These desktop CPUs became available in the early 2000s, with significant releases starting around 2003.
These desktop CPUs became available in the early 2000s, with significant releases starting around 2003.
I question whether ARM CPUs will become accessible for building PCs from scratch. They seem more likely to appear in ready-made designs. Look for All-In-One designs or compact NUC systems. Also, I’m skeptical about strong interest in custom ARM machines at the moment, since there are no operating systems available beyond ChromeOS, various Linux distributions, and a limited Linux version for ARM, which we can’t legally purchase (excluding MacOS for obvious reasons)
Thank you for your inquiry! I'm not sure if x86 will be replaced soon, but it's evolving rather than disappearing completely.
Who can predict what lies ahead? I believe competition will remain strong for a long time. It's not just about x86; it's the clash between RISC and CISC. While I don't think CISC will disappear completely, I imagine RISC could become a viable alternative depending on the tasks or perhaps a new hybrid design in the future.
It's quite lengthy here. We shouldn't aim for such a distant timeline—this won't happen until later than expected.
The focus should shift away from the ISA concerns. What truly matters is the core µArch. I wouldn't even think about high-performance ARM cores being RISC anymore—their decoders are overly complicated and just introduce instructions without real advantages, except to simplify compiler work.
I do not see x86-64 been replace with anything for a long long time. To be true fully they a lot of hype on ARM, and it just hype. There no need for desktop to ever change from x86-64 to ARM. Also note that ARM is not new but very old. It bit like big small core thing, it great for mobile and laptop but pointless in desktop. There just too much software that work only on x86-64 for people to move to ARM.
Well, "RISC" serves more as a branding concept than a technical standard, and you can trace its roots back four decades when the 6502 was labeled "RISC." It also refers to similar opcode lengths that simplify decoding. This approach gives RISC an edge in many aspects when other factors match. Apple leverages distinct benefits like moving away from 32-bit architecture and utilizing TMSC's 5nm technology exclusively. X86, on the other hand, has invested heavily in what was once considered outdated two decades ago. Ultimately, it never truly competed across most sectors—smartphones, tablets, automotive, IoT—leaving Windows and gaming consoles as its primary platforms.
Its going to be a while before we see something. I mean technically there is a Surface Tablet that runs on a Snapdraggon CPU. The issue is, Qualcomm is no Apple. Their CPU's are not really built for desktop use like the M1. And seeing how Apple is unlikely going to share its silicone with anyone we have to wait for a company to recreate what Apple did for the PC's. I personally dont think Qualcomm cares enough to really want to be competitive in the desktop market. I mean many phones and tablets already use their silicone and they really have no reason to care about the desktop market, when they pretty much own a good chunk of the mobile market. Now keep in mind that Nvidia has their Tegra ARM CPU which is used in the Nintendo Switch. And I guess AMD is developing its own ARM CPU, AMD does have CPU design under their belt, so maybe in a generation or two they might have a good contender. The other side of the coin is Windows and Windows software was developed for X86. No one is going to go back and rewrite a lot of this old software. The fact is the market share Microsoft does have is really only due to the legacy support that Windows has. While a lot of stuff has moved to the cloud, there are still companies out their using old outdated software because why fix what's not broke. Some of that said software was developed in house and they dont want to pay to update it. While there is emulation you can do to run x86 on ARM, Microsofts implementation is not that great, as in no where close to Apple's Rossetta 2. Even then emulation has its limits. For Microsoft to move to ARM it would have to give up on legacy support. While they could do short term legacy support like Apple is doing with Rossetta 2, they would be stuck if they did it long term. I would say a good portion of their customers are their customers due to legacy support. So while Microsoft wants to have ARM computers, they have to treat them like a secondary project because their customers dont want change. Keep in mind the backlash they got when Windows XP got cut. Because a lot of especially old people where like "You will have to drag my Windows XP out of my cold dead hands".
I believe the emphasis on the M1 being designed for desktops was exaggerated. In reality, it's essentially a modernized version of the last year's iPhone chip with some minor adjustments (more details on I/O). The majority of the innovation happens in software, and Apple has maintained control over the platform, preparing for this shift over the years. Most of the older software still functions well on older hardware, which means many smaller companies are still running on 10-year-old PCs. While large corporations may still rely on a three-year update cycle, they will eventually recognize that an affordable Raspberry Pi-style board can match the performance of more expensive systems. They won't use outdated software on laptops while working in the field. Consequently, the demand for CPUs and their profit margins will decline as the market shifts toward x86 solutions.