By 2030, Intel CEO predicts chips would account for 20% of car costs.

That's a daring prediction. But not so far fetched.
But that means that a mechanism would be rendered useless without software/chips to operate it. And mechanics will have to adapt. Or their branch will be also reduced to simply checking software updates?

 

For an industry that routinely would shun a supplier for raising the price of a 20$ board by 1$ this is beyond far fetched. If you look at a car built 2 years ago without any context, judging only by the tech inside it, you'd assume they built it 10 years ago.
Arguing that the chips alone are going to be that expensive is an overly optimistic statement even if you assumed full self driving would be the standard by then. The only way I could see this happening is automation taking over and cars going down in price by at least half if not more.

 

What you say is true for petrol and diesel cars, but by 2030, most cars will be electric, in which computer chips plays a MUCH bigger role.

 

LOL, most cars yea right buddy. the car industry takes decades to change.

 

On the side of these news, insurance and security regarding cars will evolve in something more IT field fused into automotive industry.
Car hacking will evolve, cyber security will be in demand.

 

There's a ban of non electric cars in place for 2030.

 

What @PrMinisterGR said.

 

Quite sure thats new cars though, the existing diesel/petrols will still be there for 10-20 years after that, only lack of infrastructure would speed that up.

This news is only to keep Intels shareholders happy with the potential.

 

In a profitability standpoint, it is very farfetched. Back around 2014 when economy cars started including touchscreen head units, they'd use the cheapest crappiest processors so they could, until people started complaining that using their phone was faster and easier (which has its own implications of issues). Tesla has invested in making their own processors, likely to reduce cost.
The way I see it, Gelsinger is envisioning a monopoly of a future where he somehow manages to outpace Nvidia. I'm sure it's all stuff to appease investors, as they like the idea of something raking in so much cash.

We already have cars with all the technology necessary to drive by themselves; they just need more refined software. For head units, the processors are for the most part are powerful enough to give a smooth experience. Of course there is room for improvement and advancement, but I'd say we're 80% of the way there when it comes to computerizing vehicles. It is just preposterous for the computer hardware to make up more than 20% of a vehicle's cost. Even 10% sounds absurdly high.
Remember too that EVs are the primary future of vehicles. We need efficient hardware, not necessarily faster. Again: we today already have the technology we need, we just need to refine it.

 

In Intel's perfect world where they can continue driving up the cost of sand so that's it's absurd.

 

Silicon supplies may be come stressed. There are signs it is already happening.

 

If you'll notice, Intel has become very fond of predicting what will be happening in the next 20-30 years, lately. But, just three years ago, Intel did not predict the rise of AMD's Ryzen and EPYC, oddly enough. It's worrisome when a company has almost nothing to say about its immediate product line, but starts spouting off predictions that are decades away in the future. Intel needs to concentrate on the next 2-3 years, imo, and leave future decades to themselves. Often, companies do this to try and take the focus off of what they are not accomplishing today.

 

hehe...that;s been said and asked before when ECU's came out.

but hey this is already impacting supply and fab queues
and mr. Gelsinger unlike others is in a position of knowledge.
idk why some are scoffing at the numbers being tossed about. they've been thoroughly vetted and are part of Intel's strategy moving forward.
no-one is talking about one massive chip that's 20% of a car. we're talking about dozens of custom silicon chips and often two or more motors in a drivetrain, not to mention the integration of sensors and controls.

electric cars are inherently simpler than internal combustion, so certain types of mechanical jobs would be unnecessary, while others are just as important. idk about fewer mechanics until service is completely automated which isn't happening soon.

 

Thats the scary bit.

 

Considering the whole thing is about new cars it does kind of apply....

 

more polysilicon fabs have already broken ground in the US (Intel, UMC, GF for sure) or are about to under Biden's supply line reinforcements. why wait when you can have the cost partially subsidized?
since we are talking wafers we only have to worry about purity, not production nodes.

but we all know TSMC is going to build a "mega-fab" in the US and it's in the final stages of decision making. Intel is making such a massive investment that they're abandoning fabbing lower margin items. UMC is going gangbusters and GF has gotten a lifeline.

the last two years has taught the industry a generation's worth of lessons.
but this stuff (fabrication) take time to build out, but both Intel and TSMC have expanded production and built more fabs. by 2030 all of this will be a roadbump in the rear mirror.

 

I prefer my car to be simple yet powerful.
All this connected crap isnt necessary.
All I need is a car that will do what I want when I want.

Basic computer systems ok.
Anything beyond that, get f****d.

 

For only new cars?

Will I still be able to hunt for my perfect 1999 Subaru Impreza WRX STI 22B ?

 

I think you will have that choice. As with combustible engine vehicles, all budget levels will be catered for in electric cars too. From super-smart to dumb cars, they will all be there.