It is well-known that the A9 chip inside of Apple's (AAPL 1.27%) recently launched 6s/6s Plus phones is manufactured by both TSMC (TSM -0.34%) on its 16-nanometer FinFET Plus manufacturing technology as well as by Samsung (NASDAQOTH: SSNLF) on one of its 14-nanometer process flavors. (I have confirmed with multiple sources privately that it is Samsung's higher-performing LPP process, not the LPE that was used to build Samsung's Exynos 7420.)

However, multiple reports have recently hit the Web -- this time from the generally reliable KGI Securities analyst Ming-Chi Kuo as well as from Japanese website Touch Lab -- claiming that the A9X is manufactured solely by TSMC.

Let's take a closer look at the implications of this.

A very good sign for next year's A10 processor
According to a source I believe to be reliable, TSMC's 16-nanometer FinFET Plus technology offers better transistor performance/electrical characteristics than Samsung's 14-nanometer LPP technology (something I first published here).

The fact that Apple chose to use TSMC exclusively to manufacture the A9X is probably a good indication that this is indeed the case.

In a recent article published right here on Fool.com, I argued that because of the above fact, the A10 -- which I believe will be produced exclusively by TSMC -- should be able to see a nice boost in clock speed/performance as a result of being produced solely on TSMC's 16-nanometer FinFET Plus.

I think the fact that the A9X uses the same CPU cores at the A9 but clocks them significantly higher (+22%) is a fairly good indication that Apple will, indeed, be able to get a "free" clock speed boost by building the A10 exclusively on TSMC's process next year and not having to dial things back to allow the chip to be manufactured on the inferior Samsung process.

This clock speed improvement should, of course, come in addition to any architectural enhancements Apple brings to the next-generation CPU. Although I don't expect as big of a boost as we saw with the A9, I think Apple will do some interesting things with its next-generation CPU. 

I expect TSMC to win the A10X; what does the future hold?
I suspect that in addition to winning the entirety of the A10 orders, the superior performance/power characteristics of TSMC's 16-nanometer FinFET Plus technology should allow it to win the entirety of the orders for the next-generation iPad-exclusive chip, likely to be called the A10X, next year.

After all, the A10X is a higher-performing, harder-to-manufacture variant of the A10, so it seems only natural that if Samsung's 14-nanometer LPP wasn't good enough for the A10, then it's not going to be good enough for the A10X. 

The picture is a little less clear for what happens after the 14/16-nanometer generation. TSMC seems to be quite confident that it will be able to begin mass production of its 10-nanometer technology by the end of 2016. Unlike the situation with 14/16-nanometer, where TSMC had a bit of a late start relative to Samsung, TSMC should be in a position to start building the next-generation A11 chips on its 10-nanometer process beginning at the end of next year.

From what I've been told by a source with knowledge of the situation, Samsung is still in the running for the A11 orders. Although I'm sure Apple would like to foster competition between Samsung and TSMC in order to get the best deal it can on wafer prices, I am inclined to think TSMC has the upper hand at the 10-nanometer generation given its strong technical execution at the 14/16-nanometer generation.

If Samsung can deliver a 10-nanometer process with comparable performance/power/area characteristics to TSMC's 10-nanometer process at good manufacturing yields, then I expect Apple to "split" the orders as it did with the A9. However, if one foundry winds up with a significant lead in performance/power/area as well as manufacturing yields, then I would expect Apple to go with the foundry with the better solution.

I suspect we'll learn more about Apple's process choice(s) for the A11 by next year.