The new Apple M1 processor is quite impressive, but the first realistic ratings have shown that it is actually far from what Apple is talking about: it is not the fastest processor for laptops or the most efficient, not even proximity.
Intel and AMD x86 portable CPUs leave Apple M1 high and dry
Interestingly, not many people consider Apple’s M1 processor based on the production process at 5 nm, and that it currently has two nodes ahead of Intel (10 nm) and one node before AMD. (7 nm) at least. All in all, this advantage of lithography has nothing to do with Apple’s architectural and engineering expertise and is simply the logography provided by TSMC, but despite this dire situation, both Intel and laptop processors from AMD have left Apple’s processor at the top of the bitumen.
Unlike GeekBench, it uses multiple algorithms and downloads the geomean (which can delay results if the performance of a particular algorithm is found in architecture), Cinebench measures the power of a powerful processor and is the standard for most enthusiasts when it comes to processor power. An offer benchmark that can fully use the performance of the processor and all of its components without any inefficiency.
The Apple M1 Chip collects 1498 points in a single test on the Cinebench R23, as we can see in the graph above. AMD’s x86 chips, which are a unique backdrop and don’t have the most aggressive clock speeds (they prefer multi-core performance as we’ll see later) to get up to 1284 points; on the other hand, the Apple M1 easily outperforms Intel processors even though it is behind two in the production process.
After all, single-core benches are cosmetic simply because no CPUs limit their use to a single core and in all cases always use their full power, so let’s take a look at test results using multiple cores:
Here we can clearly see that the 7nm AMD chips completely eliminate Apple’s M1 chip, and the 14nm Intel processor has almost the same notes, while the 10nm variant (with half cores) registers less than 1000 points below Apple’s processor. The variant of the 10nm Intel TGL with 8 cables is expected to hit the market in a few months, and it is clear that there will be a slight increase in this table you can see above judging by the results obtained with only 4 cores.
As we can clearly see from these benches, the Apple M1 is by no means the fastest chip out there as Apple claims, and is actually far from it.
Apple and its ARM processes
It should be noted that the M1 chip makes a lot of sense with Apple. Thanks to your special contract with TSMC, you are able to access the nodes ahead of any other company and you can use this node benefit to integrate it with the construction of your ARM to maximize revenue streams without exposing too much performance. In fact, with traditional code and mobile form factor, performance differences are almost negligible and will not make a difference.
High performance, on the other hand, is another matter. Desktop processors were not deliberately added to the above graphics because even multi-processor processors are crashing the Apple M1: both Intel and AMD processors are playing in another league, directly and no matter how hard Apple tries.
The problem is that ARM design does not work with very high frequencies and has high leakage at high power levels (this is also true and in retrospect, all can be said, because x86 processors are not very suitable for low power conditions). This means that the most active components in the market are safe from Apple’s ARM’s influence, unless companies such as NUVIA decide to enter the fray and begin to take advantage of this market segment.