Fully agree. And only way people will pay attention is clearly better price, performance and power use. All at once. And even then it's not clear cut but at least for me it would be an option as a laptop would only ever be a secondary device for me where I might not need all the x86 tools.Microsoft will never treat Windows on ARM as a first class citizen, so ARM laptops are irrelevant to the overall market. People who follow tech like to pretend they matter, to everyone else they will always be as invisible to them as Linux is as a desktop OS. No matter how successful a laptop chip Qualcomm can make, the market is going to yawn and ignore it. It will at best achieve the same token single digit share that Linux desktops like mine have.
They should force hardware vendors to write ARM drivers if they want their stuff signed (for new products). That should help...
Forgetting SQ3? Windows on ARM is Microsoft's bet on evolving into Apple No.2. There's a reason Windows 11 is taking so many cues from the MacOS UI. They want to reach a point where consumers find the two OSes so similar that the price conscious majority stops caring about MacOS and feel "cool" using whatever Microsoft feeds them.Any increase in ARM PC sales is essentially a decrease in x86 PC sales, it won't expand the market for Windows. Why should Microsoft create friction with OEMs over something that doesn't make them more money?
Apples customers are not price conscious nor does that come into play when buying an Apple product. They look for the Apple logo and the name of the product ie iPhone or MacBook.Forgetting SQ3? Windows on ARM is Microsoft's bet on evolving into Apple No.2. There's a reason Windows 11 is taking so many cues from the MacOS UI. They want to reach a point where consumers find the two OSes so similar that the price conscious majority stops caring about MacOS and feel "cool" using whatever Microsoft feeds them.
There's a long way to go software wise before Windows on ARM catches up to MacOS for major apps.Forgetting SQ3? Windows on ARM is Microsoft's bet on evolving into Apple No.2. There's a reason Windows 11 is taking so many cues from the MacOS UI. They want to reach a point where consumers find the two OSes so similar that the price conscious majority stops caring about MacOS and feel "cool" using whatever Microsoft feeds them.
Indeed.
They would be in a much better position if for instance they could push out their processors with more modern Cores. The 8cx Gen 3 is ANCIENT, and was old already when it got released; they can't compete that way with AMD or Intel for that matter.
For years ARM has been pushing out designs that allow for 8 core Cortex-X + 4 core A700 series cores. We haven't seen any of such designs on the market yet, and will not happen as Qualcomm still have their exclusivity deal with Microsoft.
And even worse, they want to price their products like flagship, like what they charge for the $1000+ Samsung S23 Ultra; for everything, even if not achieving the flagship-expected performance on PC.
Perhaps is what's in their best interest, keep pumping money from the mobile market; where they are almost a monopoly, instead of spending money on a more competitive market with lower yields like the PC.
Microsoft absolutely cares about it which is why they built an entire binary platform for the express purpose of easing compatibility pains - Windows 64EC. They also ported Visual Studio to Arm64 last year along with .NET 7 with a fairly intricate focus on Arm64 performance. You confuse "first class" with their hardware partners failing to execute thus far. There's no need to put this as some zero sum binary choice: They can support both, and they are. Almost the entirety of BUILD 2022 was focused on bringing Arm up.Microsoft will never treat Windows on ARM as a first class citizen, so ARM laptops are irrelevant to the overall market. People who follow tech like to pretend they matter, to everyone else they will always be as invisible to them as Linux is as a desktop OS. No matter how successful a laptop chip Qualcomm can make, the market is going to yawn and ignore it. It will at best achieve the same token single digit share that Linux desktops like mine have.
I'm not sure what you mean here seeing as successful would directly imply they sold well, but if you mean "no matter how performant, efficient" I don't know that I agree with that.No matter how successful a laptop chip Qualcomm can make,
Ignoring the driver comment this is replying to but addressing the independent claims in this:Microsoft has no incentive to do that unless they care about the success of ARM as a PC platform. Why should they? ARM doesn't bring anything to the table they don't have with x86, and they already have a competitive market for x86 now that AMD has put the Bulldozer days behind it. Any increase in ARM PC sales is essentially a decrease in x86 PC sales, it won't expand the market for Windows. Why should Microsoft create friction with OEMs over something that doesn't make them more money?
Microsoft would get a lot of complaints from OEMs if they tried what you say, because that it will increase their cost without zero prospect overall sales will increase for the same reason it won't increase Windows sales.
Maybe OEMs decide to tell Microsoft "OK we're done updating our drivers for all our existing products to avoid having to port them to ARM, and if you make changes to Windows that break them, when customers call us we will be blaming you!"
You think MS is happy about Apple's creeping marketshare of premium with the M-series? They have to have something in thin and light that can seriously compete with those SoC's on idle and "real-world" (read: light use interested by heavy use and/or background tasks) use and an increasing proportion of both activity and sales are brought by mobile AKA laptops, not desktop DIY nerd stuff.Any increase in ARM PC sales is essentially a decrease in x86 PC sales, it won't expand the market for Windows. Why should Microsoft create friction with OEMs over something that doesn't make them more money?
Yes, but also no. It's surprising how much is already ported for a developer for instance. Obviously a lot of Adobe stuff has been ported for a while now but the fact that IntelliJ Ultimate has an Arm64 .exe maintained to the latest version and Python has a full Windows Arm64 install since last fall is impressive for example.There's a long way to go software wise before Windows on ARM catches up to MacOS for major apps.
Oh no doubt the remaining software that already has ARM MacOS ports will come to WARM faster, especially for the larger software suites a la Autodesk Maya/3DS Max, SideFX Houdini etc.Between Apple's initial push making a huge difference for porting at all to Arm64, and then MS's Arm64EC + toolkit availability in VS, and probably Qualcomm working with adobe, it's not good but it's surprisingly to the point where e.g. I could actually use it if someone released capable hardware that had even a modicum of interest.
Yep. Like I said, if you look where we are with development having VS, WSL, Python, plenty of OpenJDK and similar ported, MS throwing time into .NET performance for Arm, the ground structure is there.Oh no doubt the remaining software that already has ARM MacOS ports will come to WARM faster, especially for the larger software suites a la Autodesk Maya/3DS Max, SideFX Houdini etc.
Once they get past that hump of the big stuff I think we will see more smaller devs following suit.
Thankfully foobar2000's dev added native ARM64 support in the recent v2 release, so my music is already covered once all the plugins I use are ported too.
That MediaTek Nvidia thing is BS. They had a dGPU deal and it’s over for Chromebooks now. It’s just Nvidia on their own for MS’s surface like and most likely elsewhere too come 2025, per Charlie.Qualcomm have claimed that Microsoft and other "PC" OEMs want product in hand 12+ months prior to release, compared to less than 6 months for mobile phone OEMs. This was their explanation for why the 8cx was a generation behind phone SoCs in terms of the cores used.
As for the exclusivity deal, rumors have been around for a few years that it expires in 2024. I haven't seen any official confirmation, although there was a report earlier this year that MediaTek and NVIDIA have an agreement to make a Windows-on-ARM compatible chipset for 2024.
Thanks for providing me with this context. It's sad that ARM releases a core design that is "kind of competitive", but then it take a bunch of time for chip makers to work it into a sellable product, and then to be integrated by OEMs. The 8cx saga has been really unfortunate.Qualcomm have claimed that Microsoft and other "PC" OEMs want product in hand 12+ months prior to release, compared to less than 6 months for mobile phone OEMs. This was their explanation for why the 8cx was a generation behind phone SoCs in terms of the cores used.
You think MS is happy about Apple's creeping marketshare of premium with the M-series? They have to have something in thin and light that can seriously compete with those SoC's on idle and "real-world" (read: light use interested by heavy use and/or background tasks) use and an increasing proportion of both activity and sales are brought by mobile AKA laptops, not desktop DIY nerd stuff.
Thanks for providing me with this context. It's sad that ARM releases a core design that is "kind of competitive", but then it take a bunch of time for chip makers to work it into a sellable product, and then to be integrated by OEMs. The 8cx saga has been really unfortunate.
I'm curious, do AMD and Intel provide PC OEMs with a product in hand at least 12 months prior to release?
I'm also wondering if the newer ARM strategies, such as designing full SOC, could potentially help speed up the process of bringing their chip designs from the official release to the actual sale release.
I'm curious, do AMD and Intel provide PC OEMs with a product in hand at least 12 months prior to release?
I'm also wondering if the newer ARM strategies, such as designing full SOC, could potentially help speed up the process of bringing their chip designs from the official release to the actual sale release.
Huawei were pretty prompt in the years they brought a new core into their Kirin SoC.It's sad that ARM releases a core design that is "kind of competitive", but then it take a bunch of time for chip makers to work it into a sellable product, and then to be integrated by OEMs
It seems this year Qualcomm and Mediatek are following thatHuawei were pretty prompt in the years they brought a new core into their Kirin SoC.
Usually October ish the same year the core was announced.
Qualcomm often announce a new SoC the same year around Novernber/December, but I can't recall an OEM ever actually shipping a product based on it in the same month.It seems this year Qualcomm and Mediatek are following that
Thinking further on it Huawei also had the benefit of a product stack closer to Apple's given that they made both the phone and the SoC, so they could plan both in succession for a faster release cycle.It seems this year Qualcomm and Mediatek are following that
Xiaomi, Motorola and other Chinese brands most of time release immediately after Qualcomm anouncement, Xiaomi 12, 13 and this year Xiaomi 14 are good examples.Qualcomm often announce a new SoC the same year around Novernber/December, but I can't recall an OEM ever actually shipping a product based on it in the same month.
It's usually March+ of the next year before we actually see the new OEM phones ship.
I think Huawei didn't spread their engineering corps as thin as Qualcomm does though, so they were basically just architecting the high end chip only making for a more focused effort on their end.
Not so certain about Mediatek's timeline though, I don't follow any of the manufacturers that use their high end Dimensity 9xxx SoCs.
That's definitely in Geekbench 6 though if true.
Nuvia core in Snapdragon 8 gen 4 reportedly reaching 2600-2900 points.