Just to note....
If the author were to base the argument 100% on transistor density - that would indeed be a foolish argument. The assumption, however, is that Apple has delivered a power/performance gain that goes beyond the expected gains from the process shrink itself.
To be fair, Apple has built a huge amount of credibility over the years, and has more often than not exceeded their targets, if anything. It's just that this time, a few data points are pointing to the possibility that they may have exaggerated their gains. Here's a few points to consider:
1. The claim a 25% performance gain over A7, even though the core frequency gain is from 1.3GHz to 1.4GHz - a 7% increase. The explanation was that an enhanced core design has made up the rest - but we have yet to see any official benchmarks. But what we have seen are a few leaked measurements that suggest it doesn't get anywhere close to 25%. So that's one assumption that may have been exaggerated.
2. The entire density argument is based on a figure of "about" 2 billion transistors. It wasn't a precise value. "About" could mean 2.1 billion. Or it could mean 1.9 billion. Or it could even mean 1.55 billion - rounded up. :-) We'll just have to see about that. Anandtech's own analysis couldn't even get a precise value of the transistor count in the A7. They just called it ">1B". Well, how much greater? 1.4 billion in A7 compared to 1.55 billion in A8. <grin!> Point is, the data is coarse (at best!), and yet people are trying to compare it to some more precise figures that Intel gave out - and jumping to conclusions.
2b. Does everyone realize that A7 was built on a Samsung 28nm process? Are they taking that into account when they compare A7 vs. A8 - because some of the comparisons I've seen here want to conclude something about TSMC's 28nm to 20nm process improvement. Only problem - again - is that A7 wasn't built on TSMC 28nm....
3. What _really_ matters? I mean, does transistor density sell a chip at the end of the day, because I thought it was power/performance. No one has benchmarked an A8 device compared to a Core M device, nor have they compared battery life, device-based industrial design, etc. I suppose once iPad Air Gen2 is available, we can compare that to a Core M platform, and see which one delivers better performance, power, and battery life - things people care about. And if we really want to be spiteful, we can compare performance per transistor - and see how A8 looks in that metric <grin!!!>
News 
Market Data 
Discover 
AI
