So, while Apple’s Far Out September 7 event was definitely a mobile event, there was something for those of us who dig deep into computer nuts and bolts.
I was honored to test some of the best processors on the market, from the best Intel to the best AMD processors available, and even dived into Apple Silicon recently. Computer processors have acquired an aura of uniqueness that makes even the best phone processors seem almost peculiar, but take a look under the casing and we begin to see some convergence, something that the Apple A16 Bionic shows more than ever.
To be clear, the new A16 Bionic is not a desktop chip or even one that can compete with the best laptops, but given its specs, there are many things that point to it being more like the latter two than you initially think.
16 billion transistors is a lot for a processor, any processor
The core component of the processor is a transistor, a nanoscopic electronic switch that converts electrical pulses into zeros and ones that can represent data and logic operations. This is a neuron from any microprocessor, so the more neurons you have, the more powerful the processor will be.
The ability to squeeze 16 billion aB transistors onto a mobile chip is unusual, especially compared to the Apple M2 chip which contains 20 billion. This puts the A16 Bionic at around 80% of the density of Apple M2 transistors, but even more important is the size of these transistors.
The A16 Bionic is built using a 4nm TSMC node as opposed to the 5nm node used to manufacture the Apple M2. This means that although the A16 Bionic matrix is smaller than the M2 matrix, it still approaches the density parity in absolute terms.
It also allows the 5-core GPU and 16-core neural engine to be placed on the SoC along with the central processor, and while the A16 Bionic GPU is half the size of the M2, it should still show high ability for impressive graphics, especially for the mobile phone CPU .
Meanwhile, the neural motor is the same size as the M2 and this is where the much improved power of the phone can show up, especially when it comes to editing photos and videos on the fly.
Moore’s Law still exists
Another thing to consider when it comes to the Apple A16 Bionic and Apple M2 chips is that there is still a physical limit to the number of transistors you can ultimately fit on each of them.
This hard hat is set by physics, in fact, because the transistors are already operating on a truly atomic scale. And while desktop processors will have more room for a physical increase in size – much more than laptop processors, and definitely more than those found in phones and tablets – we’re really at the point where the physical constraints of matrix size dictate potential power, not some remarkably smaller transistor in a stationary product.
Given these kinds of limitations, what the A16 Bionic is likely to achieve in terms of performance is fantastic, but it still comes to the limits of transistor densities much harder than stationary chips for a while.
This means that next-generation phone processors will still lag behind desktop and laptop processors, and this difference may only deepen in the future as they have to be physically smaller to fit in a phone or tablet.
This is where the issue of transistor density really comes into play, because as we’ve seen with the M1 Pro, M1 Max and especially the M1 Ultra, utilizing the physical space you need to achieve to get the performance gains is a major resource for these desktop chips and laptops.
So while the A16 Bionic looks very powerful, there’s literally only as much space it needs to expand, unlike Apple’s M-series chips, so any performance gains the A16 Bionic can squeeze in are really limited to an effectively smaller size 4nm node transistor compared to the 5nm node used in the A15 Bionic and especially something like the Apple M1 chip. The latter was able to physically evolve into larger M1 Pro and M1 Max chips where transistor density has a much greater impact on performance.
Phone processors will hit the wall long before the MacBook chips
While it is clear that the A16 Bionic can easily handle a computer from just a few years ago, including some of the best MacBooks and Macs running some of the best Intel processors of the time, the performance gains on Apple’s MacBook and iMac lineup will be accelerating ahead of their mobile chips in terms of a generation-per-generation performance increase.
As powerful as the A16 Bionic could be, it would have a hard time booting up the MacBook Air, although it could theoretically do so, with some limitations. After all, the Apple M1 chip also had 16 billion transistors, although it had a larger 7- or 8-core GPU.
And while it might be able to power the hardware in an older MacBook Pro with an older Intel chip, in no universe a modern MacBook Pro, even a 13-inch, could run on the A16 Bionic without seriously limiting your expectations.
The gen-on-gen, A15 Bionic, had 15 billion transistors, while Apple’s latest iPhone chip has 16 billion, an increase of about 6.7%. Meanwhile, the Apple M2 saw a 25% increase in CPU density compared to the Apple M1. Apple is unlikely to repeat exactly this stunt with the M3 (although 25 billion transistors are not entirely out of the question), but it will almost certainly exceed the 6% to 7% increase that we will likely see on the A17 Bionic.
By the time the M3 is launched, no iPhone chip will come close to the competition with its crude performance in the next few years, and it is likely that this gap will only widen over time. Not that the A16 Bionic is not that impressive, but what we expect from even the best low-cost processors with integrated graphics in the next few years is likely to be even greater.