Yes, and its been that way forever. Back in the RISC workstation days when you'd see PA-RISC or Alpha or POWER gain a temporary advantage, it never moved the market much - the rule of thumb was you needed a 2x advantage in performance or half price for same performance to get people to consider going through the hassle of migration. The market shares moved around a bit but didn't make any huge moves until Intel was able to deliver that 'half price for same performance' when their fab advantage became too much for the RISC workstation model to survive.
Some people want to point to these Ampere results and say "well, ARM has caught up, and costs less, x86 is doomed". Things don't happen that quickly, the people making the multi million dollar purchase decisions are cautious by nature, and aren't going to make such a choice without a lot of study and preparation and seeing others with a higher risk tolerance take the plunge first.
I'd put it differently.
People talk about diversity in CPUs as though that's some sort of wonderful flower that everybody wants -- OEMs want to support 5 different CPUs, customers want to own seven different OS's running on 8 different cores.
Utter nonsense! Diversity is a massive expensive PITA -- for OEMs, for developers, for customers.
There are all sorts of consequences of this fact (for how Apple does things, for RISC-C's future, for why MS and Intel made so many mistakes) but the one I want to focus on right now is that a non-obvious consequence of this is that change, when it happens, happens very fast.
People's mental model of CPU change is something like a linear increase in the number of competitor CPUs every year, so maybe ARM server fraction was 0% in 2018, maybe is 90% in 2038, and between those two dates it increases linearly.
I don't think that's correct because of the costs of diversity. Rather what happens is more like supercritical heating -- there will be some period of time during which ARM is clearly better by almost every criterion because it's still the *other*, and supporting two things is a hassle. Then, very rapidly (like over two or three years) that will flip. Individuals and companies will make the experiment, see that it works, solve the stupid lingering bugs and incompatibilities, and over a two or three year buying cycle transition close to their entire fleet.
This has two consequences:
- yes, there is an element of baseline delay. This year and next are that delay. This is the delay as cloud SW gets ported to Graviton, as FOSS gets ported to M1, as one developer after another tries either Graviton or M1 and discovers that not only is ARMv8 pleasant to work with, it also delivers everything promised.
- BUT it also means that when the transitions come, they will come fast, much faster than expected. Look at how opinion about Apple going ARM flipped. For years it was obvious (to those of us who follow Apple and understand the engineering) that it would happen; but most of the world (including much of the supposedly computing thought-leader world) threw up one reason or another why it wouldn't -- x86 emulation would be impossibly slow and work badly; people would demand to run Windows; the Apple CPUs would just have to be slower than Intel; blah blah. All vaguely plausible -- if you know nothing about Apple, nothing about ARM, and are uninterested in ever learning.
But what happened? Apple ships M1 and over the course of a month these people are forced to confront reality, to give up their illusions. And sure, you get the idiot denialists still going on about one stupid complaint or another. But most people do in fact accept reality when it bangs them hard enough. In other words you didn't get a slow growth in acceptance that M1 was better; you got a total flip in people's world pictures within a month.
This does't mean everyone will buy M1's tomorrow. People have buying schedules, they may (temporarily) have compelling reasons to stick with what's known and works, they may have reasons to wait for the M1X or M2 or even M3. But the argument two months ago was "I can't see how Apple switching to ARM makes sense"; within two months that changed to "I don't want to buy an ARM Mac because reasons, but I can see how it will be very popular".
That's what will happen with ARM generically over 2021.
- Graviton3 will make the AWS proposition so compelling that no-one will want to stay on the x86 virtual machines
- M1X then M2 will show what can be done on the desktop if you're willing to burn a little more power
- Even ARM's cores lagging behind Apple like X1 and V1 will be good enough that people will start to consider their usability in various situations (Graviton2 and Altra, with pre X1/V1 cores, show this)
- the FOSS missing pieces will be filled in
- MS will finally (ten years late) get their ARM Windows and software story together (M1 has basically forced this)
By end of 2021 to be in denial about ARM's future will be like being in denial about M1 today -- it's possible if you want to behave like an idiot, but it won't be a mainstream position.
At which point it's just a question of the length of buying cycles.