Or eat crow and license it from someone else, as GF did from Samsung for 14nm.Spread some of the R&D cost around, so they don't have to eat it all.
Or eat crow and license it from someone else, as GF did from Samsung for 14nm.
I didn't say it was viable, point of fact Intel eating crow on this front is as likely as me swallowing my own foot.People should stop bringing this up as a viable strategy.
I'll say the same thing as I said before, people need to consider the wider implications for the US with respect to Intel's Fabs.
Indeed, the market for 0.1+ micron process tech shows that fabs definitely don't have to be leading edge to be viable.The existing fabs have a lot of value, they could go the GloFo route and just not stay leading edge on their fabs.
This is something they could have done while being real undisputed leader. (Do they still tout self-praises for "process leadership"?) I doubt their handling of 14nm and the never ending 10nm saga instills confidence in their capability of introducing future nodes. 7nm is make or break in that regard (and that's late already as well).They may have to come up with a real foundry business model, instead of fiddling around the edges. Spread some of the R&D cost around, so they don't have to eat it all.
Because Intel makes more chips than all phones combined? They don't. Also their current fabs won't dissolve into thin air. All it would do is potentially making products based on bleeding edge nodes more rare and expensive ("special") in consumer product, at least as long supply is lower than demand.If they went fabless, the world would slow to a crawl. That's how much volume they handle. The top 5 fabs would be so overloaded no one would get their phones or chips.
This.This is something they could have done while being real undisputed leader. (Do they still tout self-praises for "process leadership"?) I doubt their handling of 14nm and the never ending 10nm saga instills confidence in their capability of introducing future nodes. 7nm is make or break in that regard (and that's late already as well).
I didn't say it was viable, point of fact Intel eating crow on this front is as likely as me swallowing my own foot.
The existing fabs have a lot of value, they could go the GloFo route and just not stay leading edge on their fabs.
If Intel's Fabs basically capitulate ala GF than the US in effect would be ceding position against China. There is serious implications with this compared to GF. You'll see the US government come along the lines of too big to fail should that should hypothetically happen.
Because Intel makes more chips than all phones combined? They don't. Also their current fabs won't dissolve into thin air. All it would do is potentially making products based on bleeding edge nodes more rare and expensive ("special") in consumer product, at least as long supply is lower than demand.
If Intel hypothetically wanted to go this along route they'd be better off hiring talent away from TSMC/Samsung. Much like what Samsung did with TSMC, and what China is now also doing.
Chinese fabs are effectively 2 generations/nodes behind the current, they are only just hitting 14nm finFET now for logic.If Intel's Fabs basically capitulate ala GF than the US in effect would be ceding position against China.
Even if they do catch up, at best they will be stuck playing the catch up game as TSMC and others once were themselves.If 7 nm is a flop, Intel is pretty much going to be unable to be able to catch up to Samsung and TSMC anyway.
Hmm.. Swan said that they expect the 7 nm GPGPU in Q4 2021. If they are only expecting to be what sounds like the equivalent of the foundries' risk production ability/levels, that makes it tough to think they will be able to move that quickly to bring up other products other than to satisfy Aurora's needs. Kind of rough when you consider that TSMC is saying they expect to go volume production with 5 nm in a couple months.
Chinese fabs are effectively 2 generations/nodes behind the current, they are only just hitting 14nm finFET now for logic.
Even if they could catch up in tech, they would need to signficantly expand their fab capacity to match international competitors in order to supply the demands of big fabless companies (AMD/nVidia/Apple/Qualcomm) on top of their domestic market.
If Intel 'capitulate' as you say, they would be behind TSMC and Samsung, but still in front of China - believe it or not semicon IP is guarded much more closely than any US military tech secrets (of which China are admittedly good at pilfering or copying). After the Samsung/TSMC mess you can bet whatever precautions already existed will have been increased, and then some.
I do not expect China to surpass the current industry leaders anytime soon without some insanely bold moonshot maneuver to leapfrog them with new transistor devices and/or materials.
If anything they might leapfrog the rest of the industry in one specific area by investing in new memory deeply, as worldwide industry players are very conservative in investing in non-Flash technologies if I have read the situation clearly.
It's not like China needs to manufacture products for Apple, Sony, MS and who knows how many others, but they do - it's not like China will collapse in on itself like a deranged hermit the moment it has technological parity with the outside world, profit is profit - if they can fab Axx or SD 8xx at quality for cheaper than TSMC or Samsung, you can bet Apple and Qualcomm will want to buy from them, at least for chips not sold to the US in any case.It's not like China has to appeal to foreign markets. Their domestic market is plenty big enough to develop both competitive and independent technology ...
Great point and a bit scary, IMO. Intel has DG1 back from the fab already, two years before launch - making it a very curious pipe cleaner.
It's not like China needs to manufacture products for Apple, Sony, MS and who knows how many others, but they do - it's not like China will collapse in on itself like a deranged hermit the moment it has technological parity with the outside world, profit is profit - if they can fab Axx or SD 8xx at quality for cheaper than TSMC or Samsung, you can bet Apple and Qualcomm will want to buy from them, at least for chips not sold to the US in any case.
Also, merely having the desire alone isn't enough to enable a catch up of this magnitude - for TSMC and Samsung it required Intel to stumble in a truly spectacular manner for the first time in basically forever on process tech.
Mind you, a Chinese catch up is more likely to happen as R&D of new nodes becomes increasingly difficult and drawn out for all.
This is inevitable after previous US restrictions on export of Intel and AMD product for their supercomputers.Could you imagine how devastating it would be to Intel to be banned and replaced with a chinese alternative ? (~20% of Intel's revenue comes from China)
This is inevitable after previous US restrictions on export of Intel and AMD product for their supercomputers.
I can only imagine what a potential loss to their bottom line it was.
There's still India, a huge population and an increasing tech sector.Intel parts are still available to chinese consumers ...
Them being replaced could make or break Intel's logic fabrication future altogether ...
Ah. Thanks. Guess I need to learn a new Intel dGPU nomenclature.IIRC DG1 is the first gen discrete GPU, ie: Arctic Sound and not the 7 nm GPGPU.
There's still India, a huge population and an increasing tech sector.
What, steal India?Of course providing that China doesn't steal that as well ...