And this is why AMD and Apple are eating their lunch now. I'm super happy Intel is going to build a next gen fab though, mostly because I'm concerned about the imminent threat of China taking over Taiwan and TSMC with it.
Apple, on the other hand, has put up a bit of a fight. They bought out the entire 5nm node from under the rest of the industry, for better and worse, and then they changed the game again by putting the ball in their court (mobile chipsets). At this point it's hard to even say they're competing against AMD and Intel; the two are chasing entirely different markets at this point. When they do butt heads though, watching the sparks fly has been spectacular. Despite being on decidedly worse silicon and a veritably slower ISA, Intel has managed to refute every one of Apple's offerings with a decidedly cheaper processor. I really do look forward to their response to the M1 Ultra, because I feel like the chip wars have only just begun. With Pat Gelsinger at the helm, Intel has set a warpath through the industry, and their recent performance metrics seem to cement their intent on being the best in the biz again.
Godspeed to all you chip manufacturers out there.
Core design is astronomically expensive so given the opportunity to build only one core (or two in the case of big.little) it seems reasonable that Apple's perf/W would be superior since they have had a very pressing business need for it. Add on that they have so far been built on one to two nodes ahead.
The point being, I think if Intel still can't get close to Apple in 2-3 years then there is a bigger problem, but as of now I think there are still some reasonable explanations as to why their current products are behind that don't come down to technical inability.
At this point the only way out for us, from both iOS and Android, is if they both miss the next big form factor, maybe smart glasses and AR/VR. Basically, the next OS.
It is using cash as buy back while not giving enough budget to R&D or product improvements. Apple didn't do that. They invested accordingly and still had too much money and didn't know what to do with it. Although one could argue they could invest and bring even more value to their customers. But Tim Cook think those were better for shareholders.
Compared to Intel, they stopped investing in foundries. Partly because of their 14nm were over provisioned and partly because of their broad. Along with their 10nm failures. It really was a perfect storm.
Isn't TSMC building infrastructure out in Arizona? It might not be full featured but it creates a launch point for protecting a world wide asset. https://tsmccareers.com/tsmc-arizona/