The discussion seems to have settled down to "M1 is great! But you'll soon be able to get a Zen 3 CPU that's just as good." Which may or may not be true, but:
- Apple has a material (and apparently perceptible) performance advantage against the majority of the laptop market - which will do wonders both for their and Arm's brand.
- The M1 has killed the idea (which I've heard a lot) that you can't get performance out of an Arm core.
- There is now a credible desktop platform on which to develop for Arm in the cloud.
I think the real impact of this will be in 2-3 years and it's not good for Intel.
> The discussion seems to have settled down to "M1 is great! But you'll soon be able to get a Zen 3 CPU that's just as good." Which may or may not be true
Even if it is, I doubt there will be a Zen 3 laptop out that has the screen, keyboard, trackpad, and case quality of the MBAir. Even if it were slower, literally no one is building machines of that physical quality.
Some are attempting to compete, like the Dell XPS, but they're still falling short. They're high quality laptops, yes, but people who think they're "just as good" need to take a better look at the current MBAir - and this has 0% to do with the chip. Nobody is making laptops this nice.
I'm actually thinking about picking up an Intel MBAir (the one that just got discontinued) as it's likely the single best Linux-able laptop (no longer) on the market, even if it needs to do an opaque, closed-source, online activation each time you wipe the disk (to reactivate the T2).
> The discussion seems to have settled down to "M1 is great! But you'll soon be able to get a Zen 3 CPU that's just as good."
It feels worse than that to me. Those people are actually saying "But Zen 4 in 2 years will be just as good!" because their claims depend on combining IPC improvements from Zen 3 _and_ a 5nm process that AMD has no plans to switch to for Zen 3. It's a weird bit of gymnastics in order to reach emotional parity with the M1 widely available right now today.
Without commenting on the gymnastics (I'm not qualified) the big issue for me is about actually having the machines in the market. Some of the Ryzen laptops seem to be impossible to buy and who knows how much TSMC capacity AMD has (after the Xbox and PS cores) plus how much and for how long have the PC laptop manufacturers already committed to Intel.
Well that big impact will only materialize if Apple can convince developers to make apps for this platform, something they have not been doing well as it is. (with intel offerings)
They need to stop ceding the desktop game market to windows/amd/intel. When Catalina dropped it remove a lot of software, productivity, utility, and games, had many examples where they no longer ran. Worse some developers just wrote off the platform.
Even Apple's M1 debut could not line up more than two or three known developers. Most of it was from developers people had to resort to google for and even then some didn't show up in search lists.
No, Apple's problem with the Windows platform is and has always been software. Apple never seemed keen on pushing the Mac as a platform with iPhone and iPad being the focus on so many levels. Now that they have a truly impressive offering we have to hope they go recruiting developers to the platform because Rosetta won't be here forever and it doesn't solve the availability issue.
- Apple has a material (and apparently perceptible) performance advantage against the majority of the laptop market - which will do wonders both for their and Arm's brand.
- The M1 has killed the idea (which I've heard a lot) that you can't get performance out of an Arm core.
- There is now a credible desktop platform on which to develop for Arm in the cloud.
I think the real impact of this will be in 2-3 years and it's not good for Intel.