Desktop Macs appear to now be on a two-year cycle (at best). Notebooks way outsell desktops, so I think we’ll get at least M4 base models this year, maybe Pro and Max options for the MBPs as usual. The M4 Ultra is almost certainly going to wait till mid-next year.
The kind of double-digit speed increases we’ve seen suggest to me that the Mach5 (my name for the M5 … c’mon Apple that name’s just sitting there) is a 2025 chip, and that major chip changes will likely take longer than we’re used to going forward.
Apple has the wrong priorities. Instead of focussing on the AI hype, better release a M4 27 inch iMac.
Same stale argument you trot out again and again, and oh so true given how successful they are, armchair fucking CEO.
I don't understand what "AI hype" means. Sure, it's been in the media, but that's because major breakthroughs have lead to beneficial new services. It's also something Apple has been including in their own chips for many years now. Finally, AI in their M-series chips has no bearing on whether Apple feels that a 27" iMac update is viable for the market or now. It's all just weird.
Apple would be leaving money on the table if they don’t update the Studio & Mini
Apple net income 2023 = $97b Mac revenue 2023 = $29b Mac desktop revenue = ~10% = ~$3b Mac Studio & mini revenue = < 5% = < $1.5b Mac Studio & mini net margin ~25% = $0.37b
That's how much would be left on the table, assuming nobody bought the older ones that are available now. How much potential revenue determines most companies' priorities.
It's not like the Studio and Pro are outdated, they were refreshed 12 months ago. The mini is 1.5 years out. They can throw an M4 in the mini whenever they want but it's not impacting Apple's income.
The big number of total revenue is the sum of a lot of little numbers. If you rationalize ignoring each little number, it will eventually affect the big number.
Hopefully Apple plans to respond to the Qualcomm and Microsoft challenge Qualcomm in particular has been talking trash since last November about their new Elite Snapdragon processor
Qualcomm (and others) are playing catchup to Apple, they compare their chips to the lowest-end M3 but Apple already has M3 Pro and Max:
M4 is even higher. The Qualcomm chips are aiming at mainstream low-end performance, they don't scale to Max-level performance (just Max-level power usage), which is why they get paired with Nvidia GPUs (M3 Max = Nvidia 4070). Pairing with a 3rd party GPU loses the memory sharing so you get 8-16GB VRAM.
Comments
https://nanoreview.net/en/cpu-compare/qualcomm-snapdragon-x-elite-vs-apple-m3-pro
M4 is even higher. The Qualcomm chips are aiming at mainstream low-end performance, they don't scale to Max-level performance (just Max-level power usage), which is why they get paired with Nvidia GPUs (M3 Max = Nvidia 4070). Pairing with a 3rd party GPU loses the memory sharing so you get 8-16GB VRAM.