Actually, TSMC's N3E process is somewhat of a regression on the first-generation 3nm process, N3. However, it is simpler and more cost-efficient, and everyone seems to want to get out of that N3 process as quickly as possible. That seems to be the biggest reason Apple released the A17(M3) generation and now the M4 the way they did.
The N3 process is in the A17 Pro, the M3, M3 Pro, and M3 Max. The A17 Pro name seems to imply you won't find it trickle down on the regular iPhones next year. So we'll see that processor only this year in phones, since Apple discontinues their Pro range of phones every year; only the regular phones trickle downrange lowering their prices. The M3 devices are all Macs that needed an upgrade due to their popularity: the Macbook Pro and Macbook Air. They made three chips for them, but they did not make an M3 Ultra for the lower volume desktops. With the announcement of an M4 chip in iPads today, we can expect to see the Macbook Air and Macbook Pro upgraded to M4 soon, with the introduction of an M4 Ultra to match later. We can now expect those M3 devices to be discontinued instead of going downrange in price.
That would leave one device with an N3 process chip: the iMac. At its sale level, I wouldn't be surprised if all the M3 chips that will go into it will be made this year, with the model staying around for a year or two running on fumes.
N3E still has a +9% logic transistor density increase on N3 despite a relaxation to design rules, for reasons such as introduction of FinFlex.[1] Critically though, SRAM cell sizes remain the same as N5 (reversing the ~5% reduction in N3), and it looks like the situation with SRAM cell sizes won't be improving soon.[2][3] It appears more likely that designers particularly for AI chips will just stick with N5 as their designs are increasingly constrained by SRAM.
SRAM has really stalled. I don’t think 5nm was much better than 7nm. On ever smaller nodes, sram will be taking up a larger and larger percent of the entire chip. But the cost is much higher on the smaller nodes even if the performance is not better.
I can see why AMD started putting the SRAM on top.
It wasn't immediately clear to me why SRAM wouldn't scale like logic. This[1] article and this[2] paper sheds some light.
From what I can gather the key aspects are that decreased feature sizes lead to more variability between transistors, but also to less margin between on-state and off-state. Thus a kind of double-whammy. In logic circuits you're constantly overwriting with new values regardless of what was already there, so they're not as sensitive to this, while the entire point of a memory circuit is to reliably keep values around.
Alternate transistor designs such as FinFET, Gate-all-around and such can provide mitigation of some of this, say by reducing transistor-to-transistor variability by a factor, but can't get around root issue.
There is also the fact that we currently have an iPhone generation where only the Pro models got updated to chips on TSMC 3nm.
The next iPhone generation is said to be a return to form with all models using the same SOC on the revised version of the 3nm node.
> Code from the operating system also indicates that the entire iPhone 16 range will use a new system-on-chip – t8140 – Tahiti, which is what Apple calls the A18 chip internally. The A18 chip is referenced in relation to the base model iPhone 16 and 16 Plus (known collectively as D4y within Apple) as well as the iPhone 16 Pro and 16 Pro Max (referred to as D9x internally)
The N3 process is in the A17 Pro, the M3, M3 Pro, and M3 Max. The A17 Pro name seems to imply you won't find it trickle down on the regular iPhones next year. So we'll see that processor only this year in phones, since Apple discontinues their Pro range of phones every year; only the regular phones trickle downrange lowering their prices. The M3 devices are all Macs that needed an upgrade due to their popularity: the Macbook Pro and Macbook Air. They made three chips for them, but they did not make an M3 Ultra for the lower volume desktops. With the announcement of an M4 chip in iPads today, we can expect to see the Macbook Air and Macbook Pro upgraded to M4 soon, with the introduction of an M4 Ultra to match later. We can now expect those M3 devices to be discontinued instead of going downrange in price.
That would leave one device with an N3 process chip: the iMac. At its sale level, I wouldn't be surprised if all the M3 chips that will go into it will be made this year, with the model staying around for a year or two running on fumes.