Apple is nearly through with its shift to its own chips. The Mac Studio and its M1 Ultra CPU, the company’s most powerful piece of silicon to date, were unveiled during the company’s recent spring event. But it also hinted at what Apple’s laptops might look like in the future.
For the first time, Apple has put all of its chips on the table.
The first and most important point is that Apple is now (if it wasn’t before) a chip-making force to be reckoned with. The company’s credentials were established by the overwhelmingly enthusiastic response to the first wave of M1 computers, as well as the similar success of its M1 Pro and M1 Max-powered MacBook Pro laptops last year. But with the M1 Ultra, Apple unleashed their most powerful chip yet, claiming it to be the “world’s most powerful chip for a personal computer.”
These chips are already gaining popularity as computer selling features. Buying a Mac is no longer only about acquiring Apple’s software or aesthetic design; it’s also about getting performance and battery life that no one else can match.
Apple took aim at Intel’s top-tier processor, the Core i9-12900K, claiming a 90 percent increase in multi-threaded performance at the same power level over its M1 Ultra, as well as the ability to match Intel’s greatest stats while using 100W less power. Apple performed a similar victory lap over Nvidia’s RTX 3090 GPU, which it claims outperforms while requiring 200W less power. (Of course, we’ll be putting those numbers to the test in the following days and weeks.) The Apple Silicon transition is no longer an experiment; it is Apple’s future, and it is something that PC manufacturers will have to consider in the future.
Then there’s the way Apple is constructing its CPUs. Apple’s Arm-based M1 chips currently come in four different versions, blurring the border between product form factors in a way that we don’t generally see from semiconductors. Apple has taken a different approach: instead of designing semiconductors for specific devices, the company has effectively created just one excellent chip: the A-series processor. And all it’s done is keep scaling it up, seemingly indefinitely. Apple’s secret recipe appears to be nothing more than doubling the size of each of its CPUs and adding additional cooling at each step, from a phone to a laptop to what is reportedly the most powerful desktop. But it’s extraordinary because no firm has ever done it before — and because it allows Apple to build a complete computer portfolio based on a single point in their semiconductor architecture roadmap, ranging from $430 to $8,000 (and counting).
The M1 chip found in a MacBook Air or iPad is the same one found in Apple’s iMac and Mac Mini desktop computers, and it operates at similar speeds and efficiency. With the Mac Studio, the M1 Max from a MacBook Pro laptop makes the transition to a desk. Even the company’s ultra-powerful M1 Ultra, which is effectively two M1 Max CPUs in a trench coat, isn’t a completely desktop-focused design. Devices are classified based on unique features or form factors, rather than on their processing capability.
Apple’s next Mac Pro, which Bloomberg’s Mark Gurman believes would arrive later this year with up to 40 CPU cores and 128 graphics cores on a single chip, is anticipated to follow the same scaling pattern (the equivalent of four M1 Max processors combined together, or two M1 Ultra chips). It’s a second doubling, this time with even more cooling to compensate.
In the same way that Apple differentiates the Mac Studio from the Macbook Pro with distinct form factors, ports, and feature sets, we should expect a similar change to help the new Mac Pro differentiate itself from the Mac Studio. The current Mac Pro is Apple’s most powerful (and most costly) computer, and it fills a very different niche than some of the company’s other PCs — one with plenty of Apple errors over the years as the company miscalculated what power customers require from their technology.
To satisfy pros, an M1-powered Mac Pro would likely require more than simply a doubling of the M1’s core count; it would also require scalability, modularity, and customisation. PCIe cards, user-accessible memory slots, and compatibility with discrete graphics cards and external hardware accelerators — these are the same features that made the recent 2019 redesign a success (and whose absence almost immediately doomed the 2016 “trash can” model). None of Apple’s Arm-based designs have included any of these features, and it’s unclear whether Apple is interested in doing so at all.
Apple’s CPU and GPU cores are so powerful that they may be able to defeat an RTX 3090 today; a 128-core GPU in a Mac Pro would provide an even wider cushion for a longer period of time. However, without user-upgradable parts, Apple would force a prospective Mac Pro buyer to plan ahead of time for all of their demands. However, we’ll have to wait for a more official announcement to see if Apple can avoid falling into the trap of relying too heavily on non-upgradable technologies once more.
While the number of cores and the split between efficient and performance cores varies from model to model (and even within processor families, where Apple offers a variety of configurations), the cores themselves are nearly identical: a Firestorm performance core on a $999 M1 MacBook Air is nearly identical to a $3,999 Mac Studio’s M1 Ultra, right down to the 3.23GHz clock speed, although the more powerful M1 Ultra has a higher clock speed. The Firestorm cores in the M1 aren’t substantially different from the ones in the A14 in an iPhone 12, either, albeit the iPhone cores are clocked somewhat slower.
From its most powerful desktop chips to its most battery-friendly ones for lightweight laptops, Intel’s latest 12th Gen CPUs are created following a similar scaled strategy, with a blend of horsepower and efficient cores. However, Intel’s CPUs don’t scale as well as Apple’s, with products still divided into various categories for different laptop and desktop models. Intel’s desktop chips aren’t making the transition to laptops and tablets in the same manner that Apple’s are.
THE M2 CONCERN
Finally, there’s Apple’s processor future. Apple Silicon is, without a doubt, here to stay (at this time, Apple sells just a pair of Intel-based machines: a legacy Mac Mini with severely outdated hardware and the soon-to-be-replaced Mac Pro). This indicates that, at some point this year, we’ll start to see the next generation of CPUs, whether they’re called “M2” or something else.
However, the next generation of Apple CPUs is unlikely to be as revolutionary as the jump from Intel to M1. Instead, it’ll most likely be a more slow, incremental improvement, akin to how the A-series iPhone processors were upgraded from one generation to the next.
When it comes to upgrading processors, there are essentially two options available. You can either employ a new (or refreshed) architecture that introduces more powerful or efficient CPU or GPU cores, or you can go to a smaller manufacturing node, which allows you to pack in more transistors in the same amount of space or shrink identical hardware even further.
We already know Apple has better silicon designs: the A15 chipset includes Avalanche high-performance cores and Blizzard energy-efficient cores, which are (at least on paper) superior to the Firestorm and Icestorm cores they replaced (which Apple originally debuted back with the A14 chip in the iPhone 12 lineup). With its A-series semiconductors for the iPhone, Apple has traditionally focused on refining individual core designs, but the benefits have been reduced from year to year.
A future “M2” might continue in this vein, bringing Avalanche and Blizzard cores to Apple’s semiconductor lineup, potentially giving similar performance or efficiency increases as the iPhone 12 to iPhone 13 update. According to 9to5Mac, Apple is considering doing so for its M2 portfolio, as well as adding more GPU cores to some of its chip models.
Apple might alternatively make a more modest change, as some rumors have suggested, and move the existing M1 designs to a more advanced production node. That could happen this year, according to rumours that Apple would release a new MacBook Air with a substantially identical chipset produced on TSMC’s 4nm node, rather than the 5nm node it presently uses for its M1 CPUs, allowing Apple to improve performance and/or power efficiency.
With the exception of the Mac Pro, Apple appears to have succeeded in launching its first generation of computer hardware with a bang. But its rivals aren’t sitting still: Intel has begun releasing its own next-generation laptop CPUs, and AMD’s offerings are better than ever. That’s not even taking into account Arm-based competitors, such as Qualcomm’s upcoming intentions to take on Apple in 2023 with its Nuvia-designed chips.
Apple Silicon was a new beginning for the company’s computers, propelling it ahead of the competition. However, now that the transformation is virtually complete, Apple must do more than simply impress once – it must maintain that momentum for future products as well.