321392/a-short-history-of-financial-euphoria-by-john-kenneth-galbraith/

[Cerebus]

The "South Sea Bubble" of 1720 was actually securities fraud.

Yes, it's just when the term popularly entered the language.

[TimFox]

I was just pointing out that there were different causes for the historical "bubbles".
Another term for mining or commodity fluctuations is "boom and bust", implying cyclical behavior.
My personal favorite fictional description of fraudulent corporate securities activity is "How We Live Now" by Anthony Trollope (1875).

[Benta]

The "shortage" is to 90% psychological. Happens around once a decade.
There are several elements:
1: the really big customers (automotive/cellphone etc.) get a scare, eg, Corona and start cancelling orders, expecting falling demand (big mistake). This really pisses off the semi manufacturers, because they are operating on minimum profit margin here.
2: the end customer demand falls much less than expected. Panic ordering sets in to replace the cancellled orders. Supply chain breaks down.
3: FT (or whoever) writes an editorial about chip shortage.
4: the medium-sized customers start double/triple/quadruple ordering at the distributors, creating the bubble.

End result: total chaos in the supply chain management. There may be a smal deficit in wafer fab capacity, perhaps a couple of percent, but calling out for the semis to build new wafer fabs is hypocritical at best.

The reponsibilty of the whole mess lies squarely with purchasing departments at a few major customers in automotive and consumer markets.

PS: I loved selling semis to industrial customers. No purchaser ever got fired for paying a bit more, providing he got the parts when needed. Very loyal and reliable customers. In automotive and consumer it's all about squeezing the price til the manufacturer screams. Not nice.

[Cerebus]

Go back and read what he said again. "Psychological causes" ≠ "imaginary phenomenon". You can see it at the moment in the UK, a minor difficulty with petrol and diesel distribution but no actual shortage of petrol at the pumps has been turned into petrol pumps drained dry left right and centre by panic buying. There was no petrol shortage, people believed there was and turned it into a reality.

[Benta]

Who's "he"?

[MT]

Dont exclude and underestimate Klaus Schwab and the Davos crowd etc and their planetary eugenically feudal Great Reset agenda in this component shortage.

[Cerebus]

Who's "he"?

"The cat's father!" (traditional retort). The reference was to you. Did I make a gender boo-boo?

Edit: Now I see the confusion. Hareod made a remark that suggested you were talking out of your hat because he/she/it was experiencing actual shortages not what he/she/it took as you implying by taking psychological as "not real, imagined". That message has disappeared making mine into a non-sequitur. (I don't normally quote things I'm replying to if the reply will naturally follow the comment.)

[Bud]

Dont exclude and underestimate Klaus Schwab and the Davos crowd etc and their planetary eugenically feudal Great Reset agenda in this component shortage.

Dont also miss appointments with your psychiatrist. 

[floobydust]

This is trade wars, I think it has nothing to do with the usual excuses and "supply and demand" curves economists believe in.

Look at an input cost, that of polysilicon it's up 250% this year. I think almost 80% of the world's polysilicon comes from chinese plants.
US Senate passed the Uyghur Forced Labor Prevention Act which bars imports of all goods produced in Xinjiang, a manufacturing hub for silicon metal and polysilicon. Cotton and tomatoes were already blocked.

china recently ordered plants to reduce polysilicon output by 90%... throttle down to only 10% of your August output for the rest of the year at least. This despite adding capacity.
I think just to create a shortage and spike up the prices, and retaliate. High-grade is used for IC's and lower grades for solar cells.

Prices will continue to skyrocket based on politics alone.

[peter-h]

The cost of silicon must be a negligible portion of a USD 5-10 chip.

Also prices aren't really going up (except for Maxim but they are a joke anyway) - unlike previous bubbles of long ago where a 74LS245 would go GBP 0.25 to 2.50 and then back to 0.25 when it all crashed, and I am talking of normal distribution, not the cowboy dealers. This time round the cowboy dealers are also around but selling STM 32F4 for USD 50-100.

[Benta]

Who's "he"?

"The cat's father!" (traditional retort). The reference was to you. Did I make a gender boo-boo?

Edit: Now I see the confusion. Hareod made a remark that suggested you were talking out of your hat because he/she/it was experiencing actual shortages not what he/she/it took as you implying by taking psychological as "not real, imagined". That message has disappeared making mine into a non-sequitur. (I don't normally quote things I'm replying to if the reply will naturally follow the comment.)

Hahaha! No gender boo-boo, I don't give a hoot about that kind of thing.

But posting something provocative and then deleting that post after someone else answers is trolling at its worst. Your analysis is completely correct, I lost the connection between your post and the troll.

[Cerebus]

Who's "he"?

"The cat's father!" (traditional retort). The reference was to you. Did I make a gender boo-boo?

Edit: Now I see the confusion. Hareod made a remark that suggested you were talking out of your hat because he/she/it was experiencing actual shortages not what he/she/it took as you implying by taking psychological as "not real, imagined". That message has disappeared making mine into a non-sequitur. (I don't normally quote things I'm replying to if the reply will naturally follow the comment.)

Hahaha! No gender boo-boo, I don't give a hoot about that kind of thing.

But posting something provocative and then deleting that post after someone else answers is trolling at its worst. Your analysis is completely correct, I lost the connection between your post and the troll.

I don't think it was intended to be provocative or trolling (slightly sarky maybe, but nothing extreme), I think it was a genuine "didn't read carefully enough". Hareod probably figured it out either spontaneously or because of my comment, went "Doh!' and withdrew the post either not realising or spotting the discontinuity it would cause.

[Benta]

This is trade wars, I think it has nothing to do with the usual excuses and "supply and demand" curves economists believe in.

Look at an input cost, that of polysilicon it's up 250% this year. I think almost 80% of the world's polysilicon comes from chinese plants.
US Senate passed the Uyghur Forced Labor Prevention Act which bars imports of all goods produced in Xinjiang, a manufacturing hub for silicon metal and polysilicon. Cotton and tomatoes were already blocked.

china recently ordered plants to reduce polysilicon output by 90%... throttle down to only 10% of your August output for the rest of the year at least. This despite adding capacity.
I think just to create a shortage and spike up the prices, and retaliate. High-grade is used for IC's and lower grades for solar cells.

Prices will continue to skyrocket based on politics alone.

I know the Bernreuter report. But what's missing in this conspiracy hypothesis is, that it's got very little to do with semiconductor manufacturing (as in semicinductor chips).
The Chinese polysilicon manufacturing is almost all about solar panels - and China is very succesful in delivering those. The main suppliers for the semi fabs are still Wacker, Hemlock et al.

No, the issue is still wafer fab capacity, which might be a couple percent low, and irresponsible pruchasing strategies from big customers.

[Benta]

@Cerebus: I've no idea, I didn't see the post. Let's leave it there. Cheers.

[tooki]

It is just a bubble. The fact that most passives are easily available confirms what is already obvious: the demand for the end product is not significantly larger.

I think your conclusions are wrong.

1. Making passives isn’t nearly as difficult. There’s a lot of manufacturing capacity, and it’s easier to ramp up than semiconductor manufacturing, which is insanely complex and time-consuming. The start to finish process for modern semiconductors takes around 3 months due to the sheer number of steps involved (about 700 steps).

2. The high availability of passives is likely partly due to the fact that they’re useless without the semiconductors they’re used with. No chips = no boards being made = less passives being bought + factories still pumping them out = plenty on the market.

[floobydust]

The fab reports say production is already up, 200mm is +5%. World Fab Forecast Report (2020 to 2022) shows historic investment, $90B in equipment wow.
Intel celebrates breaking ground. Oh just 3-4 years to go on a puny $20B fab

I thought it was strange Wacker AG 5-year contract for selling 70MT polysilicon to chinese JinkoSolar. As if the chinese need any for solar lol. Wacker Germany is doing excellent but USA is reducing shifts, something to do with the chinese 57% duties on US PV. So it doesn't really make sense to buy German polysilicon (PV, not IC grade) when you are throttling back your own production PV grade 90%.

I think geopolitics and a trade war are underlying the shortages.

[T3sl4co1l]

It is just a bubble. The fact that most passives are easily available confirms what is already obvious: the demand for the end product is not significantly larger.

I think your conclusions are wrong.

1. Making passives isn’t nearly as difficult. There’s a lot of manufacturing capacity, and it’s easier to ramp up than semiconductor manufacturing, which is insanely complex and time-consuming. The start to finish process for modern semiconductors takes around 3 months due to the sheer number of steps involved (about 700 steps).

2. The high availability of passives is likely partly due to the fact that they’re useless without the semiconductors they’re used with. No chips = no boards being made = less passives being bought + factories still pumping them out = plenty on the market.

Modest increases perhaps, but mind that bringing new capacity online takes almost as long.  Remember the 2018 capacitor shortage.

Tim

[Simon]

Dont exclude and underestimate Klaus Schwab and the Davos crowd etc and their planetary eugenically feudal Great Reset agenda in this component shortage.

reset to what? what does anyone stand to gain other than the grey market dealers that have no influence and can cash in on parts. Which is nothing new. Just look for obsolete infineon parts, if the part is popular they are 4x the price, if the dealer messed up and the parts were not so popular they are 0.25x the price. This is obvious with infineon stuff as they just obsolete stuff no matter how popular it is as popular to them means the car makers buy them direct in the millions of peices.

I have just got my employer to pre-order 1'000 SAMC21 microcontrollers, they are due in November, the next predicted shipment is september 2022. That is us fixed for any microcontroller required in a design for at least the next year as I know it will cover any requirement. next I will move on to voltage regulators and then general MOSFET's.

There were already issues with silicone based parts before covid, demand was growing, electric cars are on the rise, people don't "just" build fabs's that cost billions of dollars. There was an earlier shortage of passive a couple of years ago, obviously the industry caught up and now there is little use for a ceramic bypass capacitor for a microcontroller that you cannot get.

So as for your conspiracy theories, please take them elsewhere - maybe to a shrink.

[harerod]

Benta, Cerebus - sorry for the confusion. A couple of minutes after posting, I decided that my post didn't live up to my standards, checked that nobody had answered yet and then deleted the post.

My post was aimed at Benta's "psychological", which allows huge room for interpretation. Anyways, here is the original post that I deleted, verbatim:

The "shortage" is to 90% psychological.
...

Maybe you are just the doctor that I have been looking for. In my line of work I have been hallucinating for nearly a year that 90%+ of the silicon in my BOMs are unobtanium or at least beating platinum in price per mass.
Seriously, I know that I am using a lot of special components, but without those ADI/ST/TI single source devices we couldn't do the stuff we are doing. The other day I looked for opamps. It has been a long time since I settled for inferior components, just to get something into manufacturing. My old East German mentor keeps quoting GDR jokes. I haven't heard that many GDR joke since the wall came down.

[Simon]

The problem is that we no longer live in a world populated by:

74 logic
741 opamps
78 regulators
16F84 micro controllers
1N4007 diodes
2N7000 mosfets
IRF540 power mosfets

Each subgroup of silicone devices is now full of a host of different devices, unless there is demand for all of them in volume how can we expect anyone to make them and keep dead stock on the shelf. I don't know what the minimum run for a silicone device is but yea.....

ST seem to be very popular because they are cheap and from the little I have seen not great but the world runs on cheap. I chose the SAMC21 range which means I'm competing less and I get a chip that may cost a bit more but will do all I want.

Currently work are struggling to get a microchip 3.3V regulator. It was clearly used by the idiot designer because it was cheap, the cheapest. Well guess why it is so cheap, it's only 1.6x1.6mm, it also seems to be the only regulator in that package. Had he chosen a regulator in a 3x3mm package you would be able to get some right now from several manufacturers making pin compatible devices.

We need to move back to the days of more standardization in some parts, now your £100 board is held up by a voltage regulator of all things that costs pence.....

[Just_another_Dave]

The problem is that we no longer live in a world populated by:

74 logic
741 opamps
78 regulators
16F84 micro controllers
1N4007 diodes
2N7000 mosfets
IRF540 power mosfets

Each subgroup of silicone devices is now full of a host of different devices, unless there is demand for all of them in volume how can we expect anyone to make them and keep dead stock on the shelf. I don't know what the minimum run for a silicone device is but yea.....

ST seem to be very popular because they are cheap and from the little I have seen not great but the world runs on cheap. I chose the SAMC21 range which means I'm competing less and I get a chip that may cost a bit more but will do all I want.

Currently work are struggling to get a microchip 3.3V regulator. It was clearly used by the idiot designer because it was cheap, the cheapest. Well guess why it is so cheap, it's only 1.6x1.6mm, it also seems to be the only regulator in that package. Had he chosen a regulator in a 3x3mm package you would be able to get some right now from several manufacturers making pin compatible devices.

We need to move back to the days of more standardization in some parts, now your £100 board is held up by a voltage regulator of all things that costs pence.....

I find the lack of standardization in electronics an important problem nowadays. Last year, we had to solder backwards a regulator in a prototype as a consequence of a wrongly specified reference voltage by our client (it needed to be provided to one of their boards from our pcb). The manufacturer of the regulator that they recommended had assigned the pins differently in each of the regulators of that family and the one that they told us and the required one had their footprint mirrored

[Simon]

3 pin regulators tend to be the same unless they have an enable pin although I have not compared those. MOSFET's in common 3 pin packages seem to have common pinouts as do the SOIC8 ones which saved my ass recently.

But microcontrollers - all different. any other more or less specific IC like a SM converter will be unique there is no stopping that, it's more a case of no one makes drop in replacements for others parts. It's not a case oy why make my new part interchangeable with an existing one that means people can use it instead? well it works both ways, making your part the same footprint and pinout as another means people can use it with no fear.

In my last job we changed a fan for cheaper and better one on the basis that it was the same size and mount hole pattern so we could revert to the original fans if we wanted to.

[DC1MC]

What about JEDEC standardistion efforts, are they still ongoing ?

[peter-h]

Yes; regulators are scarce too. The old LM2936 is disappearing from stocks, unless you buy the version with the enable pin (which far fewer people want) and buy it from Mouser (who charge more than anybody else - how the hell did they become one of the very few TI distributors in the UK??). The MIC5201-5.0YM-TR has the same pinout but the enable pin works backwards, which matters if you are using a 2936 with the enable pin. But there are loads of regulators...

I don't really agree that old components are not used/usable anymore. The 1N4007 of decades ago is now SMT and might be a BYG10M or one of dozens of equivalents. So if an LM358 does the job, use that!

Choosing components, using old ones as far as possible, is an art, which few people have unless they have been at it for a long time and preferably have been responsible for their own company. But if you do it right, you eliminate 99% of headaches.

What has happened, over past few decades, is that manufacturers have focused on unusual parts, with some unusual feature, to entrap less experienced designers into designing-in unusual parts. That's where the money is. It is not in LM358s, or BYG10Ms.

ST are successful because they deliver good value. They do lots of cheap chips. Look at the STLED316 display driver. €0.75 1k+ and that's from Mouser which is always expensive. You get a driver for 6 digits for that. They also do loads of less well known op-amps, comparators and such like. And the 1k price for the 32F417 is about £5.

[Simon]

I'm not saying that the old ones are not usable or used, but that those jellybean part numbers are now an ever shrinking portion of the varied mix. If you search 1N4007 you will find few that are really called that but are a spec match. Now for 2N7000, those do not even have pin compatibility between manufacturers while marketed as the same number, the 2N7002 will have as they do that on SMT parts but look carefully at the specs. You get parts like NX7002.... these are the same thing but may have slight variations like in exact gate threshold but overall are interchangeable.

But as you get more complicated than a transistor pin compatibility gets more unlikely unless some sort of silent standard has established itself.

I'm almost suggesting we build in house as we will soon have to hold all of our own stocks and free issuing means you get reals back from a 50 peice build with 5 parts floating in the bag as the tape has to be pulled back by a few parts in the feeder.

ST data sheets give me the heebeegeebees , they read like they are written by lawyers and I found them the worse for getting to the bottom of what a part can actually do. Compare an ST datasheet for a switch mode converter to an LT one...... I'm making a few hundred, I can't spend hours saving pounds on the entire product run of 300 plus worry I misunderstood an intentionally ambiguous datasheet.

[Cerebus]

The manufacturer of the regulator that they recommended had assigned the pins differently in each of the regulators of that family and the one that they told us and the required one had their footprint mirrored

Grrr. People who design a family of parts with different pinouts within the family should have their toes desoldered.

[TheDane]

As soon as the shipping world gets back to normal.... delivery rates will go down and shipping time will decrease - and chips be delivered cheaper and faster.
- and when is that going to happen 

I read this on SlashDot a while ago, with some interesting informative comments:
https://news.slashdot.org/story/21/06/16/2254236/why-we-are-in-a-shipping-crisis-thats-sparking-shortages

[Simon]

Shipping is not the issue, it's the manufacturing.

[Cerebus]

A year ago shipping a container from China to the UK cost £1600, this year you're talking £10,000; and the same thing has happened to shipping prices all over the world. That is of course going to have an effect on everybody. Remember, not just finished parts require shipping, so do raw materials, sometimes several times (e.g. feedstock => resin manufacturer => chip encapsulating plant => distributor).

[Simon]

Yes, but here we are talking just plain not being able to get hold of stuff.

[Cerebus]

The increase in shipping prices is mostly being driven by the demand curve. That is, shortage of shipping capacity is driving shipping prices up. You can't get stuff if you can't find a ship to carry it at a price you're willing to pay.

[Simon]

https://youtu.be/KrTmvqwpZF8?t=2958

[Bassman59]

There were already issues with silicone based parts before covid,

Silicon and silicone are NOT the same thing!!

[Simon]

true, I mis-spelt it in my haste.

[Infraviolet]

The chip shortages, and associated higher prices end when, and only when, there is enough manufacturing capacity back online to cover for the demand. That won't happen until several new factories have been built, or existing ones repurposed, to overcome things like the factory which brutn down in 2020 and those which have closed. Could take anywhere from one year to 4 years I'd guess, the incentive for new semiconductor plants is good, but setting them up will take time.

[peter-h]

No; there is no real "demand".

The buying we are seeing is panic buying. People with cash are stocking up. Eventually they will get bored with it, and then there will be a crash.

[daqq]

Not sure of your definition of a bubble.

The processes and requirements for manufacturing of passives are trivial in comparison to the manufacture of low complexity semiconductors. Not sure it's a good comparison.


edit:

Grrr. People who design a family of parts with different pinouts within the family should have their toes desoldered.

Yes oh yes oh yes, this a thousand times this!

[Simon]

No; there is no real "demand".

The buying we are seeing is panic buying. People with cash are stocking up. Eventually they will get bored with it, and then there will be a crash.

Demand is up and we know that, I am now looking to prestock. If the manufacturer tells me that my microcontroller can be had in November 2021 or September 2022 what do I do? I ordered 1'000 with the assurance that of any of the projects I am aware of those ones will do all of them. I don't know how many are made at a time but only 8000 were available on preorder for 2 months time.....

[VK3DRB]

The component shortage is not a bubble. It is caused by serious issues such as poor planning, government stimulus, electric vehicles, people working from home, the big freeze in Texas caused by global warming, but the biggest overall cause is the China virus causing global havoc. Opportunistic brokers is only a small aspect of the problem.

ST Micro, Texas Instruments and Bosch (Sensortec) and a number of other players are in strife. TI's website has virtually nothing about the global chip shortage other than some frogshit about protecting their employees. I have read statements released by some of the big chip company CEO's that clearly mask the truth.

The "Notify me when they are available" tick box does not cut it. These companies won't communicate honestly to their customers about what is going on. Maybe because they are more interested in their share price than their customers. Their websites advertise parts that are simply not available anywhere, now or in the foreseeable future.

Analog Devices does keep an internal cache of chips within their company for design engineers. They are a smart company, realising looking after the engineering community will pay dividends further down the track. Prices are high, but have very good IP in their chip design, but at least we can get some for R&D purposes  .

I expect 2022 will be the same as 2021. 2023 may start seeing some normality in the chip supply chain unless another disaster occurs like China invades Taiwan in which case all bets are off.

[jonovid]

need to discern the times. In many ways this is like the 1930s only this time its both Russia and China ramping up the wardrums. 
a slow closing off from the west. growing suspicions & chaos as shipping containers start piling up everywhere,  food price inflation.
and the slow lost of western freedoms. disruptions everywhere,    looks like normal may be lost forever.  prove me wrong!

[peter-h]

" I am now looking to prestock"

Opportunism in the supply pipeline is a part of this. Some months ago I got onto one well known UK disti regarding the STM 32F417. They said if I order 500 I can have 20 right away and a certain delivery time for the other 480.

I had no use for 500 anytime soon since this product won't be in production until into 2022 but had to order the 500.

In the meantime, of course, the originally quoted lead time has vapourised, so this was a con.

I wonder how much of this went on. Distis are opportunists and have always been major players in starting off the a-word ("allocation") rumours, and 6-12 months later the world goes mad. Every time.

I reckon it will all melt down before 2022. Medium size manufacturers will soon be sitting on massive overstocks.

[Simon]

I'm looking at what we are likely to use and getting some. I can't buy 10 micro controllers, spend weeks writing software then find we can't go into production sure to no stocks. A product I am working on now has parts that are on a 2 year lead time, obviously the alternatives will also be vaporising fast. We would rather just ask a subcontractor to make the boards but by the time we do the boards will be impossible to make if we don't free issue stock.

[chickenHeadKnob]

need to discern the times. In many ways this is like the 1930s only this time its both Russia and China ramping up the wardrums. 
a slow closing off from the west. growing suspicions & chaos as shipping containers start piling up everywhere,  food price inflation.
and the slow lost of western freedoms. disruptions everywhere,    looks like normal may be lost forever.  prove me wrong!

Currently there is strong inflation, particularly in food, correct. In the 1930's it was the opposite, profound deflation, the velocity of money crashed.

 As far as the demand for electronic components  being real I would say yes and in many sectors. Demand for PC's, graphics cards web cams, ect. Automotive demand is also real. It doesn't take much of an increase in demand to cause a shortage when the production side is inelastic.

[station240]

It is just a bubble. The fact that most passives are easily available confirms what is already obvious: the demand for the end product is not significantly larger.

I think your conclusions are wrong.

1. Making passives isn’t nearly as difficult. There’s a lot of manufacturing capacity, and it’s easier to ramp up than semiconductor manufacturing, which is insanely complex and time-consuming. The start to finish process for modern semiconductors takes around 3 months due to the sheer number of steps involved (about 700 steps).

2. The high availability of passives is likely partly due to the fact that they’re useless without the semiconductors they’re used with. No chips = no boards being made = less passives being bought + factories still pumping them out = plenty on the market.

https://www.kitguru.net/channel/generaltech/joao-silva/passive-components-may-be-hit-by-shortages-too/
"According to DigiTimes, Chemi-Con, Nichicon, and Rubycon are responsible for over 50% of the capacitor market worldwide. Currently, most of their factories are located in Malaysia and Indonesia, which have been lockdown during August and July due to the increasing number of Covid-19 pandemic cases found in the population. Unfortunately, the lockdown severely affected the factories, forcing the companies to reduce working staff and production, which resulted in a 30% to 60% decrease in shipments."

"Lead times for the passive component order are also increasing significantly. Usually, customers need to wait four to six weeks to get their order. Now, lead times range from three to six months. To make things worse, the demand for these components has skyrocketed in the last year. They are used by a multitude of products, including 5G, consumer electronics, electric vehicles and renewable energy technology. As these industries grow, demand for capacitors has increased proportionally.

With the lack of capacitors from these companies, Taiwanese component manufacturers have become the go-to alternative. During H1 2020, these manufacturers’ YoY revenue grew by 20% or more. The shortage is expected to last for the rest of 2021"


So short version.
Capacitors made in Taiwan are taking up (some of) the 30% to 60% decrease in production in Malaysia and Indonesia based factories.

[peter-h]

There is never a shortage of scare stories.

Like the current UK petrol shortage. 100% engineered.

Nobody in the industry will ever say component x is freely available.

[peter-h]

I am already buying Maxim chips for 1/3 of franchised disti prices, which themselves crept up to close to mouser.com prices which are always over the top anyway.

From somebody's overstock, of course...