Vehicle Semiconductor Chips 101

This is a reprint of a great primer on vehicle semiconductor chips -- by a member of sister-site Bronco6G.com


Words by ocbucks1:

Administrator said:

This is a reprint of a great primer on vehicle semiconductor chips -- by a member of sister-site Bronco6G.com


Words by ocbucks1:

There are a lot of gear heads here that can rattle of how engines and drivetrains work and how to make many modifications on our Broncos. But they might struggle to describe what a semiconductor does and why is there a shortage. I had a reply to a post that said "I know just leave them out". Not​

​

First full disclosure I like this stuff and I do have an engineering degree but I am not am expert but thought I would provide a basic overview since this is such a big topic now.​

​

• What does semiconductor mean?Metal is a conductor of electricity and glass is a insolater (does not conduct). A semiconductor can be either. Weather electricity will flow through the semiconductor or not is changed by energy. In electronics, the energy is electromagnetic fields produced by transistors. The transistors and software, for example will dictate off or on , up or down, locked or not locked and a lot more detailed decisions about when to open the circuit.
  • How many semiconductors are there is a car? Overall the world uses one trillion semiconductor chips a year! You probably know they are in every electronic device. The auto industry uses about 10% of the total. In one car there can be 3,000 chips.
  • How are they manufactured? This is where it gets very complicated. A chip takes 2 months to make and goes through 3000 manufacturing steps. They are made on a silicon wafer. The silicon wafer is the "semiconductor" material. This goes through a dozen or more machines before complete. It takes massive amounts of pure water to purify equipment. One factory reported they used 60,000 tons of water a year. Because of the requirements and complexity, there are only a handful of factories in the world. The order lead time prior pandemic was 10 weeks and is now roughly 18 weeks.
• Why is there a shortage? There were multiple factors that came together to cause the current shortage.
  • -First at the pandemic auto maker reduced orders expecting a reduction in car buying. At the same time there was a huge increase in buying of other stay at home electronics. For a somewhat brief period, the chip manufacturers focused on providing chips to other industries and auto orders were now behind
    • -Then the strained relations and trading with China, where the biggest chip manufactory was located ,reduced supply. But this was not really a direct hit to auto.
    • -A majority auto chips were made in Japan and in one factory. This factory had a severe fire. This is what hurt Ford the most.in getting behind on orders. This plant is now back but not full capacity.
    • -Another major factory in Twain was shut down due to a drought there and they did not have enough water.
  • What did Ford do? Contrary to popular opinion here Ford did make some moves. They made an exclusive partnership directly with a foundry, global foundries, This vertical integration is not common in auto. Overall they are in slightly better or the same position as most auto makers. (Tesla did do some creative software engineering to make better use to available chips.)

Want more Semiconductor Video.​

​

I hope that helps. We know the track record with the Bronco and delays. But this one is truly farther reaching than Ford and the bronco. We have less of a gripe in my opinion.​

​

​

​

Click to expand...

VERY useful information… thank you for taking the time to write this up and share it.

Administrator said:

Why is there a shortage?

Click to expand...

Adding to what is listed - Texas chip plants were closed and lost product due to the freeze in 2021.
https://www.statesman.com/story/bus...exas-freeze-cost-samsung-millions/4891405001/

While its true that we had the perfect storm in terms of chip shortage with the fire in Renesas plant in Japan, the blackouts in Texas, and on top of all that Covid, the shortage of chips for automakers was something that did not happen overnight, and car manufacturers have been very slow to adapt to new times

Most chips being used on automobiles today are obsolete, most use 45 nanometers or even 90 nanometers, where things like computers and phones have moved to <10 nanometers.

Nobody is going to invest on increasing production of old technology, car manufacturers need to update their designs to utilize latest chip technology and better integrated designs, and for the future of EV's even consider "producing" their own chips or doing specific partnerships to insure they can meet demand

JcMarin said:

While its true that we had the perfect storm in terms of chip shortage with the fire in Renesas plant in Japan, the blackouts in Texas, and on top of all that Covid, the shortage of chips for automakers was something that did not happen overnight, and car manufacturers have been very slow to adapt to new times

Most chips being used on automobiles today are obsolete, most use 45 nanometers or even 90 nanometers, where things like computers and phones have moved to <10 nanometers.

Nobody is going to invest on increasing production of old technology, car manufacturers need to update their designs to utilize latest chip technology and better integrated designs, and for the future of EV's even consider "producing" their own chips or doing specific partnerships to insure they can meet demand

Click to expand...

I've read and passed on the opinion that there is a "demand" side to the chip shortage. I have no idea of its validity. e.g. "BLAME IT ON THE BITCOIN MINERS"
Spiking development of "bitcoin miners" (huge collections of computers that "create" cryptocurrency wealth) has contributed to the global chip shortage. They use CPU and GPU chips which may or may not be different from the chips used in the auto industry? Your comment suggests that cryptocurrency miners are not in competition for microchips with automakers. Is that an over-simplification?

Also, in 2021, Tesla produced >800,000 BEV while Ford produced about 65,000 MME with "thousands" still waiting for chips. It suggests Tesla has better microchip supply lines or better foresight to deal with competition for chip supplies. Some automakers are already building their own plants to assure microchip supplies. Ford has announced they will "cooperate"' with a chip supplier to improve their sourcing. Don't have a clue what that really means to a 2022 MME delivery that was ordered in 2021.

Well I think TESLA suffered the chip shortage as well, they just have advantages in several areas compared to other manufacturers:

1) Do bare in mind that although the production of EV is as you point out much greater for TESLA, thats not true on overall production, while TESLA produces 800K a year Ford is closer to 2MM, (Toyota to 7MM) and although ICE use an order of magnitud less chips (~300 vs 3000) they still use a lot

2) TELSA has a more integrated design, less parts and more modern platform in all the electronics, controllers and such. There is an interesting Munro video comparing ID4, MME and Model Y

3) TESLA is a Software company as much (or more) than a Car manufacturer, so they were actually able to adjust their software to bypass or even use other chips depending on availability

4) They are willing to do things other manufacturers will never do, like deliver a car without a part to be installed later or remove a function all together if its missing chips (like passenger lumbar support)

I think Ford is learning from all this, and beginning to simplify their designs (removal of kick-open trunk anyone?). and moves like the partnership with the foundries will also help, its just going to take a while before things settle again and even more time until they catch up to TESLA

Btw. One area where TESLA definitely has more chips is in the paint job they deliver.... thats why I prefer my MME even with legacy platforms than the Software-Gadget-Car from TESLA

And, on it goes

https://www.autonews.com/manufactur...up-production-week-due-semiconductor-shortage

This was an interesting piece from a tech podcast I listen to about how the current war in Ukraine could impact the chip shortage. Audio, text, and source below.

“There are also going to be hits to the supply chain. The Conversation notes that Russia and Ukraine are key exporters of neon, palladium and platinum which are critical to chip production. About 90% of neon used in chip lithography comes from Russia and 60% of that is purified by a company in Odessa, Ukraine. Alternative sources would require significant long term investments. Chip makers are estimated to have about 2-4 weeks of inventory on hand but disruptions that last longer than that would cause another stress on chipmaking and prolong shortages.”

Source: https://theconversation.com/amp/five-essential-commodities-that-will-be-hit-by-war-in-ukraine-177845

Interesting . . . . ?

I know that we waited 7 months for our MMe and I was told it was because of a "chip" shortage. I also know that in 7 months, not once did Ford publish and official announcement regarding a chip shortage as the reason for the production delays. All of the information I received was from other, 3rd party sources, i.e. news/industry media, blogs, social media, but never an official statement to the consumer, buyer, customer. I thought this was a poor PR strategy on the part of Ford, it was as if they were hiding something. As a result, I no longer trust Ford to tell the truth.

Regarding, the current ongoing chip issues, I am curious; Do all of the global automakers use proprietary chip designs or are they all using a universal automotive chip? I am suspect, everyone has gone proprietary and that will add to the delays. Wouldn't it be better if the chip for a phone, computer, farm tractor, etc. could be used a car/truck? Instead of everyone having their own special design? Look at laptops/computers. Many components are "standard" just the firmware is custom.

I am sure some "engineers" will contend it is better to have proprietary chips. However I don't see it. You can expand your "RAM" in many computers just be buying a "ram chip" that fits in many different computers. It is cheap and easy to do.

Wouldn't universal chip designs take some of the pressure off the chip supply chain? I understand the logic: A chip specifically designed for the car manufacturer creates a higher profit margin scenario, but it also increases supply chain risk. A single design, single source strategy also reduces negotiation leverage.

Just a thought.

There are both factual and grammatical errors in that post. I leave the search to each reader. As an EE for over 47 years, and having taught Electronics for ten, it was obvious to find some errors.

JcMarin said:

While its true that we had the perfect storm in terms of chip shortage with the fire in Renesas plant in Japan, the blackouts in Texas, and on top of all that Covid, the shortage of chips for automakers was something that did not happen overnight, and car manufacturers have been very slow to adapt to new times

Most chips being used on automobiles today are obsolete, most use 45 nanometers or even 90 nanometers, where things like computers and phones have moved to <10 nanometers.

Nobody is going to invest on increasing production of old technology, car manufacturers need to update their designs to utilize latest chip technology and better integrated designs, and for the future of EV's even consider "producing" their own chips or doing specific partnerships to insure they can meet demand

Click to expand...

This is a distortion of the facts. While there are semiconductor applications that require the latest technology. Most applications do not need super high component densities. Manufacturers are more than willing to serve these profitable markets. It would be a mistake for a company to implement a new more expensive chip that has both higher cost and complexity and lower reliability in a situation that does not require it. When designing a product, you typically want to use the most reliable, least complex, and lowest cost component that will do the job.

efisher said:

This is a distortion of the facts. While there are semiconductor applications that require the latest technology. Most applications do not need super high component densities. Manufacturers are more than willing to serve these profitable markets. It would be a mistake for a company to implement a new more expensive chip that has both higher cost and complexity and lower reliability in a situation that does not require it. When designing a product, you typically want to use the most reliable, least complex, and lowest cost component that will do the job.

Click to expand...

While you are right in that Car manufacturers don't need the more complex chips and the current ones are tried a reliable, the problem is that they are more than 10 year old technology and the rest of the industry (which accounts for more than 90% compared to automobiles) has moved to newer technology. Its no longer desirable for semiconductor or chip manufacturers to maintain the infrastructure needed for these obsolete technology

Automotive manufacturers dread the idea of spending millions on changing their designs and testing and certifying the new chips, but as the chief of Intel (Pat Gelsinger) put's it

"Rather than spending billions on new 'old' fabs, let's spend millions to help migrate designs to modern ones."

I'm an executive at a company that manufactures integrated electronics into electromechanical devices so I'll give you guys my observations from an insider point of view.

The electronics industry is highly segmented internally. That is to say that we build designs that are often fragile when it comes to parts shortages not because other parts aren't available, but because of arbitrary design decisions. For example we used an ethernet connector made by a Japanese company that had a shortage from one of its suppliers (plastic resin I believe). You would think, no problem, we just switch to another supplier, but this was a "low profile" connector and was 1.5mm too small, and we didn't leave enough space on the board for the replacement part (8-12 weeks re-spin/re-layout).

We are often told 1 -2 weeks before the parts run out. Often the electronics industry runs on very short lead-times. We don't warehouse components (although that is changing....That's not helping). For example, I'm sitting on 40k of a particular part that I am not using right now because my primary supplier had a shortage and then subsequently looked like they were getting their act together. In fact, I was just telling our supply chain to sell the parts on the open market, when suddenly they told us that we just received our last delivery until November (I guess we're going to use the alternate part). Other companies are doing the same thing. Buying up supply of alternative parts while praying that suppliers don't cancel.

One final example is power regulators. These component are very generic and help with power regulation throughout the board. All of a sudden our main supplier told us "no more" regulators. So we built 6 different versions of the board for multiple alternative supplier so we could shift the supply chain whenver we needed to. Normally board designers HATE designing for multiple parts. It adds complexity and potentially impacts reliability.

So my point is that the supply chain, design and inventory models used in our modern electronics industry are not well suited to disruption of any kind. and we've seen wave after wave of disruption. It isn't as simple as "just build more chips". It's a combination of more capacity, distributed capacity, changes in design techniques, and possibly a return to more warehousing of components.

Sorry for the book. Hope this gives a sense of what we're dealing with.