For fifteen years Mac hardware development was largely determined by Intel’s product release cycles. This year’s WWDC marks the start of Apple’s second M-series release cycle, promising valuable incremental improvements in performance, with particular emphasis on the Neural Engine (ANE). With the first two Mac models using new M2 chips due to ship next month, this article speculates on the rest of the M2 cycle.
Apple’s M1 series started with four lower-end Mac models, each equipped with the basic M1 chip offering two four-core clusters, one Performance (P) and the other Efficiency (E):
- 11/2020 MacBook Air
- 11/2020 MacBook Pro 13″
- 11/2020 Mac mini
- 05/2021 iMac 24″.
It took almost a year for the next M1 variants to become available in higher-end MacBook Pro models. The M1 Pro and Max have the same CPU core architecture, doubling the number of P cores to eight, divided into two four-core P clusters with a two-core E cluster:
- 10/2021 MacBook Pro 14″
- 10/2021 MacBook Pro 16″
After a further five months, Apple completed the M1 series with its doubled M1 Ultra chip, containing four four-core P clusters and one four-core E cluster, in the Mac Studio, released in 03/2022 but delayed until April-May because of problems with chip supplies.
Starting from month 1 of the first year as 01/01, a diagram of the whole cycle shows its three phases spread over a period of just under 18 months, giving an overall length of about two years, matching that for Apple’s devices.
For Apple’s second release cycle, the first M2 Macs are due to ship next month (July), marking the first month in the cycle. Later this year, we should expect M2 Mac minis and iMacs, most probably around October-November. It looks unlikely that the Pro and Max versions of the M2 will be available in production quantities this year, but might be released in 14 and 16-inch MacBook Pros as early as late Spring 2023. That leaves the high-end Studio with its M2 Ultra for July or September 2023 at the earliest, and possibly as late as November, ready for the M3 to be announced at WWDC in June 2024.
One model not taken into account in this is any replacement for the Mac Pro 2019, Apple’s last Intel Mac which coincidentally was announced at WWDC three years ago. Whether Apple will release a replacement using its M1 Ultra later this year is likely to be driven by specialist demand, and whether its small market is ready to switch from Intel processors yet.
M1 v M2
Stepping up from M1 series chips to those of the M2 isn’t the big leap we’ve been experiencing over the last year or two when switching from Intel to ARM. Most of the improvements in the M2 are incremental: 18% improvement in CPU core performance, and up to 24 GB of faster memory. GPU performance improvement is more substantial at 35%, but the biggest difference comes in the Neural Engine, claimed to be 40% faster, which remains one of the less-used parts of the chip.
Although Apple could always spring a surprise, if you’re looking for a Mac with a Pro, Max or Ultra series chip, and don’t want to wait until 2023, you’re unlikely to see a suitable M2 model before Christmas this year.