The latest numbers on the microchip shortage: Worst-case estimate grows worse(autonews.com)
autonews.com
The latest numbers on the microchip shortage: Worst-case estimate grows worse
https://www.autonews.com/manufacturing/latest-numbers-automotive-microchip-shortage-1
8 comments
I wonder how many perfectly usable/recyclable chips are abandoned in landfills worldwide.
They aren't perfectly usable if they've been used. Semiconductors are kind of like toothpaste - if they've been used, you don't really want to buy them.
This is largely because the cost of any individual component is so much less than the cost of a finished assembly, or the cost of a defective assembly. Inserting used components into your manufacturing process (unless they're very high value or rare parts), is asking for trouble.
These components need to be stored properly to be effectively assembled, and you'd never be sure of why the used one were thrown out. Especially as many types of damage or degradation would either require very substantial and specific testing rigs, or else be challenging to identify in the first place.
Putting them inside a vehicle isn't a viable option from an safety or economic perspective.
This is largely because the cost of any individual component is so much less than the cost of a finished assembly, or the cost of a defective assembly. Inserting used components into your manufacturing process (unless they're very high value or rare parts), is asking for trouble.
These components need to be stored properly to be effectively assembled, and you'd never be sure of why the used one were thrown out. Especially as many types of damage or degradation would either require very substantial and specific testing rigs, or else be challenging to identify in the first place.
Putting them inside a vehicle isn't a viable option from an safety or economic perspective.
To pile on to this thread on how hard it would be
My understanding is that most of the cost of analog chips is the time in the testing machines to characterize and bin them. I assume the same is still somewhat true for CPUs, etc.
If you were somehow to acquire old CPUs that were up to the task in old modules, you'd have to remove them, clean up the leads or solder bumps, and then test/bin them before you could trust them enough to send to manufacture.
My understanding is that most of the cost of analog chips is the time in the testing machines to characterize and bin them. I assume the same is still somewhat true for CPUs, etc.
If you were somehow to acquire old CPUs that were up to the task in old modules, you'd have to remove them, clean up the leads or solder bumps, and then test/bin them before you could trust them enough to send to manufacture.
Which, do you know how much way more advanced science go on inside of those chips in order to create them? The issues you mention are a logistical nightmare, and due to the costs involved it currently isn't lucrative to reclaim them from built devices. Let's say that someone was willing to pay $1,000 for one entirely boring, through-hole 74-series logic chip. Would that change the calculus on it being "too hard" to put them back in the supply chain? Junk VCRs in the garage would suddenly be potentially worth a several thousand dollars each! At a thousand dollars per chip, it would be worth my time,
personally, to do the testing and clean up, and for someone at the other end to verify I've actually done that, and that the chips are functional.
That astronomical price is obvious fantasy, but it's not too difficult as a task, it's that capitalism can't and won't care about the environment until after it's too late (which, it might already be).
That astronomical price is obvious fantasy, but it's not too difficult as a task, it's that capitalism can't and won't care about the environment until after it's too late (which, it might already be).
You know maybe the current batch of running chips can't be recycled. But I wonder what would need to change in the manufacturing and installation of them to make them reusable without it adding to much cost?
As a species we need to stop this throw away culture.
As a species we need to stop this throw away culture.
We would need to accept larger, thicker and more unreliable devices, I guess, unless there's an alternative to DIP or PGA/LGA sockets. The removal, testing and binning process could largely be automated, but just disassembly is going to be hugely labor- and capital-intensive.
"The projection is up by nearly 55 percent from AFS’ estimate two months ago.The auto industry’s struggle to acquire adequate supplies of semiconductors continues to go badly, with worldwide tallies of production cutbacks significantly worsening this month.
According to a new estimate by Auto Forecast Solutions, which has been tracking the supply chain crisis all year, automakers have now eliminated a total of 8.2 million cars and trucks from their factory schedules because of microchip scarcity — an increase of approximately 888,000 vehicles from AFS’ estimate just one week earlier.
AFS now forecasts that the industry ultimately stands to lose 9.4 million vehicles world wide because of the chip shortage.
That worst-case scenario is up by nearly 55 percent from AFS’ estimate two months ago.At the core of the shortage is a growing competition for available manufacturing capacity dueto increased demand for chips from makers of consumer electronics, cellphones and computers. But in recent weeks, the supply strain has been exacerbated by work force issues due to the resurgence of COVID-19. Chip production lines in Asia — particularly in Malaysia — have been forced to shut down in response to rising coronavirus infections there"