GM Slows Production in Some Plants Due To Semiconductor Chip Shortage (axios.com) 61
General Motors is reducing production in some of its North American plants due to a global semiconductor chip shortage. From a report: The chip shortage is affecting automotive companies around the world, with semiconductors functioning as a key component for steering systems, car brakes and other automobile features. GM has temporarily closed some plants, with expected downtimes ranging from a week to several weeks. GM expects the closures will cost them between $1.5 billion and $2 billion in operating profits this year. The chip shortage stems from slowed production and manufacturing in 2020. Semiconductor chips require long lead times due to their complicated technology, resulting in a backlog of demand.
Small wonder (Score:1)
Bill Gates bought them all to put in people.
Re: (Score:2, Insightful)
Bill Gates bought them all to put in people.
I wasn't sure if this was the dumbest theory related to Bill Gates ever created, but I did forward a bunch of emails about it.
I'll let you know how it goes when I get my check.
Re: (Score:2)
The smallest IC is the Hitachi Mu-Chip, which is 0.6 mm by 0.4 mm. That's about the size of a grain of salt.
Covid vaccines use a 23 gauge needle with an internal diameter of 0.35 mm.
So it is obvious that Bill Gates has either figured out how to make smaller ICs, so has somehow slipped bigger needles into the supply chain.
We still aren't sure how he is harvesting power from the 5G network (centimeter-scale wavelengths) for a 0.4 mm chip immersed in electrically conductive blood plasma.
Re: (Score:2)
Re: (Score:2, Funny)
We still aren't sure how he is harvesting power from the 5G network (centimeter-scale wavelengths) for a 0.4 mm chip immersed in electrically conductive blood plasma.
One of the lesser known discoveries from the LHC back in 2012 (when they accidentally created the Mandela effect), was Unicorn Farts.
Pretty sure that's how they're doing it. That's also why COVID was invented. Lose your sense of smell, and you can't easily tell when Communist Farts, are smelling you.
Re: (Score:2)
Chips are able to burst cell walls and harvest power from ATP found in cytoplasm.
Re: (Score:2)
I searched all over when I got my vaccine, but I couldn't find Ponch and Jon to install my CHiP.
12 - 16 weeks for PCs (Score:2)
We've been told to expect any PC ordered to take 12 - 16 weeks to be delivered as a direct result of the chip shortage. That means if we want to replace machines in September/October, we need to order them now to have any chance of staying on schedule. We can't even get things like mice and keyboards because of the chip shortage.
Sorry, not sorry, it shouldn't take four months to get
Re: (Score:2)
Just in Time was always a poor idea (Score:2)
In addition to not removing cost, it also introduces an increased number of points of failure. Your JIT part is in a container? guess what, that critical part is stuck behind a ship that managed to block a canal. On a truck? well, lot of things can go wrong there.
It is a f
Re: (Score:3)
Stockpiling screws makes sense because the only downside is the inventory cost. They will eventually be used.
Stockpiling ICs makes less sense. If they are found to have a defect or are replaced with an upgrade, a lot of expensive components are going to the landfill.
Nonetheless, some ICs were stockpiled. But we are more than a year into the pandemic and the inventory FIFO is not infinite. The stockpiles are now exhausted.
Re: (Score:2)
This sounds like a business opportunity.
Now presumably these aren't the most sophisticated chips on the planet; they just solve particular problems in car designs. The vendors aren't making them all time, they just tool up when they have an order. But while they do that, everyone all around has money sitting around tied up, doing nothing. A vendor who could cut that time by a few weeks would save its customers a lot of money and be more financially efficient itself.
Re: (Score:2)
Now presumably these aren't the most sophisticated chips
Clearly you've not worked in semiconductors. Anything going lower than 100nm, which is pretty much everything MOS except maybe a few families like CD, LS, HC, HCT, and what not, is not some trivial task to make. As for microcontrollers, PICs, PLCs, and FPGAs (among others), I can't name a single one that's been made in the last decade that isn't FinFET at the least.
they just solve particular problems in car designs
From what I understand, which isn't much so take with grain of salt. Pretty much you don't have a modern car without ICs. So if the ICs are t
Re: (Score:2)
I've read one of the bottlenecks is an LCD display driver chip.
If fabs worldwide are always operating at full capacity, and there is nothing that can physically be done to increase their capacity, and there is a global shortage of chips, clearly we need more fabs.
I didn't make any assertions about whether it is *simple* to address this problem. I assume if there were a simple solution it'd have been found by now. More or less I think I'm stating the obvious.
Re: (Score:2)
clearly we need more fabs
Yes. Yes we do. They are serious investments, so for me, the interesting part will be who would be willing to upfront the capital to begin construction.
I didn't make any assertions about whether it is *simple* to address this problem.
Hmmm. Alright then, I must have misread your comment then. It had come off as such to me on first reading.
Re: (Score:2)
Dunning-Kruger is so rampant it's probably a reasonable heuristic to assume most people have no idea how hard things are.
Re: (Score:1)
Re: (Score:2)
Re: (Score:2)
JIT is extremely efficient
Efficient in what respect? There are numerous ways to measure efficiency.
and cuts cost substantially
The additional cost to store extras of certain parts can be trivial, though I'll grant this is situational and highly variable (depending on fixed asset, e.g. building, requirements). There are no additional costs (beyond the opportunity cost of capital), and there can actually be cost savings in bulk ordering and bulk delivery.
No, there is one overriding reason that JIT took the manufacturing world by storm, and that's based on a fina
Re: (Score:1)
Semiconductor chips require long lead times due to their complicated technology, resulting in a backlog of demand.
Meh...I want Fullyconductor chips. None of that "Semi" crap. If they made fullyconductor ships they could simply snap them in two and you'd have twice as many semiconductors! Shortage resolved!
Re: (Score:2)
Relays use conductors instead of semi-conductors. You could make a circuit out of a few thousand of them and haul it behind your car.
Re: (Score:2)
The chips facing shortages are the chips that were produced in huge quantities and where there typically is a forever supply of them.
It's not in the big chips - the processors, th
Re: (Score:2)
This is not true. Embedded processors are at greater than 6 month lead times. Just go on digikey and try to buy 1000 pieces of some low powered STM32 ARM cortex M0 something. There are lots of processors in modern cars. $10,000 dollars of inventory is literally nothing. Nobody cares about $10,000 of inventory. That will keep your line up for a day or two. Nobody buys from digikey in mass production. They are just for prototype quantities. There may be shortages of other IC's also. But I have not noticed pro
Old news (Score:2)
Not every article on this chip shortage is useful information. So what? The world won't stop turning and a minor production speedbump is trivial.
It may result in the rich getting almost immeasurably less rich but has no practical effect in the field since new cars are purely luxury items and people needing transportation buy used. Those prices are slightly high since the economy is booming but that's a minor inconvenience.
Re: (Score:2)
Re: (Score:3)
You don't understand modern supply chain logistics. Parts are made all over the world and shipped to factories on schedules. A single part, if it is truly unavailable and cannot be substituted with an exact work-alike can bring the whole line down. Once the line is down, all the other orders may stop, too. Some of the demand for the final product will be permanently lost because the people will buy something else or change their mind or do without or just won't have the need anymore. It is very damaging. So
too big to fail (Score:5, Insightful)
We should have let GM fail in 2008.
Re:too big to fail (Score:4, Insightful)
Just as we should have let J.P Morgan, Goldman Sachs, et al fail a year later.
Re: (Score:2)
Just as we should have let J.P Morgan, Goldman Sachs, et al fail a year later.
J.P. Morgan needed a bailout. Goldman Sachs did not.
GS was making bets against housing, offsetting most of their losses. They only lost $2B in the last quarter of 2008, a trivial percentage of their capitalization.
GS accepted bailout money since it was cheap capital, but they did not need it.
Re: (Score:2)
Funny how the free market party bailed them out with dirty socialist money.
Re: (Score:3)
So you don't work at GM then? Or any of the myriad suppliers for GM? Or any of the ancillary businesses catering to GM personnel?
How gracious of you.
Re: (Score:2)
Did I create the situation that brought GM to the brink of failure? No.
I'm sure the idiot executives that actually did help the company fail got bonuses of more money than I'll make in a lifetime. Go ahead an blame the folks that caused the problem rather than those that don't want to help bail them out.
You act like GMs market share would just disappear and then there would be a smaller market forever. The reality is that other companies would work to meet market demand. Don't put all of your eggs in th
Re: (Score:2)
Did I create the situation that brought GM to the brink of failure? No.
What you did has no bearing on the concept of being "too big to fail". The reality is them failing would have an impact on you either directly or indirectly.
You act like GMs market share would just disappear and then there would be a smaller market forever.
No, he acts like the hundreds of thousands of people would be out of work, possibly thousands of businesses making the 10s of thousands of specialist parts would go under, and many millions of citizen's retirement savings would be decimated through the resulting upset to their 401k
Re: (Score:2)
Do you even know what you are talking about?
https://www.thebalance.com/aut... [thebalance.com]
Re: (Score:2)
We lost over 11 billion bailing you idiots out.
https://www.thebalance.com/aut... [thebalance.com]
Yep, great investment. Its obviously working out well for the supposed 1M jobs we saved, many of which will be side lined due to problems with order timing. But keep blaming me while Mary Barra makes over $20M this year. Idiot.
Comment removed (Score:5, Insightful)
Weber and SU (Score:2)
Re: (Score:2)
Sure, why not?
It's not like they designed the engine around the precise volume and timing of computer-controlled fuel injection.
Plus, modern emissions and efficiency regulations are totally obtainable with decades-old carburetor designs.
Re: Weber and SU (Score:2)
The emissions would be fine - for about 10 minutes until the catalyst was completely poisoned by all the unburned hydrocarbons and leaded fuel.
Re: (Score:2)
Re: (Score:2)
Sure, if emissions tests only lasted 10 minutes and it was towed to the test site having only just been built or had a new catalyst installed.
Re: (Score:2)
And they can't be registered in any state as an on highway vehicle. GM could put an off-road only sticker on it to get around the laws. And take a 50% hit on retail price as consumers won't be able to get car loans for $50k+ toys.
Failure to plan for supply disruption. (Score:2)
When designing circuits, there are always a slew of different components you can use. Cost optimization dictates which components you use but that doesn't mean there aren't other parts that do the job but might cost a nickle more. What they should have done is designed each modular component to work with different sets of ICs to avoid getting slammed like this. However, when you have MBAs optimizing everything for cost and never planning for anything to go wrong (because plans cost money) then you end up
Re: (Score:3)
One of the main shortage components is microprocessors. If you change out a micro you have to write and test new code. Possibly even re-certify (if applicable). I am guessing you have never actually gone through that type of cycle at a real company? It is not so easy to make sure every component has multiple sources. Especially for processors.
Re: (Score:2)
The shortage is not for ECU processors. The ICs they are short on are for auxiliary components. However, if you knew about ECUs then you would know that the code is mostly portable MISRA C code with a small amount of code for peripheral I/O. Yes, they would have to test and validate the alternative ECU but it wouldn't be a large burden to test multiple since almost all the code is shared.
It is not so easy to make sure every component has multiple sources. Especially for processors.
I never claimed it was, I wrote that they make no attempt to do so.
Re: (Score:2)
I don't think your idea of a "large burden" and my idea of a "large burden" are necessarily the same. But there are a lot of processors in a car. Not just the ECU. No matter what you say or think, there is a substantial amount of effort to switch out one processor for the next and port the code over and re-test. I have never worked in the auto industry but I have worked on consumer electronics. You have to budget some weeks for a change like that. I do design PCB's for a living and work with FW engineers fr
Re: (Score:2)
I don't think your idea of a "large burden" and my idea of a "large burden" are necessarily the same.
Probably not because I put it in the perspective of designing the entire car. Design a handful of modules to use chips from different vendors does take time and effort but it's not a monumental task. Honestly, power and RF systems are far more difficult.
But there are a lot of processors in a car. Not just the ECU.
And they are all a part of modules that speak a common language: CANbus. The ECU is the one that really matters the most though because it's the thing that can trigger a recall. There are plenty of flaws in the computational monstrosities they call "info
Thanks. no confusion with potato chips. (Score:2, Funny)
So one must qualify it as "semiconductor chi shortage".
The fallacy of post-scarcity on display (Score:5, Insightful)
Many of the chips in question get made in places like Indonesia or Thailand. We're not necessarily talking about $100 cpus even $20 cpus. We're talking about pennies-apiece PIC16s or equivalents. I've never seen one that wasn't made in Southeast Asia in over nearly 20 years.
My point is that the scarcity is happening almost a year after the initial round of order cancellations and factory closures on the other side of the world away from the view of many board rooms and media commentators on this side of the world.
The result would seem counterintuitive: we have outsourced "low-value" manufacturing and hyped up the "advanced manufacturing" we did retain. We also hype ourselves up as being a consumer-driven and service-driven economy. We're living the post-scarcity dream, they say. The automation took the jobs away.
And yet here we are: there ain't enough stuff to go around (in a noticeable way) because human beings on the other side of the world aren't working.
Don't underestimate the power of "out of sight out of mind" to cloud people's thinking about both this particular problem and the nature of work, and its intimate relationship with prosperity, the next time you hear a seemingly well-informed and educated individual telling you why outsourcing manufacturing (both clean and dirty, high wage and low wage) can be done without incurring any penalty.
This isn't a left vs right thing btw. Lefties don't like manufacturing because of quasi-religious reasons having to do with environmentalism and righties have few qualms about outsourcing in the pursuit of lower up-front costs.
Milton Friedman went so far as to say that manufactured goods critical to military readiness (his example in the 70s and 80s was steel) "can be stockpiled" negating the need of having a robust domestic steel industry. He was flat wrong then, and any Gordon Gecko type who thinks microchips, or petrochemicals, or computer screens, or batteries, or facemasks, or antibiotics, can just be stockpiled from cheap foreign sources is just as wrong. This being a case of stupid more than evil. I don't think many politicians or cxo types really understand the concept of shelf life of stockpiled materials or the human capital required to utilize those materials even if they are shelf stable.
Re:The fallacy of post-scarcity on display (Score:4, Insightful)
We're talking about pennies-apiece PIC16s or equivalents. I've never seen one that wasn't made in Southeast Asia in over nearly 20 years.
Funny you mention PIC16s, since they are quite specifically made in either Arizona or Oregon. Automotive chips are also made in Colorado, France, Germany, and Italy. But honestly who knows what specific part is being discussed here. These things comes from all over the world. The display driver shortage being talked about last week, they are almost exclusively made in China.
Re: The fallacy of post-scarcity on display (Score:2)
I promise you there are PIC16s made in Indonesia. Perhaps there is or was domestic production as well I wouldn't see ordering through digikey or mouser or jameco back in the day but nonetheless a good number of them are made abroad and the little PIC12 or whatever you find in some key fob remotes could be from anywhere.
Re: (Score:2)
There are lots of PIC16 clones made all over the world. But it is possible for an authentic Microchip Technology company product to come from Indonesia. Looking at the export control data on Microchip's website [microchip.com],
The products of Microchip Technology Inc. might be assembled in any of the following countries: Austria, Germany, Hong Kong, N.T., Indonesia, Israel, Japan, China, Republic of Korea (South Korea), Malaysia, Philippines, Republic of China (Taiwan), Singapore, Thailand, Vietnam, and the United States.
Re: (Score:2)
I promise you there are PIC16s made in Indonesia.
You can promise whatever you want, you just would be breaking that promise (strange use of english, promising something which is false). PIC16s and similar chips of that process are made absolutely everywhere, including the US company's own facilities in the US which are geared for exactly that production process node size.
Doesn't add up (Score:5, Informative)
Or put another way, for a 4-week (1/13 of the year) closure of all plants (not just "some" as in TFA) to cost them $1.5-$2 billion profit, they'd have to have expected an annual profit of $19.5-$26 billion this year.
Sounds like they were on track to lose a ton of money this year, and are trying to blame it on the chip shortage, instead of fessing up to the other factors which are the main cause.
Re: (Score:2)
https://www.thebalance.com/aut... [thebalance.com]
Would Now Be A Good Time? (Score:1)
To repeal absolutely all the business-throttling income taxes and start building things like semiconductor chips in the USA without the bloodsucking income taxes making such things unprofitable? Hmmmm? There were reasons that the Founding Fathers pointedly did not allow income taxes, and this giving of advantage to foreign business was probably one of them. Repeal all the income taxes - personal, corporate, payroll, estate, self-employment, capital gains etc. etc. and replace them with a luxury tax on