I don’t agree with that. Two entirely seperate scenarios, in two entirely seperate application areas (smartphones vs. Laptops, and Qualcomm vs. Intel).
There is a world of difference between the reasons for each. Apple went with its own designs for the iPhones because they couldn’t find what they wanted if the shelf. They wanted to add dedicated accelerators for various things like cameras, secure enclave, Face UD, as well as getting the fastest possible CPUs.
The AS SoCs for the Mac are being forced by other factors, the major one being Intel’s repeated failure to deliver on process improvements, and unavailability of volume (as was alluded to in last week’s Analyst’s/Earnings meeting). Add to that, the A series SoCs finally getting to the point where acceptable performance under MacOS was achievable. There was no real intent to reduce parts cost, it just happened out of the need to get away from Intel as a supplier (and that Intel charged Apple a very premium price for CPUs) due to non delivery on Intel’s own timeline, and the relentless improvement in A series SoCs.
The A series SoCs are more expensive than the Qualcom, or for that matter, Samsung Exynos type parts. But Apple gains some functionality that they want and is unavailable in off the shelf parts.
The Mac SoCs will be cheaper, overall, than the Intel parts. This is due to a number of factors. Firstly, Intel was charging a lot of money for CPUs (I have heard at least $300) plus more for the accompanying chipset. No external chipset with an SoC. Secondly, the R&D for the basic building blocks, CPUs, GPUs and common accelerators will be spread ocer not just the 4-5M Macs sold each quarter, but 10M iPads, and 50M+ iPhones as well.
There will be a net BOM cost reduction, and more to be saved in manufacturing, from simpler and easier to manufacture logic boards, to reduced cooling requiremts (fewer fans) and maybe smaller/cheaper batteries in the laptops.