I'm not saying it's not better than Tiger Lake - it is better than Tiger Lake and that's a fact.
The discussion i'm trying to have and argue is that it's not a chip design breakthrough like Apple is claiming for, especially when u take into account the fact that TSMC took the lead in manufacturing technology which gives them the edge they have now finally, exactly like AMD have released a better desktop chip than Intel with their new Ryzen series.
they didn't invent the wheel of chip design and gave us 3X times performance from anything else existing in the market - they did a good progress in chip performance (after a while now it had been quite stuck due to Intel's difficulties in advancing their manufacturing technology...) , and kept the prices low, which for us consumers, is the most important thing.
What is remarkable about Apple chips is that they match the performance of anything else out there at significantly lower power consumption — not 10% or 20% lower (which would already be a noteworthy achievement), but 70% lower. In another words, Apple needs 3-4 times less power than AMD (and about 4-5 times less power than 10nm Intel) to reach the same performance target.
Sure, you might say that this is all because TSMC's advanced process... and it certainly plays a role, but you cannot use the 7nm to 5nm shrink to explain all of A14 performance and energy efficiency, because A13 already was both performant and energy efficient. And A14 is not just a shrink, it significantly increases the cache size and adds additional execution units, making the CPU even wider.
Apple's main technological achievement — as I see it at least — is that they are able to make a relatively "slow" CPU (frequency-wise) that has an unprecedented level of instruction level parallelism. Both Intel and AMD rely on high frequencies to reach high performance. Apple instead relies on out-of-order execution of instructions. It is frankly a bit of a mystery how they do it. The "common wisdom" states that the data dependencies in the code will quickly limit how much instruction you can run in parallel, but somehow Apple broke it. Their CPUs are so superscalar that it puts any other design to shame.
The drawback of their approach of course is that their CPUs are so complex that they cannot run the same high frequencies as x86 CPUs can. What you say is essentially — their chips are not any faster than state of the art Intel or AMD chips, they just can do this with less power. But — and forgive me for using this unoriginal phrase —
you are missing the point.
In the last decade, as CPU designers found themselves unable to further increase the speed of their CPUs, they kind of managed to hack the problem by going around it. Multi-core designs became a norm. As many important workloads benefit from multithreading, this allowed the performance gains to continue. But it created a new problem: power-hungry x86 CPUs had to be clocked down
significantly in order to prevent multi-core chips from melting, which in turn had negative impact on performance. It was not long time ago when gamers would choose dual-core CPUs before quad-core ones, because the dual-core were running faster clock. This problem was "solved" by turbo boost technology. Now the cores could dynamically switch between power-hungry single core mode and more energy-efficient multi-code mode, making these CPUs very flexible. But turbo boost again brings it's own share of problems (unpredictable performance, user confusion, etc. etc.).
Apple's CPUs kind of break the system. They might not be significantly faster per se. than other high-end CPU designs, but their energy efficiency allows than to stack multiple cores without having to downclock them that much. An Apple A14 CPU core can run at 3.0ghz consuming around 5 watts, while having the performance that equals a Tiger Lake core at 4.5-4.8 ghz. Apple could pack 8 those cores into a 45W CPU literally tomorrow, if they wanted (and in fact, we will see them coming out next summer). When do you think Intel or AMD are going to release an 8-core CPU that can run with each core sustaining 4.5 ghz? Two years? Four years? Will they ever?
And this is the point that you are missing. We already see a quad-core entry-level Apple CPU, running at 10-15W that outperforms an 8-core AMD CPU running at 25W. This is indeed a 2x performance advantage. They could make it a 3x performance advantage by scaling up the number of cores and using higher TDP targets. And of course, they have GPUs that are 2x-3x faster per watt than any competition...