“Unleashed”, in English literally means “released from a leash”. The correct translation is “unleashed”, but the etymology is different. Whether they are leashes or chains, Apple is convinced that it has got rid of them with the presentation of the new M1 Pro and M1 Max chips and two new MacBook Pros during the event on Monday 18 October, entitled “Unleashed”. In this case, the implications of the term are not just Apple’s hyperbole: M1 Pro and M1 Max are two unprecedented processors, which bring to Apple’s professional laptops levels of power – graphics, in particular – never seen on MacBook Pros.
On the previous episodes
Apple explains that “unleashed” works well because it can have multiple interpretations. First of all the new M1 Pro and M1 Max finally show the full potential of the transition from Intel to Apple Silicon. The M1 chip that had debuted at the end of 2020, already surprising for performance and power management was just a first taste.
The other leash from which Apple frees its new professional laptops is the power cord. An important aspect, which has passed relatively quietly after the event, is that the performance shown in comparison with professional laptops of the competition do not change if you use the new MacBook Pro by connecting them to the mains or with the battery. No other laptop in the pro range with Intel or AMD chips and dedicated graphics can do the same.
Finally, Apple doesn’t make it explicit, but the other yoke that it has taken its MacBook Pro away from thanks to the M1 Pro and M1 Pro Max chips is the dependence on Intel’s roadmap for the laptop market. The producer from Santa Clara, which Apple has relied on for 15 years after the transition from Power PC to Mac Intel in 2005-2006, it lost the train of the mobile revolution, remaining anchored to paradigms that today show serious technological limitations. In the meantime, Apple has capitalized on the experience gained with the production of the ARM series A chips for iPhone and iPad and translated it into the M1 processor line.
It is interesting to note that iPhones and iPads, often taken as examples of Apple’s alleged loss of interest in the professional market, are the reason why Apple can now bring its professional laptop chips to market. Without the iPhone and the iPad the most powerful and “Pro” MacBooks of all time could not exist.
Change of paradigm
For those who have been involved in the technology and PC market for a long time, Apple’s chips require a review of how the performance of a professional computer is evaluated. Not because the Apple Silicon line is complicated by an excess of technical factors, but if anything for the opposite reason: the engineers of Cupertino are managed to simplify it drastically, differentiating chip configurations only on the basis of the number of cores – the compute units – of the CPU and GPU and the amount of unified memory available.
The summary of the main features of the M1 Pro chip
The diagram of configurations available for MacBook Pros clearly shows this simplification. The versions of the M1 Pro There are three to choose from, with increasing power and performance:
-
M1 Pro con CPU 8-Core e GPU 14-core
-
M1 Pro con CPU 10-Core e GPU 14-core
-
M1 Pro con CPU 10-Core e GPU 16-core
L’M1 Max it can be had instead in two configurations:
All five versions of the chips also have a dedicated 16-Core Neural Engine, a section for hardware encoding and decoding of the video Pro Res, a secure enclave for security and Thunderbolt 4 support, which enables support for up to two external monitors on M1 Pro and up to four on M1 Max. 200GB / s (double the M1 of late 2020), while the M1 Max reaches 400GB / s. The maximum amounts of installable memory also change compared to the M1, which stopped at 16GB. On the M1 Pro you can choose whether to have 16 or 32GB of RAM, while the M1 Max supports up to 64GB.
Here we need another conceptual leap: because the chips also contain the graphics unit, the availability of 64GB unified memory means that the GPU potentially has access to more unified RAM than the best discrete graphics cards for PCs. One of the animations shown during the presentation keynote of the new MacBook Pro, we are told by Apple, is made up of 137 million polygons and it requires about 35GB of memory to run smoothly in real time. Such an animation would therefore not have been achievable on a professional laptop based on Intel chips and equipped with a discrete GPU.
Making comparisons
The M1 Pro and M1 Max also disrupt the concept of integrated graphics chip, which in the x86 universe is synonymous with low performance. A System on a Chip cannot be enough for any professional PC. So the CPU is always accompanied by dedicated GPUs To be able to compare its new SoCs with the competition, Apple had to show different comparisons for CPU and GPU performance respectively with Intel processors and best NVIDIA dedicated graphics cards.
In the case of the CPU, M1 Pro and M1 Max in the 10-Core version are equivalent. In multithreaded performance, with the same power consumption (peaking at about 30W) the two chips have 1.7 times the performance of the 8-Core Intel Core i7-11800H, one of the more powerful variants of Intel’s Tiger Lake 10nm processors. Apple chips use 5nm processes from Taiwanese TSMC (see Figure 1).
In the case of GPUs, Apple has compared the M1 Pro with 16-Core graphics and the NVIDIA GeForce RTX 3050 Ti 4GB discrete graphics card (see Figure 2). The two systems offer comparable performance, but the M1 Pro achieves them by consuming 70% less energy: the peak is about 30W of power against the 110W needed by the NVIDIA card. In the case of the M1 Max, the maximum relative performance is instead slightly lower than that of the GeForce RTX 3080, an even higher model. The difference here is greater: the Apple chip requires 100W less to achieve comparable performance. (see Figure 3).
It is also worth noting that the comparison graphics card is only found on heavy and bulky desktop “laptops” (some super laptops from MSI). The M1 Max can be mounted on laptops that they weigh 1.6 or 2.1 kg respectively and achieves this performance without having to plug the computer into a power outlet.
The name problem
Given the substantial leap in performance compared to the first Apple chip, it is surprising that in Cupertino they decided to keep the M1 name, adding only the Pro and Max suffixes to the new SoCs. The new processors are not in fact direct elaborations of the first M1, which was inspired by the structure of the A14, but rather represent a high-performance evolution of the A15 of the latest iPhone 13 and iPad Mini.
The reason for this choice – suggests DDay – it is probably linked to marketing considerations: the M1 has been so successful that it has become a brand in itself that is difficult to modify at such a short distance from the original launch, making the letter and the number inseparable.
The explanation is perhaps even simpler than that. Excluding the insiders, very few are really interested in knowing how the nature of the cores that make up the Apple SoC and the marketing name assigned to them. Moreover, as confirmed by Cupertino, the chips are internally indicated with code names completely different from the commercial ones. With this distinction independent of the build and generation of the chip, Apple can build a precise business history that brings together in a single arc all the chips for Mac, from the entry level M1s to the professional SoCs presented a few days ago.
The next steps
After all, Apple’s aim is not to sell its chips, but the Macs that contain them. The first Macs equipped with the M1 chip, a year after launch, for example, maintain a significant competitive advantage compared to the competition of the same range. Considering that a computer’s update cycles are longer than those of smartphones and tablets, Apple is in no rush to update the nomenclature and numbering of its laptop chips as it does on iPhones and iPads. It is not excluded that the entire first phase of the transition from Intel chips to Apple Silicon takes place with M1 processors, differentiated only on the basis of the suffix.
Looking at Apple’s current SoC line-up, we don’t feel a particular need for further differentiation in the short term. Apple will be able to upgrade low-to-mid-range computers keeping the M1 “smooth”, but also bringing back to the less expensive MacBook Pro features such as the HDMI port, the SD card reader or the Magic Keyboard without Touch Bar. At the same time the M1 Pro and M1 Max chips can also end up on the successor of the iMac Pro and on a Mac mini with a professional cut, also exploiting in this case the modularity of the configurations based on the number of cores.
The megachip on the way
And precisely this modularity will allow Apple to bring the first professional M1 also on its Mac Pro. Rumors are already running: thanks to the simplicity with which the computing units can be scaled, Apple’s engineers will be able to churn out an additional ultra-powerful chip with 20 or 40-core CPU and 64 or 128-core GPU. The chip will end up on a Mac Pro where the ease of managing dissipation and higher power consumption will allow to reach higher wattage peaks than those of the M1 Max, and therefore performance that could be up to 4 times higher in the most extreme configuration.
For this chip-monster, however, the unknown name remains: it will be interesting to see if the marketing division of Cupertino will maintain in this case too the name M1 (Ultra? Ultimate? Finedimondo?) or if instead it will be an opportunity to experiment with a different name to be destined only for very high-end computers.
.