Chip shortage? WHAT chip shortage!

The Rocketry Forum

Help Support The Rocketry Forum:

afadeev

Well-Known Member
TRF Supporter
Joined
Sep 21, 2017
Messages
1,888
Reaction score
1,092
“I’ll make them as many Intel 16 [nanometer] chips as they want,” Intel chief executive Pat Gelsinger told Fortune last week during his visit to an auto industry trade show in Germany.​
“It just makes no economic or strategic sense,” said Gelsinger, who came to the auto show to convince carmakers they need to let go of the distant past. “Rather than spending billions on new ‘old’ fabs, let’s spend millions to help migrate designs to modern ones.”​

 

FredA

Well-Known Member
Joined
Jan 19, 2009
Messages
2,300
Reaction score
462
16nm is OLD by PC standards ..... so sure, let the automakers use that old stuff.
16nm, is not a standard process anywhere not even Intel -- who's tooled for it? What's the leadtime to masks?
Also - Intel fab's are usually not building automotive grade parts.
Empty words......
 

MJW

Well-Known Member
TRF Supporter
Joined
Jul 31, 2019
Messages
85
Reaction score
108
Pat has a great point! Auto makers should really invest the NRE to migrate to more modern technologies instead of complaining that legacy capacity is too constrained. Mid decade chips are a whole lot cheaper than cutting edge nodes and there is a lot of manufacturing capacity available to high margin designs like automotive grade on mature processes.
 

Funkworks

Low Earth Orbit, obstructing Earth's view of Venus
Joined
Jul 28, 2018
Messages
2,109
Reaction score
1,949
Have the newer chip generations been proven to be as reliable as the older ones yet? Older stuff tends to be seen as more reliable, just because longer-term reliability is demonstrated. So it's a bit tricky.
 
Last edited:

cerving

Owner, Eggtimer Rocketry
TRF Sponsor
TRF Supporter
Joined
Feb 3, 2012
Messages
4,596
Reaction score
2,057
Not sure that Intel is even a big player in the automotive industry, except maybe in the infotainment systems. The real shortage is for commodity parts... microcontrollers, FET's, drivers, and MLCC capacitors. GPS chips too... that's what's killing us. Covid had a huge impact, primarily on the shipping industry... I was at the beach yesterday and you could see the line of cargo ships stacked up waiting to get to the port to be unloaded. Pre-Covid, that just didn't happen.
 

FredA

Well-Known Member
Joined
Jan 19, 2009
Messages
2,300
Reaction score
462
Unless Intel makes a NEW process geared toward LONG-LIFE, HIGH VOLTAGE, AUTOMOTIVE GRADE parts, you don't want these parts in your cars.

Sure Pat wants to use his excess old capacity....at some point he will have 14nm Fabs go idle after 12 years of hard use instead of the usual four.

Tooling is SUPER EXPENSIVE and LONG LEADTIME. Automotive NRE is better spent on a more targeted process.
 
Last edited:

jderimig

Sponsor
TRF Sponsor
Joined
Jan 23, 2009
Messages
3,646
Reaction score
1,055
Pat knows better than his stupid statement.
 

heada

Well-Known Member
Joined
Jan 18, 2009
Messages
4,442
Reaction score
1,626
Location
Indianapolis, Indiana
I've seen news stories about container ships backing up on both the east and west coasts. I even read where smaller ships are going through the great lakes to get unloaded in Chicago now. All the backups are related to getting enough truck drivers to move the containers out of the docks and not the capacity of the docks themselves. Don't see that getting any better any time soon.

About the auto industry....they want to use what they know and that is out-dated chips from old fabs. They (the auto industry) told the fabs that they wouldn't need much capacity since they thought COVID would put a stop to most sales. The fabs happily changed their plans to making other chips for other customers. Now the auto industry is asking for chips but the fabs are booked for the next 2 years with other customers. They can't get their old chips because the fabs are busy and they can't use new chips because that would involve a redesign and that takes years. Kinda stuck between a rock and a hard place and it's of their own doing by bad forecasting.
 

neil_w

Tenet Rocketworks
TRF Supporter
Joined
Jul 14, 2015
Messages
12,446
Reaction score
5,390
Location
Northern NJ
About the auto industry....they want to use what they know and that is out-dated chips from old fabs. They (the auto industry) told the fabs that they wouldn't need much capacity since they thought COVID would put a stop to most sales. The fabs happily changed their plans to making other chips for other customers. Now the auto industry is asking for chips but the fabs are booked for the next 2 years with other customers. They can't get their old chips because the fabs are busy and they can't use new chips because that would involve a redesign and that takes years. Kinda stuck between a rock and a hard place and it's of their own doing by bad forecasting.
It's not all on the auto industry; there have been considerable supply disruptions. Our own orders never changed, but all of a sudden we can't get parts we ordered more than 6 months ago. The whole thing is a mess and it's going to take a while for it all to unwind and get back to normal. My own suppliers are suggesting that things start to get better in the second half of 2022. That makes for some rough sledding between now and then.

We are doing some urgent redesigns right now to swap in chips we can get for those we suddenly can't. This is not fun.
 

afadeev

Well-Known Member
TRF Supporter
Joined
Sep 21, 2017
Messages
1,888
Reaction score
1,092
We are doing some urgent redesigns right now to swap in chips we can get for those we suddenly can't. This is not fun.
That's precisely what Tesla did:

Given an average automaker's level of component outsourcing and multi-year supply chain management practices, something like what Tesla did would take others 1-2 years to RFP out, and another 2-3 years to inject into ongoing production flows. Optimistically.
 

OverTheTop

Well-Known Member
TRF Supporter
Joined
Jul 10, 2007
Messages
5,907
Reaction score
3,443
Location
Melbourne Australia
Just flipping a design to a smaller fab is not as simple as it sounds. There are not insignificant costs of getting masks updated and manufactured to get the whole design ported across, and then the "die shrink" causes other problems. We did it recently on a CCD imaging device and it was millions of dollars to swap fab houses and re-do for the different size wafer.

Die Shrink is what happens when a manufacturer decides that their new process (usually smaller so they get more parts on a wafer) is good to use for their old devices. A new device with the same voltage and current characteristics is easy to make in the newer fab so they put the new, smaller, chips into existing packaging and call them "identical" and don't even change the part number. Their parts are cheaper and that is what matters to them.

Where this is a problem is that the smaller devices have less capacitance and sometimes inductance so can switch much quicker. These fast-changing voltages cause internal EMC (electromagnetic compatibility) problems in equipment. I know of two of our products that have suffered this problem over the years. Shrinking the die on a microcontroller, and other parts, will have similar ramifications unless significant design changes are made. That all costs money and the automotive industry have a difficult time parting with that.
 
Top