[Jtsummers]

It's probably a good idea to skip the first generation in general, but this is a bit different than those others.

The first iPhone had a fixed function graphics pipeline and 2G cellular service. Unsuitable for the native application market that they set up afterward and changing cellular capabilities and customer expectations.

The first iPad was similarly very constrained, though slightly better. For all practical purposes, it was a blown up iPhone 4.

I suspect, given the general advancements and recent stability in Apple's ARM SOCs, that this first generation of hardware will be good, though maybe a bit underpowered compared to what people are used to with the Intel equivalent. But not the practically beta level of those other devices.

[jandrese]

I wouldn't be surprised if the first generation ships with firmware issues. The x86 emulator is also likely to be a source of friction. In the past when Apple has switched architectures the new one is so much faster that even emulated code see some benefit. That's not the case here, old x86 apps stand a good chance of performing poorly and having issues. I wouldn't blame anyone for holding off for a generation or two.

[newscracker]

> The x86 emulator is also likely to be a source of friction. In the past when Apple has switched architectures the new one is so much faster that even emulated code see some benefit. That's not the case here, old x86 apps stand a good chance of performing poorly and having issues.

This doesn’t match Apple’s claims and demos of performance of emulated x86 applications. I’d suggest you watch the WWDC 2020 keynote video for more on this and take a look at the compute and graphics intensive x86 applications they demonstrated running (emulated on Rosetta 2) on ARM hardware of undisclosed specs.

[jandrese]

I'm taking those demos with a grain of salt personally. Historical precedent suggests this is very difficult to get right.