Apple (NASDAQ:AAPL) is notorious for being one of the most secretive companies on the planet. It's hard to blame them, particularly as Apple wants to avoid the potentially disastrous impact of the Osborne Effect to its current product sales and wants to stay as far ahead of the competition as possible.
That being said, if you pay close attention to commentary from Apple's suppliers on their quarterly earnings calls, it's not too difficult to at least get a glimpse into what the Cupertino, Calif.-based device maker is up to.
On-contract chip supplier TSMC's (NYSE:TSM) most recent earnings call, I believe we got one very interesting nugget of information pertaining to Apple's next-generation iPhones. Let's take a closer look.
The die-size question
On the call, analyst Steve Pelayo made an interesting (if not entirely accurate) observation. He noted that when Apple moved from the 20-nanometer A8 to the 16-nanometer A9, Apple was able to add additional functionality to the A9 without seeing much die-size growth thanks to the smaller manufacturing process.
The reason I think this statement isn't entirely accurate is that Apple did actually grow the die size on the A9 fairly substantially compared to the A8 (104.5 square millimeters, up from 89 square millimeters on the A8) as the 16-nanometer node brought minimal area scaling compared to the 20-nanometer node.
With this background in place, he asked management about whether Apple's upcoming chip, known as the A10, would see a significant increase in die size as Apple needs to improve performance while still using the same 16-nanometer manufacturing process in which the A9 was implemented.
TSMC co-CEO C.C. Wei declined to comment on any potential die size growth in the A10 relative to the A9, but he did say that Apple added "a lot of functionality inside" in order to make the phone "better."
This might seem obvious, but it's actually quite interesting
It may seem on the surface that this is almost a "duh" kind of observation -- Apple is making its A10 better than its A9. Shocker, right?
However, I think that this is actually quite significant. For one thing, the addition of a lot of functionality does point to a meaningful increase in die size, which should mean good things for
TSMC (as this should mean more wafers required at a given yield rate).
It's also very good news for Apple and its stockholders.
Apple faces the challenge of trying to make its next-generation smartphones significantly more compelling than prior-generation ones in order to convince users to upgrade. One way to do that is to simply deliver more CPU and graphics performance to allow for a more responsive system overall.
However, doing just that isn't enough.
I suspect that with the A10 Apple will improve all of the major subsystems in the system-on-chip. I expect the CPU to receive some architectural enhancements and to run at a higher frequency compared to the CPU core in the A9.
I also expect Apple to significantly beef up the graphics capabilities of the A10, possibly even going so far as to introduce its very first custom graphics processor. Indeed, I believe that it will be upping the resolutions of its next-generation iPhones, so the A10 will need to deliver a nice boost in graphics capability to support that resolution increase.
Additionally, I'm expecting the iPhone 7 to deliver some significant camera improvements (both still and video), which means that Apple is going to need to add some additional muscle to the image signal processor in the A10 relative to the A9 (which already has quite a beefy image signal processor).
On top of that, I expect to see Apple make some significant improvements in other areas such as media encode and decode (for more efficient video playback and capture).
The A10, in my view, is going to be a pretty big deal.
It only gets better from here
I'm excited to see the improvements that Apple will make in the A10 relative to the A9, but I can't help but be even more excited to see what the A11 and A12 bring. TSMC confirmed on its most recent earnings call that its next-generation 10-nanometer technology will go into mass production by the end of 2016 and that its 7-nanometer technology should start volume production in the first half of 2018.
This means that Apple will enjoy the benefits associated with moving to new manufacturing technologies for both the A11 (iPhone 7s) and the A12 (iPhone 8). As impressive as I expect the A10 to be, I fully expect the A11 and A12 to be game changers in terms of mobile chip performance.