TSMC to Begin Production of 'iPhone 8' and 'iPhone 7s' A11 Chip in April

[mschmalenbach]

Holy crap we are getting close to Moore's law. I feel like these chips will keep getting smaller for a few years, then a little thicker with newer battery tech helping to cover the increased power budget, and then we'll be forced to move on to quantum computing. I hope Apple is doing research on this. They have one of the best chip design teams in the world, but that will be completely irrelevant in 10 years. We know that Google has a quantum team. It would be pretty wise for people to research and invest in quantum computing companies.

We're already at the point where the cost per transistor is now going to increase for the fist time ever...

Creating the mask sets for a 16, 14,11 or 10nM process costs 10s & 10s of $millions. Add in the crazy cost of the fab equipment etc, and the fact that yields are off and end user volumes are in very low single digits for growth, and these devices are going to really start costing... and at some point people may just not invest as much in new fab equipment for a while... development work will be heavily focused on improving yields and getter by other costs down - which all takes time, and these things take 12-16 weeks to fabricate...

 

[macduke]

TSMC already announced plans for a $16 billion facility for producing 3nm and 5nm chips by 2023. Fun times!

That's crazy. I remember reading somewhere that 1nm is the limit for silicon, so my 10 year prediction is probably pretty close. We'll need to figure out something else, whether it be graphene (the magical fantasy material), quantum or whatever. You can't just start splitting atoms. I mean, you could, but it won't end well.

 

[Jnesbitt82]

That's crazy. I remember reading somewhere that 1nm is the limit for silicon, so my 10 year prediction is probably pretty close. We'll need to figure out something else, whether it be graphene (the magical fantasy material), quantum or whatever. You can't just start splitting atoms. I mean, you could, but it won't end well.

Time crystals are an interesting concept that may push quantum computing forward in the future.

https://www.technologyreview.com/s/602541/physicists-create-worlds-first-time-crystal/
https://phys.org/news/2017-02-crystalshow-scientists-state.html

 

[macduke]

Time crystals are an interesting concept that may push quantum computing forward in the future.

https://www.technologyreview.com/s/602541/physicists-create-worlds-first-time-crystal/
https://phys.org/news/2017-02-crystalshow-scientists-state.html

Hah I was just reading about these the other day. Pretty wild stuff.

 

[Septembersrain]

Thank God no chipgate crap, I really would love to have a tsmc chip special iPhone 8

Nope, now let's see if they avoid "modemgate". LOL

 

[Glideslope]

Holy crap we are getting close to Moore's law. I feel like these chips will keep getting smaller for a few years, then a little thicker with newer battery tech helping to cover the increased power budget, and then we'll be forced to move on to quantum computing. I hope Apple is doing research on this. They have one of the best chip design teams in the world, but that will be completely irrelevant in 10 years. We know that Google has a quantum team. It would be pretty wise for people to research and invest in quantum computing companies.

Well, Apple has that abandoned complex in Ca. that has all sorts of loud, strange noises coming from it at 4am.

 

[macduke]

Well, Apple has that abandoned complex in Ca. that has all sorts of loud, strange noises coming from it at 4am.

The one where they were working on their car? Yeah, I agree that it's a lot smarter to spend research dollars on quantum computing than cars. Apple would be best served in developing strategic partnerships with companies like Tesla. I think a lesson modern day Apple should learn from post/pre Jobs Apple is that you shouldn't try to do everything. Saying no is a virtue to be cherished. A lot of this started after Jobs died and Apple started actually caring about their stock market value by giving out dividends while buying back stock. Jobs ignored all that nonsense and it still worked out pretty well for the investors. They're worried that they can't sustain infinite growth, but no company can. Even so, they're spreading too thin. Any company who comes out with a quantum computer with a decent, modern OS specifically designed to crunch through the incredible amounts of data that it can process—and at a reasonable price—will completely wreck Apple. And I wouldn't be surprised if Google is one of the first to get there.

 

[Glideslope]

The one where they were working on their car? Yeah, I agree that it's a lot smarter to spend research dollars on quantum computing than cars. Apple would be best served in developing strategic partnerships with companies like Tesla. I think a lesson modern day Apple should learn from post/pre Jobs Apple is that you shouldn't try to do everything. Saying no is a virtue to be cherished. A lot of this started after Jobs died and Apple started actually caring about their stock market value by giving out dividends while buying back stock. Jobs ignored all that nonsense and it still worked out pretty well for the investors. They're worried that they can't sustain infinite growth, but no company can. Even so, they're spreading too thin. Any company who comes out with a quantum computer with a decent, modern OS specifically designed to crunch through the incredible amounts of data that it can process—and at a reasonable price—will completely wreck Apple. And I wouldn't be surprised if Google is one of the first to get there.

Not so sure about Google. I'd keep an eye on Intel. They are on a new mission (finally) and we will see big leaps by them. I do agree with your Tesla comment. The only issue IMO would be the chemistry between Elon Musk and Tim Cook. Very different end games, styles, and abilities. I don't think Apple could purchase Tesla either. Musk would leave, and Apple would not be able to continue down that Tesla road under Tim's leadership.

It's actually very depressing to engage in these types of discussions. The time of Apple entering such an arena has long passed. ARM architecture can only push so far. I don't see Apple having any desire to develop for anything other than consumer Media Consumption Devices, and the occasional Term Paper.