I’ve not seen discussion that Apple likely scales performance of chips to match the use profile of the specific device it’s used in. An M2 in an iPad Air is very likely not the same as an M2 in an MBP or Mac Studio.

The GeekBench [1,2] benchmarks for M2 are:

Single Core: iPad Pro (M2): 2539 Macbook Air (M2): 2596 Macbook Pro (M2): 2645

Multi Core: iPad Pro (M2 8-core): 9631 Macbook Air (M2 8-core): 9654 Macbook Pro (M2 8-core): 9642

So, it appears to be almost the same performance (until it throttles due to heat, of course).

1. https://browser.geekbench.com/ios-benchmarks 2. https://browser.geekbench.com/mac-benchmarks

A Ryzen 7840U in a gaming handheld is not (configured) the same as a Ryzen 7840U in a laptop, for that matter, so Apple is hardly unique here.

The manufacturer often targets a tdp that is reasonable for thermals and battery life, but the cpu package is often the same.

Yeah, but the difference is that you usually don't get people arguing that it's the same thing or that it can be performance competitive in the long run. When it comes to Apple stuff, people say some irrational stuff that is totally bonkers...

Surprisingly, I think it is: I was going to comment that here, then checked Geekbench, single core scores match for M2 iPad/MacBook Pro/etc. at same clock speed. i.e. M2 "base" = M2 "base", but core count differs, and with the desktops/laptops, you get options for M2 Ultra Max SE bla bla.

Likely there is also a smaller battery as the iPad Pro is quite a bit thinner