I meant if China ends up being the only other country than Taiwan with domestic chip production.
Oh they definitely aren’t. Singapore, Korea, Germany, UK, Japan, and the US also have significant fabrication capacity.
The problem is that most of those fabs aren’t anywhere near a 14nm process, much less 7nm or less competing with TSMC. Only Intel and Samsung even come close. TSMC is just that good.
When I was at Intel by the mid to late 90s we were almost a full generation ahead in process technology. TSMC is there today, but I think people underestimate how easy it is for leaders to lose their top dog status in the tech world.
So when do TSMC reveal the alien tech from another dimension they’ve had crack teams of Gatehopper Operatives stealing for the last five years has finally been traced by the Dnthrakk Syndicate, whose Barrier Piercer warships are already beginning their assault on the fragile walls of our so called reality?
It’s funny you made a joke I was thinking about very strongly about making, but didn’t because adulting. And yet, here we are.
Yes, they are very good. Almost… too good.
As I understand it, part of the reason is that Intel tries to go whole hog on every new process, whereas TSMC treats process jumps incrementally, like “okay, half this chip will be 10nm and parts of it will be 7nm, because trying to go all 7nm at once is really, really hard.” They baby-stepped it, learned, and refined the process further.
Intel didn’t want to baby-step it, and that’s just really hard.
Insightful analysis on Pat’s return to Intel from my friend Rob Enderle
Real Answer: Profit and greed.
Be a really weird decision if true. As their 11th gen chips are only due out in a few weeks.
Intel has a lot of catching-up to do. Apple made Intel look absolutely obsolete with their first-gen M1, and there are faster Apple chips coming very soon. There’s probably a ton of pressure coming from intel’s partners to respond.
I do wonder how much the Dells and HPs of the world care about what Apple is doing. I guess you should always be paranoid, but it’s hard to imagine a big surge in MacOS no matter how performant their CPUs are.
Still, a competing product that has more power than your desktop and better battery life than your best laptop just makes you look terrible in comparison. Consumers want to know why they can’t have that on the PC side.
Also, substitute partners with “shareholders.” Intel’s been struggling for a while, and everyone knows it.
Call me skeptical, but Apples early progress here is around building out an organization and manufacturing process around early engineering. I’m not knocking what they have done. What I will knock is they aren’t old hat at it, yet.
A ten, twenty or even fifty year track record doing things like this brings both experience at the repeated process of the engineering cycle, but also a track record of being able to pivot and move as an enterprise to new things.
If we were to point at the long term successes of Apple, I would not put chip production as one of their strengths, and certainly not as high as marketing or product iteration.
So, a new player to the game, sure. Intel is always worrying about new players to the game. So is every other player here. The larger the number of players in the market, the harder it becomes to gain, or hold market share.
I was with you up until pivot. Intel’s experience is in entrenching and holding onto a de-facto monopoly, not in pivoting. All their attempts to expand beyond desktop/server CPUs and associated motherboard chipsets have failed haven’t they?
I don’t disagree with your central point that Apple’s focus is on Apple products so they aren’t a direct threat to Intel. But it’s an example that should be worrying Intel, because what if AMD does something similar (or licenses Apple tech) and gets Microsoft on board?
Good point, pivoting outside of their standard market is a weakness. To further that, tons of new chip technologies have led to new markets they don’t even compete in. (Does intel even have any significant mobile market?)
I guess when I said pivot I’m thinking of all the changes in chip design, shrinking the process, etc. As an example, if you have a good understanding of how Tesla is making cars and batteries right now you could implement one hell of a competition with them with much smaller necessary innovation. But 5 years in, now you have to go through that iterative process of innovation, and that’s all you (unless you spy/cheat/steal your way to it.)
That makes me wonder how well Intel guards its secrets.
Is that an intel only scenario though or is that an x86 as a whole scenario (i.e. includes Ryzen)?
This is odd to me. How long do you think Apple has been doing this? How much of a track record do you expect to see and on what axes?
For laptops/desktops, only recently. For mobile/ipad, since 2010. They got the bulk of a lot of their start from a purchase in 2008 and another purchase in 2019 from Intel itself (not laptop/desktop.)
I’m not throwing Apple’s other manufacturing history under the bus, we’re talking high end chipsets here. Intel has just been doing it much, much longer (for this specific market.)
Sorry @ReptileHouse I missed the last part of that question. I’m not an engineering exec, I would expect a few product -cycles- prior to getting good feet for it as a track record. No different here. I also am only skeptical of how they will do, not saying they will fail.
Late edit: I would also say they are poised for success if they can continue the iteration we’ve seen them do for mobile/tablet chips.