IDK maybe it does because apparently some people don't understand how google, ampere and amazon all do what they do. RISC vs CISC has no impact other than adding one more stage to a chip and ARM literally runs UOPs anyways so its not functionally different from x86 other than it has slightly less instructions. Even if ARM can decode some instructions faster, why couldn't any CISC design copy the same functions in its decoders? Do you know what an FPGA is and why its relevant? Have you ever heard of arduino or raspberry pi? Have you ever wondered why there are so many different types of processors out there running at all different kinds of bit depths and clock speeds? Because its about the market. None of these other companies are making high performance ARM chips, they're all budget solutions to reduce operational expenditures for working in simple server transactions.
.............
I really don't see ARM breaking through in laptops, tablets and phones are also extremely fast for basic web browsing tasks and even light gaming. Like, I have had a ton of laptops over the years, and I personally wish they could have all been as fast as my desktop. This meme about ARM popping up is always so ridiculous to me. Think about it like this, people are upset with core ultra 200 right now right? You think customers are excited for CPUs even slower than intel's current release? Just because they last as long as an iPad from 6 years ago? Nah.
I don't hate ARM at all but I hate this delusion that its somehow got inherent leverage from its physics alone. Its all about the market segments. Nobody targets high perf ARM chips because unless you're already inside a walled garden with non-stop updates like android or macos/ios, people want to run their old code. Just like how there isn't any movement or desire to port windows to a raspberry pi or like why the Amiga A1222 exists at all. Same goes for apple, nobody mass market is buying a mac because of the chip, they buy it because of the OS and the ecosystem. Hence why literally every intel mac owner doesn't mind shelling out a bunch of money for slightly cooler temps and more battery life, they aren't concerned with software unlike most other consumers in the PC markets.
First time posting, but I've been a reader here since 1997. So I'm not new. But the unseriousness of this post compelled me to finally register after all these years.
I won't argue architectures, but ARM has already broken into laptops, and you apparently haven't been paying attention since 2020 because they are far more than light web-browsing machines. The M3 Max is a top-tier performer in a wide variety of normal user workloads - the benchmarks are readily available. It often uses 1-4-1/5 of the power of its competition for dozens of workloads! 1/5! The M1 Max itself may be the most revolutionary and memorable laptop in the last 15 years. A desktop replacement that seems to never get hot or turn on a fan (it does, in theory, but it's nigh unnoticeable).
And as for software under OS X, come on. Don't be silly. There's a tremendous and wide ecosystem. Not every workflow is doable, but the choices are nevertheless very broad. And, Windows itself doesn't cover every workload either.
As for PCs, the trends are moving away from x86. Qualcomm isn't done. AI processing is CPU agnostic. Microsoft is a serious company and they and Qualcomm both have invested hundreds of millions into the ARM experiment on Windows. It's also easy on a Macbook, by the way, to run Windows 11.
One can make the case that the Surface with the Snapdragon Elite is the best PC laptop option on the market for a professional user; at minimum, it's probably the coolest and quietest for anything at its performance level. That isn't "market segments." Furthermore, Lunar Lake and Meteor Lake are proof that one doesn't just shift market segments easily - the former is being hailed because it kind of matches the Macbook Air (lol). Intel has simply invested in other areas. Had they invested in efficiency...who knows. Maybe they would be a better M series competitor. Or maybe they never had it in them.
Taking a step back, the iPhone changed everything, good and bad (including culture!). Trends take time to develop, but consumer behavior and preferences around computing ain't about to go back to loud, noisy and hot. Desktops with any performance are passé outside of technical workloads in offices. 1.5 hour laptop workstations are also passé. Finally, the DIY gaming desktop is a regressing niche. A big niche, but a shrinking one all the same as the market metrics reveal. Note that the largest game market by revenue is far and away on phones.
My crystal ball tells me x86 will go the way of the dodo, and it may accelerate if/when Intel is acquired, something I reasonably expect in the next 2-3 years if the government doesn't block it. It's already easy to get 95% of the typical usage from Windows 11 ARM. That's a game-changing development. I personally run "old code" on that thing every day. Niche cases requiring deeper x86 compatibility will take a long time to die (hey, there are still Amiga users out there), but die they will when superior hardware just overwhelms with better options.