I hate to say I told you so, but if Mark Gurman is to be believed, Apple is set to have the M4 chip out before the end of the year, just 12 months after M3 arrived in October of 2023, and the rollout looks similar, but more extensive. Maybe, just maybe, as I said in 2020 before we’d had our first Apple Silicon Ship, Apple is building out a stable, annual release schedule for the future.
Before I come to the roadmap itself, why did I predict that this would be the way Apple would go with their own Mac focused chips? Simple, Apple Silicon is based on the exact same cores as the iPhone’s A chips, But while the A17 has 4 efficiency cores and a pair of performance cores, the M3 has 4 of each, offering more performance at the cost of slightly higher power consumption and also doubles the GPU cores from 5 to 10.
So what are we getting and when?
First up, and very much in line with last year, a low end MacBook Pro and an updated 24” iMac, both with M4 coming around the end of this year, lining up perfectly with the 2023 releases that came at the same time. Bang on 12 months, happy days. Let’s keep this like clockwork.
Next, 14” and 16” MacBook Pro with the M4 Pro and Max chips, again coming at the end of 2024, just like last year, though Mark has added a little wiggle room saying between the end of 2024 and early 2025 - We saw October last year, but we have had January releases in the past, with M2’s MacBook Pros arriving in January, potentially following some delays from a targeted late year release. The Mac mini with M4 and M4 Pro is also slated to have the same release window as these, which right now suggests that Apple’s cheapest Mac will skip the M3 chip generation.
Then in the Spring, the 13” and 15” MacBook Air will get its M4 upgrade, again, mirroring 2024 with their March announcement.
We’re expecting the M powered iPad Pro M3 and iPad Air M2 in a pair of sizes each to come this year in May, so that could also become the more powerful iPads’ slot in the year, with the M4 and M3 versions respectively arriving in May 2025. The Mid year in 2025, its time for the “high end” M4 chip, with the M4 Max and potentially Ultra launching in the Mac Studio, and then a Mac Pro with M4 Ultra (and hopefully something even more powerful) in the “Second half” of 2024, which sounds vague, but the second half starts in July, so just after WWDC could make sense.
In fact, Apple Silicon has taken such a leap from the dark dark intel days that if you have any Apple Silicon Mac and your workload hasn’t drastically changed, you almost certainly don’t need to upgrade yet, even if you’re on the M1 generation. To put it in context, the M3 MacBook Air is 13 times faster than the Intel MacBook Air that was sold up until 2020. 13 times. That is insane. Chips generally get 10-20% faster generation over generation these days, so that leap from Intel to even first generation Apple Silicon is enormous. An M1 MacBook Air, which you can now pick up for as little as $699 brand new, is still a perfectly capable machine for anyone doing normal person things with a computer. And even far more than normal people computing. This video for example is shot on a couple of iPhones, and edited on a base level Mac mini M1 with 8gb of unified memory and 256gb of storage. The lowest spec Apple silicon Mac chip ever released. It’s VERY capable, even now.
But, all that being said, there is potentially a big reason that you MAY want to upgrade to an M4 generation chip this year, and that is the buzz word of the past year or so, AI. It seems that Apple is going big on AI this year, and the latest reports show that a lot of the AI features will run on device on the new iPhones, while in most cases AI heavy lifting is done in cloud server farms. Apple however has been putting Neural Engines for machine learning in their devices for year. These have been used for behind the scenes Machine learning tasks like cataloging what’s in every image in your photo library, reading and indexing the text in pictures, working as a part of the image processing pipeline for HDR, frame selection and more. But with Apple, this runs on device instead of being processed in a server farm, keeping your data safe and meaning can also run when you’re away from wifi or out of cellular signal. It also removes the cost of those high compute farms that are often built from GPU style chips, meaning the service would likely not be charged for by Apple, especially for any on device processing. A very different approach to the competition right now.
Join the conversation
http://iCaveDave.com/twitter
http://iCaveDave.com/facebook
http://iCaveDave.com/instagram
http://iCaveDave.com/youtube
http://iCaveDave.com/merch
http://iCaveDave.com/discord
56 Comments