To TOS' point, yea, I agree with that. RAM is not what it used to be with the M1. The unified memory model is a revolutionary concept that really hasn't been touted enough. Traditionally, CPUs excel at "narrow-and-deep workflows," while GPUs are best suited for "wide-and-shallow, with SIMD meant to cover some use cases in the middle. That all goes out the window with the M1. Now you can write code that mixes/matches these elements at will without having to shuttle data around, or query the host system to see if it supports the right proprietary graphics API, and so forth.
This a big deal. The story so far from the press has been: "Apple's low-end laptop chip hangs with high-end PC desktops in lightly threaded workloads". Once we start seeing apps start to take advantage of the unified memory architecture, I expect those headlines will change to: "Apple's low-end laptop chip demolishes high-end PC desktop in video editing and CAD".
Looking ahead, the even bigger point about the M1 is that Apple has built a platform that combines ALL elements of a system in a way that can be easily scaled both UP and OUT, from the lowly Apple Watch all the way up to the yet-announced Mac Pro, and they've done so in a way that allows them to get great yields across the board, while cutting out vendors in the middle. At the same time, this platform provides new pathways for optimization that are simply not possible under traditional computing models. Apple's lead on paper might look small, but in reality, Apple is YEARS ahead competitors in the x86 ecosystem. Secretly, they all know this. It's why NVIDIA is acquiring SoftBank, and it's why Microsoft announced their own SoC program.
Hack Pro: 4.6GHz i7 980X, RTX 3080, 24GB RAM, 2x 2TB HDD, 8TB SMR, 0.25/1TB SSDs, ZR30w, HTC Vive
MacBook 12" 1.2GHz Core M, 512GB Flash, 8GB RAM