Apple Reality, India, and security are shaping up to be the top three things happening for Apple in 2023, but there's one more: Apple Silicon, primarily a shift to 3nm process technology in the upcoming A17 and M3 (series) processor upgrades. We're starting to hear whispers about what these chip upgrades mean.
We already know what to expect: higher computing performance at a lower wattage than before, essentially enabling Apple's devices to do more work with less energy.
No, as the migration from M1 to M2 shows, we didn't anticipate the same huge leap that companies made when they switched from Intel processors to their own chips in Macs, but the improvements are still generous.
How generous is it? Based on claimed * Geekbench scores, we expect a 20% performance improvement from the upcoming iPhone chip A17. This boost will show up in useful features such as fast image editing and prettier visuals.
* These claims appear to come from China, so we cannot verify their authenticity. But given the expected benefits of moving to 3nm process technology, they seem reasonable. It may be worth noting that the leaker also warned of the upcoming yellow iPhone to follow.
The leaked A17 benchmark seems promising
The leaker reported a single-core score of 3,019 and a multicore score of 7,860 on Geekbench 6. The iPhone 14 Pro's A16 Bionics reached 2,504 and 6,314, respectively, but fell short of the previously claimed leak scores of 3,986 and 8,841.
Although these scores may not be accurate or real, they do seem to be tightly aligned with the expectation of a shift to 3nm chips. According to Digitimes, the processor is about 35 percent more efficient.
That's for the iPhone. Apple is also expected to launch its first M3 Mac processors this year, and most speculation suggests these will also be based on the 3nm process. Just as the M1 chip is an updated version of Apple's mobile processor, the M3 will use some shared resources.
That means it's reasonable to imagine a similar performance boost for a Mac with these processors.
To put these expected benefits in context, think of Photoshop. Adobe was thrilled when Adobe brought Photoshop to Intel's M1 Mac, which saw a massive 50% performance boost.
When Apple launched the MacBook Air running the M2 chip in 2022, it claimed it would make some key Photoshop conversions five times faster than Intel Macs and 20 percent faster than the previous M1 system. Now we're looking for another 20 percent or so.
It's clear that Apple's silicon chip team is moving forward, and as the per-watt performance of each chip improves, that means these new systems will compete more aggressively with higher-end systems.
All of this performance is contained in systems that require quantifiable less power, which continues to have a significant impact on businesses running large numbers of computers. The best recent insight came from MacStadium's revelation that Server Center, which hosts its Mac mini servers, complained that the computers weren't using as much power as MacStadium was paying for in its hosting plan. At scale, this is good for both energy bills and the environment.
These performance improvements are significant. They gave Apple a lead in low-power chip speeds, and competing systems can't seem to match them in power consumption.
This strategic importance is reflected in Apple's investments. At a ceremony on Dec 29 at its new Fab 18 construction site in South Taiwan Science Park, TSMC announced the start of mass production of 3nm chips that are expected to bring in US $1.5 trillion in revenue over the next five years.
Subsequently, Digitimes claimed that Apple bought chipmaker Taiwan Semiconductor Manufacturing's entire 3nm processor production, making it the only manufacturer able to ship mobile devices and computers equipped with 3nm chips on a large scale.
The investment is important because it also gives Apple a clear path to improving its chips in the future. "Compared to the 5-nanometer (N5) process, TSMC's 3-nanometer process provides up to 1.6-fold logic density gain and 30-35 percent power reduction at the same speed, and supports the innovative TSMC FINFLEX architecture," the company said in a statement.
FINFLEX provides Apple with a signature advantage as it lays out a roadmap for future improvements in chips made with 3nm chips. The roadmap means we'll get the M1, M1 Pro, M1 Max, M1 Ultra, M2, M2 Pro, and M2 Max because the FINFLEX architecture has let Apple's chip team accurately optimize performance and power consumption.
TSMC isn't sitting still either -- previous reports say the company is already preparing to launch the N3E, an enhanced version of its existing 3nm production technology, by the end of this year. Apple should also be the first customer.
So what is the takeaway from this dizzying array of facts, rumors and speculation?
Not only will Apple carve out a unique advantage as the sole supplier with a 3nm chip count, but it can also make incremental improvements on these processors and lay out a clear road map for new processor designs next year.
We can't say for sure if this means Apple plans to upgrade some or all of its Macs with new chips every year, as it already does with iphones, but the frequency with which it rolls out new chips suggests it might.
Perhaps it intends to upgrade laptops more frequently, while offering maximum relative improvement on less frequently updated desktops. But whatever approach it chooses to take, the days of the Mac lagging behind in terms of performance improvements are over for us. No AIM Alliance barriers, no waiting for Intel. And Taiwan Semiconductor is poised to set up chip production in the United States, which will help Apple secure future processor supplies.
The result is that Apple seems to be able to offer powerful Mac, iPad and iPhone upgrades as often as once a year. This means that if Macs don't yet provide the performance your business needs, it's only a matter of (seemingly short) time before they do.
Apple's processor story will inevitably translate into Mac market share gains, even as overall PC market growth erodes.
