New Arm v9 announced

That's great for those Acers, but I've got 12 years of accounting data showing that in our cases that our PC's by the time I have MDM, AV, and Backup software in their + time required to deal with user issues are undeniably more expensive than our Mac purchases spread over their lifetime. A conclusion similarly reached by a very large part of the Educational system which is why they are so prevalent there. This was a similar conclusion reached by IBM who found that purchasing and maintaining Apple devices was almost half the cost of their existing Linux-based PC platforms they were running previously. I'm obviously not bringing in Apples at this stage for heavy workstations and we're not even going to pretend that Apple Server is a viable anything, but in the hands of users, the data shows them cheaper over their full lifespan. That said I like my Dell G5's I bring in for my department because it's now also my department's responsibility to deal with the security cameras on the busses and buildings so I am constantly having to re-encode and video and I totally need a laptop with a 2080 for that, so much faster...

"At IBM, one of the largest Apple-using companies with 290,000 Apple devices, a 2016 study found that the company was saving up to $543 per Mac compared to PCs over a 4-year lifespan. Forrester Research came up with an even higher number, showing that Macs cost $628 less over a 3-year lifespan."
https://mbsdirect.com/mbs-blog/article-forrester-research-and-ibm-studies-show-macs-are-cheaper-than-pcs#:~:text=At IBM, one of the,over a 3-year lifespan.

What works for one organization does not necessarily work for another. We no longer support Macs. If a user chooses to connect one they are segregated and all service is up to them.
 
What works for one organization does not necessarily work for another. We no longer support Macs. If a user chooses to connect one they are segregated and all service is up to them.
Absolutely.
Machines should be looked at from a number of perspectives. I tend to agree with Lakados in terms of total cost of ownership and lowered maintenance as well as just being faster and more stable for longer as being primary reasons to run macOS machines in a workplace.

However, all of that is irrelevant if you're a Solidworks or 3D Studio Max house as an example. If you need particular pieces of software or you rely on very specific infrastructure to deploy updates and manage software or user accounts that is Windows only then obviously buying a WinTel machine makes a lot of sense. Still, it's telling that there is a high amount of Mac users at IBM.
 
Not really. One major step would be to make an open platform for ARM just like x86's modern day IBM compatible PCs. You still need the software, which ARM does have a lot of software but nothing on the scale of x86 Windows. Did you know there's an ARM Windows? Nobody knows about it because nobody in their right mind would use an ARM based Windows laptop over an x86, because of the sheer amount of software available on it. Microsoft can't take down x86 Windows. Linux can't get people to switch and make x86 Linux apps. Steam had to make Proton so that x86 Windows games would run on x86 Linux, because nobody was porting their games over to x86 Linux.
I run a number of x86 software packages on my Pi400 using Box86 no problem. Furthermore, you run ARM hardware to get away from Windows. Windows 10 struggles to run on a 4GB x86 based system, let alone a ARM based system with 4GB of ram running off an SD card.

Win32 is an API, Linux is simply now Win32 compatible. What's interesting is in many cases I'm actually getting better performance under Linux than native Windows when gaming.

It's obvious that the world is moving on from Wintel.

*Replying from my 2.3Ghz CPU/800Mhz GPU Pi400 running TwisterOS off USB3.0 SSD that makes a fantastic desktop machine for next to no $$. Furthermore its silent with stock fanless cooling.
 
Last edited:
I run a number of x86 software packages on my Pi400 using Box86 no problem.
Old DOS games don't count because they're old... and DOS.
Furthermore, you run ARM hardware to get away from Windows. Windows 10 struggles to run on a 4GB x86 based system, let alone a ARM based system with 4GB of ram running off an SD card.
You run Linux if you want to get away from Windows. If you're using ARM then you want to get away from software. If you're using ARM Mac then you just want to get away from electronics.
Win32 is an API, Linux is simply now Win32 compatible. What's interesting is in many cases I'm actually getting better performance under Linux than native Windows when gaming.
For 32-bit, sure. For anything modern and I'm having trouble. I still can't get Red Dead Redemption 2 to work on Linux. Other people have but I'm clueless to know what they did.
It's obvious that the world is moving on from Wintel.
Yet, nobody can find a graphics card for their Wintel machines. If you want to get away from Wintel then Linux and ARM needs to do a lot more to get people to switch. ARM needs motherboards conforming to ATX standards with sockets. Linux needs to get more software ported over instead of depending too much on Wine. Also, Wine and Proton need to merge at some point. I'm in favor of AppImages since nobody can get their shit together when it comes to dependencies.
*Replying from my 2.3Ghz CPU/800Mhz GPU Pi400 running TwisterOS off USB3.0 SSD that makes a fantastic desktop machine for next to no $$. Furthermore its silent with stock fanless cooling.
I have a Raspberry Pi 3 and I've learned that these machines aren't worth the money. The main advantage of these devices is their size. They aren't cheap when you need a power adapter, SD Card, and case. If I want something relatively small and cheap then I'll buy a used Zotac Zbox. That's what I've done because they're more powerful than a Pi 4, plus better graphics driver support from Linux since they use either Intel or AMD GPU's, and also they're x86 so I have access to a lot more software.
 
Yet, nobody can find a graphics card for their Wintel machines. If you want to get away from Wintel then Linux and ARM needs to do a lot more to get people to switch. ARM needs motherboards conforming to ATX standards with sockets. Linux needs to get more software ported over instead of depending too much on Wine. Also, Wine and Proton need to merge at some point. I'm in favor of AppImages since nobody can get their shit together when it comes to dependencies.
I agree with the basic premise of this, but I wouldn't put too much stock in GPU shortages. Those are triggered as much or more by crypto miners and pandemic-related supply issues as they are gamers. And you don't need ATX for laptops, which dominate PCs these days. I would say the world is moving on from Wintel, but right now it's only at a slow pace — Apple and the server world are the main drivers right now. On top of hardware support, Microsoft also needs to do much, much more to ease the transition and reduce its dependency on a handful of Qualcomm chips.
 
I agree with the basic premise of this, but I wouldn't put too much stock in GPU shortages. Those are triggered as much or more by crypto miners and pandemic-related supply issues as they are gamers.
Whatever the cause is, the main thing to understand here is that PC gaming is booming. Specifically X86 Windows gaming. The only move towards ARM right now is Chrome, and that's because Chromebooks are cheap.
And you don't need ATX for laptops, which dominate PCs these days. I would say the world is moving on from Wintel, but right now it's only at a slow pace
If you count cell phones and tablets then sure, but that's like saying AAA gaming is dying because most people are playing mobile games which are mostly bejeweled clones. You don't see much of any laptop or Desktop running ARM, for a reason. While most people are on laptops, that's because of the standards like ATX that brings it all together. If you want a more powerful ARM laptop then what do you buy? You need standards to unify the platform.
— Apple and the server world are the main drivers right now.
Apple? Why Apple? They have 10% of the PC market and they don't exist in servers. Linux does exist in servers, like all the servers, but hardly exists in the PC market. If you add phones and tablets then sure but that isn't the same thing, for so many obvious reasons.
On top of hardware support, Microsoft also needs to do much, much more to ease the transition and reduce its dependency on a handful of Qualcomm chips.
Microsoft can't do anything to ease nothing. You aren't going to get people to jump on ARM unless you offer standards. You need a standard way to install an OS, which ARM devices are sorely missing. You can't buy an ARM laptop and simply install Linux. It usually requires a lot more work to do something like install Linux on a Chromebook. You need desktop ARM motherboards with sockets and memory slots.

If people are moving away from Wintel or x86 then that's because ARM is cheaper. Which says more about the disconnect the market has with its customers than ARM vs x86.
 
Old DOS games don't count because they're old... and DOS.
Well I'm not running old DOS games, I'm mostly running Windows applications. Considering usernames, I'd expect yourself to be running something like Dos Box in all honesty.

You run Linux if you want to get away from Windows. If you're using ARM then you want to get away from software. If you're using ARM Mac then you just want to get away from electronics.
All the software I need is available. What you're saying is: You can't run Microsoft Office. I don't use Microsoft Office, I use Libre Office on all my devices and don't experience compatibility issues as they've mostly all been rectified. I've got electronics design software that runs fine under ARM, remote desktop tools available under Windows that run fine under ARM, video conferencing software available under Windows that runs fine under ARM, Email clients that sync via my Oauth authenticated email and calendar platform just fine...

Essentially, I've got everything I need regarding work and downtime. ARM is far from a software wasteland, and community support is advancing the platform quickly.

For 32-bit, sure. For anything modern and I'm having trouble. I still can't get Red Dead Redemption 2 to work on Linux. Other people have but I'm clueless to know what they did.
Win32 is simply describing the API compatibility layer present in Windows using .DLL's to communicate with the kernel, Wine and associated software take advantage of this layer. Most people are using the 64bit variant of Wine, meaning Wine is not limited to 32bit software.

I have a Raspberry Pi 3 and I've learned that these machines aren't worth the money. The main advantage of these devices is their size. They aren't cheap when you need a power adapter, SD Card, and case. If I want something relatively small and cheap then I'll buy a used Zotac Zbox. That's what I've done because they're more powerful than a Pi 4, plus better graphics driver support from Linux since they use either Intel or AMD GPU's, and also they're x86 so I have access to a lot more software.
My Pi400 complete with the keyboard (naturally), mouse, power adapter, mini HDMI to HDMI cord and manual cost me $120.00 in a country where hardware is nowhere near as cheap as it is in the US. My Pi3 cost me about $80.00 with the Pi case and power supply. Both came with SD cards. Even here, that's cheap.

Furthermore, my main workstation also runs Linux and an Nvidia GPU with Nvidia driver support that usually performs better than native Windows.

You seem to think games define a PC, gaming is but one thing you can do with a PC, and considering the Pi was in no way built with gaming in mind it does an admirable job - Even when running software coded for Windows.

The future isn't Wintel. No one platform in the history of tech has remained top Dog forever.
 
Last edited:
Microsoft can't do anything to ease nothing. You aren't going to get people to jump on ARM unless you offer standards. You need a standard way to install an OS, which ARM devices are sorely missing. You can't buy an ARM laptop and simply install Linux. It usually requires a lot more work to do something like install Linux on a Chromebook. You need desktop ARM motherboards with sockets and memory slots.
What?!

There is a standard, it's called Pi Imager. Remember: Linux is not Windows, it doesn't need to behave identically to Windows and Pi Imager is in no way hard to use. The only reason why you can install the OS on another machine is by virtue of the fact that the 'hard drive' is an SD card that can be simply and easily removed from the system and inserted into another PC.

I run Pi Imager on my Pi, then swap either SSD's or SD cards on the same Pi. How is that difficult? I can swap my entire OS by popping out one SD card and inserting another.
 
What?!

There is a standard, it's called Pi Imager. Remember: Linux is not Windows, it doesn't need to behave identically to Windows and Pi Imager is in no way hard to use. The only reason why you can install the OS on another machine is by virtue of the fact that the 'hard drive' is an SD card that can be simply and easily removed from the system and inserted into another PC.

I run Pi Imager on my Pi, then swap either SSD's or SD cards on the same Pi. How is that difficult? I can swap my entire OS by popping out one SD card and inserting another.
I think DukenukemX means that we need a universal boot and platform standard for ARM, similar to what we have with IBM-compatible BIOS/UEFI for x86-16 through x86-64 - it needs to be universal, and not just an imaging tool.
We can make images with dd in Linux as well, that doesn't mean anything for the boot loader or the platform support of the disk image and OS we are imaging to said disk.

Obviously there are x86-64 platforms out there that aren't PCs (current-gen consoles), but once an OS like Linux is running and booted, it can then execute Linux-based x86-64 code.
We need just that for ARM platforms.

Your Pi Image works great for Raspberry Pi platforms, but not for the NVIDIA Jetson, the MACCHIATObin, or any other ARM-based platform - that's the point, every ARM platform is too proprietary and segmented to be considered universal like the IBM-compatible PC x86-16 to x86-64 platform, which has now been around for 40 years.
If ARM is going to truly dominate x86/x86-64 and PCs, it needs that universal and standardized platform, and no more segmented boot loader silos.


There is a standard, it's called Pi Imager. Remember: Linux is not Windows, it doesn't need to behave identically to Windows and Pi Imager is in no way hard to use. The only reason why you can install the OS on another machine is by virtue of the fact that the 'hard drive' is an SD card that can be simply and easily removed from the system and inserted into another PC.
I am going to touch on this part again.
Linux doesn't have a universal boot loader for every single platform, regardless of the ISA, and each one is platform specific.

Yeah, I can easily make an image of an SD card and duplicate it or it's image to any other disk, be it SSD, HDD, CF, or whatever - that isn't the issue.
Go ahead, try and get that Raspberry Pi image to boot on another ARM platform and see how well that goes for you. ;)
 
I think DukenukemX means that we need a universal boot and platform standard for ARM, similar to what we have with IBM-compatible BIOS/UEFI for x86-16 through x86-64 - it needs to be universal, and not just an imaging tool.
We can make images with dd in Linux as well, that doesn't mean anything for the boot loader or the platform support of the disk image and OS we are imaging to said disk.
Why? The issue isn't that ARM is proprietary, the reality is that ARM is 'open' and open doesn't support the notion of one size fits all - There's a bootloader, and due to the fact that everything is SoC that's really all we need. To standardize means to lock down, we want to learn from mistakes of the past, we don't want to keep repeating them.

If it wasn't for the reverse engineering of the original IBM BIOS, there's every chance that the PC as a platform never would have gone anywhere. Ideally speaking, we don't want anyone to have to reverse engineer anything ever again, and there's nothing to state that an underlying UEFI needs to manage hardware in this modern day and age.

Yeah, I can easily make an image of an SD card and duplicate it or it's image to any other disk, be it SSD, HDD, CF, or whatever - that isn't the issue.
Go ahead, try and get that Raspberry Pi image to boot on another ARM platform and see how well that goes for you. ;)
Equally, there's nothing to state that my Windows 10 install is portable between differing x86/64 platforms either. The only reason this situation has improved is because Microsoft improved Window's ability to download and install the necessary drivers required for the OS to run and even then, while the situation has improved, it's still not perfect.

You can compile Linux for virtually anything, standardized UEFI or not. Once compiled, a packaged distro can be easily created for the layman to install on their device, and there's a plethora of such distributions available.

It's a moot point. ARM is not x86, Linux is not Windows. Android is globally the most popular OS, more popular than Windows - the fact there's no standardized EFI like x86 doesn't seem to be holding the platform back in the slightest.
 
I would agree that if ARM really wants to kick x86 to the curb in the next few years and not after a long slow death. Yes ARM needs a universal socket/uefi open system.
Ideally Nvidia completes their ARM purchase and to make all the regulators happy... builds a open system standard for ARM. Yes best case they ARM creates a standard socket setup.... ARM socket 1-D for desktop and ARM socket 1-M for mobile. Along with a universal uefi type boot system. IF nvidia did that I can't imagine any of the regulation pressure companies like Qcomm are throwing out would hold up for long. The up side for Nvidia would be clear although not pure profit... ARM and nvidia would have one more thing to license cause open or not I can't imagine they wouldn't take the opportunity to score a few points of everyone making a ARM open socket system. Then you get the Asus Gigabytes ect of the world making ARM MOBOs. CPUs from Nvidia could lead the way in the same way Intel did for years. I can't imagine AMD wouldn't jump in, as would Qcomm Samsung and who knows who else.

IF Nvidia actually went that route. Yes ARM would go from hobby project (I'm sorry Maz the PI is super cool but that is what it is) to serious platform. Down that road I could see Nvidia displacing Intel a lot faster then anyone could imagine. Microsoft would be more then happy to sell a new windows ARM... and heck perhaps Linux would actually get a good hold. IF Nvidia was really really really serious. They would take a page out of Intels book and build their own Linux distro. Intel has clear linux. If Nvidia went ALL in built the open specs for the industry... and then dropped a Nvidia backed Linux distro... perhaps in collaboration with Valve as they pushed the game industry to ARM/Nvidia Linux. I don't know Nvidia could make a huge dent pretty quick. Wouldn't take much for Nvidia to build out a Linux desktop in the same way Google has... build in all the steam x86 emulation stuff, and then push the way its meant to be played team to ARM hard.
 
I think ARM has a lot of potential, and I would love to see an open platform with more players.

Also, interestingly enough. It could be a boon for Linux, as if Microsoft was dropping x86 compat and running through an emulator, well you can do that on Linux already.

Mostly the reason people stick with Windows is because of the legacy software. If Nvidia did it right and supported devs with proper cross-platform tools, it could be a good way to get a solid library of ARM software (for Windows, Mac, and Linux). A clear break.

Also, as I said before, I think Microsoft is considering a Linux-based version of Windows. If they have to abandon Win32 anyway, might be a good time to build their Linux distro (if it is not already underway).
 
What?!

There is a standard, it's called Pi Imager. Remember: Linux is not Windows, it doesn't need to behave identically to Windows and Pi Imager is in no way hard to use. The only reason why you can install the OS on another machine is by virtue of the fact that the 'hard drive' is an SD card that can be simply and easily removed from the system and inserted into another PC.

I run Pi Imager on my Pi, then swap either SSD's or SD cards on the same Pi. How is that difficult? I can swap my entire OS by popping out one SD card and inserting another.
We're not just talking about the Raspberry Pi. Most ARM SoC's are in phones and tablets, with the Raspberry Pi making up a very small fraction of the market. The only other device that could have some potential for standardization is Google's ChromeOS, but even that is a hit or a miss. Fact is most manufacturers have taken the time to use ARM to isolate their customers to their platform. This isn't a problem with ARM itself, just the baggage that comes with it from greedy manufacturers.

For example iOS is terrible because you can't install whatever OS you want on it. Pretty sure jailbreaking is dead on iOS devices since Apple has successfully prevented anyone from doing anything with their devices. Android partitions the storage on devices to prevent piracy, which hasn't worked too well for Google. Because of this, Android devices have multiple partitions and you can't access the Android partition unless you root the device. You can't just install any version of LineageOS, but you need a specially made version just for your device. This is why Android devices are horrible when it comes to updating the OS to the latest version.

Compare all this to x86 which is still using IBM's Compatible setup and you can see why x86 is still so attractive. You can still buy the latest laptop and boot MS-DOS 6.22 if you wanted to. It's hard to move away from the massive library that Windows X86 has. Microsoft themselves has had horrible progress when it comes to getting people onto ARM for this reason. I bet that Apple M1 users are regretting their purchase when they realize that a lot of x86 Windows games no longer work or work poorly. It'll take a lot more than faith for ARM to dethrone X86. This is something I hope Nvidia will solve and not make things worse.
 
We're not just talking about the Raspberry Pi. Most ARM SoC's are in phones and tablets, with the Raspberry Pi making up a very small fraction of the market. The only other device that could have some potential for standardization is Google's ChromeOS, but even that is a hit or a miss. Fact is most manufacturers have taken the time to use ARM to isolate their customers to their platform. This isn't a problem with ARM itself, just the baggage that comes with it from greedy manufacturers.

For example iOS is terrible because you can't install whatever OS you want on it. Pretty sure jailbreaking is dead on iOS devices since Apple has successfully prevented anyone from doing anything with their devices. Android partitions the storage on devices to prevent piracy, which hasn't worked too well for Google. Because of this, Android devices have multiple partitions and you can't access the Android partition unless you root the device. You can't just install any version of LineageOS, but you need a specially made version just for your device. This is why Android devices are horrible when it comes to updating the OS to the latest version.

Compare all this to x86 which is still using IBM's Compatible setup and you can see why x86 is still so attractive. You can still buy the latest laptop and boot MS-DOS 6.22 if you wanted to. It's hard to move away from the massive library that Windows X86 has. Microsoft themselves has had horrible progress when it comes to getting people onto ARM for this reason. I bet that Apple M1 users are regretting their purchase when they realize that a lot of x86 Windows games no longer work or work poorly. It'll take a lot more than faith for ARM to dethrone X86. This is something I hope Nvidia will solve and not make things worse.
First of all, the RPi as a platform is very, very popular in everything from home brewed projects ranging from RGB to HDMI adapters through to IoT devices and emulation - Computing isn't limited to huge towers with RGB lighting, nor are huge towers with RGB lighting always necessary.

And you're 100% right, Apple are doing their best to lock down their ARM based devices, and creating an unnecessary UEFI simply because some ancient x86 platform used it since the IBM XT days isn't going to rectify that - Nor is it necessary. You need a variant of Android that contains the drivers needed for your SoC, ARM is essentially a standardized instruction set.

The only reason the PC as a platform took off is because the actual platform was open, it was the BIOS that was proprietary. The fact that one little rom contained proprietary code was nearly the death of an open platform until someone found a loophole around IBM's copyright. If it wasn't for the reverse engineering (more like reorganizing) of IBM's BIOS, there's every chance the IBM XT and everything afterwards would have lived, and died, in corporate environments and never really made it into the hands of consumers as they were stupid expensive at the time and there were a number of cheaper (and arguably better) products that were more suitable to the home user using 68000 CPU's.

ARM is already dethroning x86. As stated, Windows is globally no longer the most popular OS, and Apple's transition to ARM is very real and the results are (coming from the mouth of someone that hasn't really liked Apple since the death of the IIGS) very impressive. All that remains to be seen is how well will it scale.

Put the might of Nvidia behind ARM, and x86 may just need to pluck a Rabbit out of it's ass. The concept of backwards compatibility is, in many ways, holding Windows back.
 
Why? The issue isn't that ARM is proprietary, the reality is that ARM is 'open' and open doesn't support the notion of one size fits all - There's a bootloader, and due to the fact that everything is SoC that's really all we need. To standardize means to lock down, we want to learn from mistakes of the past, we don't want to keep repeating them.
I'm not talking about ARM being proprietary, and I never said that it was.
There are x86 systems in existence which are highly proprietary and locked down as well - again, I'm not talking about that.

The issue is less the ISA and more the lack of a standardized platform and bootloader, and in this case with what we are discussing is needed, specifically for ARM.
Also, not every ARM CPU or platform exists as an SoC, so that isn't really accurate; being an SoC is only part of the equation, and is nearly independent from the platform itself.

You need to start learning more about how computer platforms function and operate before continuing down this road so you can understand what DukenukemX and I are talking about.
Again, try to boot your Raspberry Pi OS and image on any other ARM platform and see how well that works for you - there is more to a platform than just the CPU/SoC.

Standardizing doesn't mean locking down, it means making a universal platform that companies can use to make their systems on, and be able to use universally compatible operating systems.
DukenukemX is right, we can still take modern PCs and boot MS-DOS from the 1980s on them because of the said IBM-compatible platform they operate on, which x86 has and ARM does not.
 
I personally see a UEFI as a legacy requirement...
I see UEFI as a requirement.

Also, interestingly enough. It could be a boon for Linux, as if Microsoft was dropping x86 compat and running through an emulator, well you can do that on Linux already.
Without any booting standards on ARM then Linux could fall apart. Anyone remember this being a huge problem for Linux when Microsoft was pushing for UEFI? Not going to be easy to get people to use Linux when they have to pull teeth. Look at LineageOS for Android devices where there isn't enough official LineageOS ports. Most other devices depend on other people to port unofficial LineageOS ports. It's a messy problem we don't want.
Why? The issue isn't that ARM is proprietary, the reality is that ARM is 'open' and open doesn't support the notion of one size fits all - There's a bootloader, and due to the fact that everything is SoC that's really all we need.
As someone who puts custom roms on his phones and tablets, I can tell you that it sucks massive ass. Sucks royal massive ass. Sucks even more ass when it's a no name brand Chinese device where you can't ever get a software update. I have a car stereo in my car that runs Android and it can't connect to the internet because apparently Google likes to phone home to verify internet connection, and my stereo is set to China. I need to root it which I haven't found a method to root it to fix it. If this was x86 I could just format and reinstall the OS but not ARM based devices.
To standardize means to lock down, we want to learn from mistakes of the past, we don't want to keep repeating them.
That's not how things work. Open stardards are open while closed standards are closed. Nearly all ARM based devices have closed boot standards. ARM needs open boot loader standards, otherwise x86 is not threatened by puny ARM.
Ideally speaking, we don't want anyone to have to reverse engineer anything ever again, and there's nothing to state that an underlying UEFI needs to manage hardware in this modern day and age.
Speak for yourself, I still want a PS4 emulator.
You can compile Linux for virtually anything, standardized UEFI or not. Once compiled, a packaged distro can be easily created for the layman to install on their device, and there's a plethora of such distributions available.
Do you know how long it took to port Linux to the PS4? Do you know how long it'll take to port Linux on the Apple M1? Hector Martin is going to make a living porting Linux to the Apple M1, because this doesn't happen over night. Compared to just installing Linux or any OS really onto any x86 PC.
It's a moot point. ARM is not x86, Linux is not Windows. Android is globally the most popular OS, more popular than Windows - the fact there's no standardized EFI like x86 doesn't seem to be holding the platform back in the slightest.
Saying ARM is popular due to Android is like saying Linux is popular because most servers use Linux. Technically both statements are correct but also this doesn't change the fact that both ARM and Linux have insignificant market share on the laptop and desktop market. Most people who own an ARM based device are also the same people who take selfies and post on social media about the food they ate today. These people do not know what an ARM cpu is, let alone care. You could put a CPU in there called the butthole3000 and they still wouldn't know. Compared to someone buying or building a PC who will at least know the CPU's name. They may even know the clock speed it runs at.

Being popular without people knowing you exist is nothing to be proud of.
 
First of all, the RPi as a platform is very, very popular in everything from home brewed projects ranging from RGB to HDMI adapters through to IoT devices and emulation - Computing isn't limited to huge towers with RGB lighting, nor are huge towers with RGB lighting always necessary.
It is a popular platform, but it is hardly universal.
And you're 100% right, Apple are doing their best to lock down their ARM based devices, and creating an unnecessary UEFI simply because some ancient x86 platform used it since the IBM XT days isn't going to rectify that - Nor is it necessary. You need a variant of Android that contains the drivers needed for your SoC, ARM is essentially a standardized instruction set.
Apple uses EFI, not UEFI, starting in 2006 with their Intel-based systems, which has been dropped on their ARM-based systems with a boot ROM similar to their mobile platforms using iOS and iPadOS.
Also, PCs only started using UEFI around 2011, which replaced classic BIOS which was used in the 1980s.
The only reason the PC as a platform took off is because the actual platform was open, it was the BIOS that was proprietary. The fact that one little rom contained proprietary code was nearly the death of an open platform until someone found a loophole around IBM's copyright. If it wasn't for the reverse engineering (more like reorganizing) of IBM's BIOS, there's every chance the IBM XT and everything afterwards would have lived, and died, in corporate environments and never really made it into the hands of consumers as they were stupid expensive at the time and there were a number of cheaper (and arguably better) products that were more suitable to the home user using 68000 CPU's.
It wasn't "open", but due to the low cost and high popularity, it did become the defacto-standard in the 1980s and 1990s because of this.
ARM needs this same thing if it wishes to thrive outside of proprietary equipment, each perhaps running the same OS but with different bootloaders due to said different platforms.

Even the Motorola 68000 had numerous different platforms that were all not universal, and required different bootloaders despite being able to run the same operating systems with the same ISA.
This is the point we are trying to make to you; the CPU ISA isn't the issue, it is the lack of a universal platform like what IBM had for x86.
ARM is already dethroning x86. As stated, Windows is globally no longer the most popular OS, and Apple's transition to ARM is very real and the results are (coming from the mouth of someone that hasn't really liked Apple since the death of the IIGS) very impressive. All that remains to be seen is how well will it scale.

Put the might of Nvidia behind ARM, and x86 may just need to pluck a Rabbit out of it's ass. The concept of backwards compatibility is, in many ways, holding Windows back.
It is, but Apple would never open their ARM-based platform up unless it were a benefit to them in some way.
I hope NVIDIA will make that universal platform standardized, just as IBM did, even if it involves licensing and/or reverse engineering the ROM/firmware like Compaq did with IBM-compatible BIOS back in the 1980s.
 
Last edited:
It is a popular platform, but it is hardly universal.
From a practicality, cost and power consumption perspective, among hobbyists, it's pretty universal - In fact it's revolutionary. The only tech better than ARM in the hobbyist community is FPGA. X86 isn't even really a consideration.
Apple uses EFI, not UEFI, starting in 2006.
Also, PCs only started using UEFI around 2011, which replaced classic BIOS which was used in the 1980s.
I never stated otherwise. However, as far as I'm aware, their ARM based devices have no EFI beyond a locked down bootloader whatsoever.
It wasn't "open", but due to the low cost and high popularity, it did become the defacto-standard in the 1980s and 1990s because of this.
ARM needs this same thing if it wishes to thrive outside of proprietary equipment, each perhaps running the same OS but with different bootloaders due to said different platforms.
It definitely was open. S-100, ISA, PCI, they're no more than connectors on a data and address bus using readily available at the time electronics. The only component that was propitiatory was the BIOS, hence why IBM copyrighted it.

txInNq9.png

Even the Motorola 68000 had numerous different platforms that were all not universal, and required different bootloaders despite being able to run the same operating systems with the same ISA.
This is the point we are trying to make to you; the CPU ISA isn't the issue, it is the lack of a universal platform like what IBM had for x86.
They were not universal because they ran different operating systems, a UEFI will do absolutely nothing to stop that from happening again. A UEFI is no more than a bootloader and a collection of libraries - All we need is the bootloader (unless you want to return to the S-100/VAX PDP days), the OS can contain the basic libraries needed for the system to run. We don't need a UEFI just because x86 uses one dating back to the days of the original IBM XT.

I hope NVIDIA will make that universal platform standardized, just as IBM did, even if it involves licensing and/or reverse engineering the ROM/firmware like Compaq did with IBM-compatible BIOS back in the 1980s.
It's already standardized, UEFI does not = Standardization just because it's what you're used to. Furthermore, implementing a UEFI will not change Apple's locked down ARM stance, not when Apple have owned 43% of the company since 1983.
 
From a practicality, cost and power consumption perspective, among hobbyists, it's pretty universal - In fact it's revolutionary. The only tech better than ARM in the hobbyist community is FPGA. X86 isn't even really a consideration.
While that statement is correct, you are totally missing the point we are trying to get across to you.
Again, please stop bringing up the ARM ISA and its popularity as the issue - that is not the issue we are discussing.
I never stated otherwise. However, as far as I'm aware, their ARM based devices have no EFI beyond a locked down bootloader whatsoever.
That is correct, they use a boot ROM with their ARM platforms, and I did recently edit the post to reflect this.
It definitely was open. S-100, ISA, PCI, they're no more than connectors on a data and address bus using readily available at the time electronics. The only component that was propitiatory was the BIOS, hence why IBM copyrighted it.

View attachment 345037
Ah, gotcha, that makes sense, and it is exactly what ARM platforms need.
They were not universal because they ran different operating systems, a UEFI will do absolutely nothing to stop that from happening again. A UEFI is no more than a bootloader and a collection of libraries - All we need is the bootloader (unless you want to return to the S-100/VAX PDP days), the OS can contain the basic libraries needed for the system to run. We don't need a UEFI just because x86 uses one dating back to the days of the original IBM XT.
A platform is more than just the bootloader, as hardware paths and I/O can be radically different from one platform to the next, regardless of the ISA.
Compare the PS4 to an IBM-compatible PC - both run with the x86-64 ISA, but both require different bootloaders due to the PS4 not being a PC, and having a wildly different platform.

This is exactly the issue with nearly every ARM-based device in existence, save for the Raspberry Pi's platform.
It's already standardized, UEFI does not = Standardization just because it's what you're used to. Furthermore, implementing a UEFI will not change Apple's locked down ARM stance, not when Apple have owned 43% of the company since 1983.
haha, I'm used to quite a few different firmware and boot ROMs outside of UEFI.
OpenBoot/OpenFirmware, PROM, SROM, PS3 Firmware, you name it, I've probably done it.

As for Apple not using UEFI, or whatever, for their ARM-based platforms is fine.
No one is relying on Apple to do this, and again, they would never do this unless it benefited them in some way; please don't put words into our mouths.

We are hardly relying on Apple to make the universal platform and standard, but we are hoping that NVIDIA will do this, even if it requires licensing and/or legal reverse engineering.
 
Ah, gotcha, that makes sense, and it is exactly what ARM platforms need.
No they don't. To do so would defeat the very mantra of the ARM platform. ARM allow chip designers to implement the architecture as they desire, right down to their own custom instructions provided they pay royalties to ARM. Therefore, what you're implying may not even be possible. In fact, the ability for manufacturers to fab their own designs based on ARM is actually quite desirable from a security perspective as it can limit the attack vector.

Basically, integration (what you're getting in a twist over) is up to the fab designer, even the individual manufacturer, and not ARM themselves. If this makes it difficult for closed operating systems like Windows to adopt such platforms, than perhaps Windows as an operating system needs to evolve? Because while compilation takes time, Linux is showing how open source operating systems are definitely a benefit when it comes to ARM designs - The OS simply offers greater flexibility.

As stated, Android has overtaken Windows as the more popular OS globally, running almost solely on ARM based designs. Whether you want to admit it or not, x86/64 is for the first real time ever under threat.

Personally, I think the days of NT/NTFS as it is now are numbered.
 
Last edited:
No they don't. To do so would defeat the very mantra of the ARM platform. ARM allow chip designers to implement the architecture as they desire, right down to their own custom instructions provided they pay royalties to ARM. Therefore, what you're implying may not even be possible. In fact, the ability for manufacturers to fab their own designs based on ARM is actually quite desirable from a security perspective as it can limit the attack vector.
Keep in mind that this isn't an ARM issue but an issue that's associated with ARM. PS4 and PS5 for example are both x86 but do not use a UEFI boot loader. Manufacturers are taking advantage of a new market by making it limited in what you can do, and ARM is just coming along for the ride.
Basically, integration (what you're getting in a twist over) is up to the fab designer, even the individual manufacturer, and not ARM themselves. If this makes it difficult for closed operating systems like Windows to adopt such platforms, than perhaps Windows as an operating system needs to evolve? Because while compilation takes time, Linux is showing how open source operating systems are definitely a benefit when it comes to ARM designs - The OS simply offers greater flexibility.
I'm not sure how you're coming to these conclusions because Android is based off Linux and Android is a horrible fucking mess. What if I sold you a PC that was running Ubuntu 20.04, then a year later 22.04 was released and you couldn't upgrade because the bootloader has locked it down? You get no security updates because it doesn't make me money, and my answer to you is to buy a newer better PC with Ubuntu 22.04. Meanwhile my 500GB disk drive has been divided in half and I can't see the other half because I don't have root permission. So now you look to XDA-Developers to see if anyone ported LineageOS to your PC, which luckily for you someone did but it's buggy and is missing features because it was done by a guy living in a basement in Russia.

That is Android in a nutshell. The reason Android devices are stuck with much older versions of the OS is because there's no standardization. Which makes Android inherently less secure than a typical x86 Linux distro. Though this is why people are now looking towards Linux Cell Phones like Librem 5, which acts more like a Linux Distro than Android. Due to the lack of money behind it, Linux OS's on cell phones aren't the greatest thing.


As stated, Android has overtaken Windows as the more popular OS globally, running almost solely on ARM based designs. Whether you want to admit it or not, x86/64 is for the first real time ever under threat.
If x86 is under threat then that's because of Intel, not due to ARM and Android. Intel has screwed up a lot for the past few years and it may take a few years more before they fix their mistakes. Microsoft would love for you to jump onto ARM, but they've failed many times trying to make Windows on ARM a thing. Microsoft like many companies would like you to buy a PC that's locked down and you could only do what they'd like you to do, and ARM is very locked down.
 
Last edited:
Whatever the cause is, the main thing to understand here is that PC gaming is booming. Specifically X86 Windows gaming. The only move towards ARM right now is Chrome, and that's because Chromebooks are cheap.

PC gaming is doing well, but the market share stats are off and don't really convey the significance. IDC had Apple's market share at about 7.5 percent as of late 2020, versus Chrome OS' 10.8 percent. However, Apple is transitioning completely to ARM, while only some Chromebooks use ARM. So that's a sizeable chunk of the broader PC market that's moving to ARM, and Apple currently represents the majority of that push.


If you count cell phones and tablets then sure, but that's like saying AAA gaming is dying because most people are playing mobile games which are mostly bejeweled clones. You don't see much of any laptop or Desktop running ARM, for a reason. While most people are on laptops, that's because of the standards like ATX that brings it all together. If you want a more powerful ARM laptop then what do you buy? You need standards to unify the platform.

Mobile helps, but I mean in conventional PCs as well. About 18.3 percent of computers are now running something that isn't Windows, and in a year a large chunk of those won't be using x86, either.

You do need standards to a degree; it's just important to note that ATX isn't one of them, and that laptops aren't as strictly tied to standards as desktops.


Microsoft can't do anything to ease nothing. You aren't going to get people to jump on ARM unless you offer standards. You need a standard way to install an OS, which ARM devices are sorely missing. You can't buy an ARM laptop and simply install Linux. It usually requires a lot more work to do something like install Linux on a Chromebook. You need desktop ARM motherboards with sockets and memory slots.

It can absolutely do something. Remember, it only recently added emulation for 64-bit x86 apps on ARM machines. Think about it: until very recently (and then only in a rough form), you couldn't run 64-bit apps at all on ARM-based Windows PCs unless they were native apps. You'd need ARM versions of absolutely every app you intend to run if you don't want to be stuck with 32-bit versions. On top of this, I'd say Microsoft also needs to make sure ARM support is truly broad (that is not skewed toward Qualcomm). The standards you mentioned are important, but not the only necessary ingredients.
 
It can absolutely do something. Remember, it only recently added emulation for 64-bit x86 apps on ARM machines. Think about it: until very recently (and then only in a rough form), you couldn't run 64-bit apps at all on ARM-based Windows PCs unless they were native apps. You'd need ARM versions of absolutely every app you intend to run if you don't want to be stuck with 32-bit versions. On top of this, I'd say Microsoft also needs to make sure ARM support is truly broad (that is not skewed toward Qualcomm). The standards you mentioned are important, but not the only necessary ingredients.
Well there are a lot of rumors swirling that Microsoft is building their own ARM chips. Qcomm stuff seems like it was always destined to fail. Qcoms chips where never more then phone grade parts. I don't know if Microsoft is in anyway capable of building something that would compete with Apple. I tend to think not... but I do believe it shows they would be more willing to tailor windows a bit to work properly with others. The main massive issue for windows as I see it is the windows micro kernel.

If Microsoft is going to dump their own kernel for the Linux kernel... which they should. This would be the time to do it. One of the big advantages for Microsoft (and consumers) would be how easy it would be at that point for any ARM supplier to jump in and supply a windows powered device. Microsoft would simply need to aggressively keep their kernel up to date. (with in reason) All they have to do is keep their Windows Linux powered OS run the latest LTS at all times. At that point chips from basically the entire industry could be supported easily. If manufactures want to use bios eufi custom crap or whatever would be irrelevant as long as it was supported by the latest kernel. The MS microkernel system is terrible for CPU support. Switch to a Linux backend and not only is everything possible... Microsoft doesn't even have to be involved.
 
Well there are a lot of rumors swirling that Microsoft is building their own ARM chips. Qcomm stuff seems like it was always destined to fail. Qcoms chips where never more then phone grade parts. I don't know if Microsoft is in anyway capable of building something that would compete with Apple. I tend to think not... but I do believe it shows they would be more willing to tailor windows a bit to work properly with others. The main massive issue for windows as I see it is the windows micro kernel.

If Microsoft is going to dump their own kernel for the Linux kernel... which they should. This would be the time to do it. One of the big advantages for Microsoft (and consumers) would be how easy it would be at that point for any ARM supplier to jump in and supply a windows powered device. Microsoft would simply need to aggressively keep their kernel up to date. (with in reason) All they have to do is keep their Windows Linux powered OS run the latest LTS at all times. At that point chips from basically the entire industry could be supported easily. If manufactures want to use bios eufi custom crap or whatever would be irrelevant as long as it was supported by the latest kernel. The MS microkernel system is terrible for CPU support. Switch to a Linux backend and not only is everything possible... Microsoft doesn't even have to be involved.
I've heard of those rumors. I hope Microsoft does it, if just to spur Qualcomm into action. Qualcomm is notoriously unable to devote its attention to any consumer platform that isn't for phones. Want something for PCs and smartwatches? Be glad if we release one chip every two or three years! A microkernel switch might help, at least so long as Microsoft can ensure backward compatibility.

Apple does so well in certain categories in part because it's one of the few companies that bothers to show up consistently. It rules smartwatches and tablets because it has at least one new model each year, every year, and often with significant (if sometimes overdue) software improvements. Qualcomm, Google and Microsoft often come across as entitled kids who skip classes in a course, but then wonder why they failed the final exam — they want all the glory without putting in the work. If QC and Microsoft aren't careful, they could see Apple walk away with the lead in ARM-based PCs simply by delivering the constant iteration it's known for.
 
Last edited:
  • Like
Reactions: ChadD
like this
The fact is:

- ARM devices cannot be 'standardized' as that's not how ARM works.
- Manufacturers like this as they can create their own silicon and therefore their own ecosystems.
- If Windows struggles with this concept, Windows needs to evolve - As the concept is not going away.
- It's actually an advantage from an attack vector perspective.
- Raspberry Pi OS works across all Raspberry Pi's, so there's a desktop platform with as much standardization as you're going to get. I see no reason why there couldn't be an installer, but I also don't see an issue with the current method of installing an OS - At least I don't have to download the OS, extract the ISO and convert it to bootable media, install the OS and then install the drivers. I just download the tarball, open the imaging software (or use DD), create my SD card, pop the SD card into the machine. Furthermore, I'm far from limited, there's a number of operating systems to choose from - I believe there's even a Windows variant.
 
Last edited:
If x86 is under threat then that's because of Intel, not due to ARM and Android. Intel has screwed up a lot for the past few years and it may take a few years more before they fix their mistakes. Microsoft would love for you to jump onto ARM, but they've failed many times trying to make Windows on ARM a thing. Microsoft like many companies would like you to buy a PC that's locked down and you could only do what they'd like you to do, and ARM is very locked down.
The rise of ARM's popularity started well before Intel ran into trouble with their chip designs and ARM is anything but locked down. Hence the reason why you cannot have one UEFI to rule them all.
 
The fact is:

- ARM devices cannot be 'standardized' as that's not how ARM works.
Anything can be standardized, or kept proprietary - that's how everything works - and we are saying that ARM-based platforms need to be standardized the same way that the x86-based IBM-compatible platform was, in order for ARM to truly take over the x86 industry all together.
ARM is a CPU ISA, no different than x86-64 or m68k or Power ISA or SPARC, and can be standardized just as the IBM-compatible PC was and still is to this day 40 years later.
- Manufacturers like this as they can create their own silicon and therefore their own ecosystems.
True, because Corporatism works so well.
I would think you, being a large Raspberry Pi and open architecture fan, would be heavily against this.
- If Windows struggles with this concept, Windows needs to evolve - As the concept is not going away.
Don't hold your breath for Microsoft, they are always late to the game, kill off products and product support far too soon, and are going down hill in terms of quality.
They are shifting to more cloud services and functions, and away from the OS market in terms of quality.
- It's actually an advantage from an attack vector perspective.
I think DukenukemX hit the nail on the head about this - might want to re-read his post above.
- Raspberry Pi OS works across all Raspberry Pi's, so there's a desktop platform with as much standardization as you're going to get. I see no reason why there couldn't be an installer, but I also don't see an issue with the current method of installing an OS - At least I don't have to download the OS, extract the ISO and convert it to bootable media, install the OS and then install the drivers. I just download the tarball, open the imaging software (or use DD), create my SD card, pop the SD card into the machine. Furthermore, I'm far from limited, there's a number of operating systems to choose from - I believe there's even a Windows variant.
It's dd, not DD, and for the third time, try and boot your Raspberry Pi OS disk image on any other ARM platform.
It's standardized for the Raspberry Pi platform, period, and will not work on any other ARM platform - this is the point we keep trying to make to you that you either don't understand or are just simply ignoring at this point.

The rise of ARM's popularity started well before Intel ran into trouble with their chip designs and ARM is anything but locked down. Hence the reason why you cannot have one UEFI to rule them all.
FFS, we aren't talking about the ARM ISA itself, we are talking about ARM platforms being standardized.
Do you even know what a platform is???
 
The rise of ARM's popularity started well before Intel ran into trouble with their chip designs and ARM is anything but locked down. Hence the reason why you cannot have one UEFI to rule them all.
You need to watch this to get what we are talking about:



PS4 x86-64 Platform:
PS4 Platform.png


IBM-Compatible PC x86-64 Platform:
PC x86-64 Platform.png



Look at how different both of those platforms are to one another - they are both x86-64, yet are totally different, and thus have different platforms, firmware, and bootloaders.
This is what we are talking about in terms of ARM platforms needing to stop being different/proprietary from one another, and be more standardized like the IBM-compatible PC platform; it doesn't mean we want it to exclusively use UEFI itself, just for the platform/firmware/bootloader to be standardized.

This is the point we are trying to make to you that you keep missing.
The CPU and CPU ISA are just one part of this equation, and it happens to be that we want this for the ARM ISA.
 
You need to watch this to get what we are talking about:



PS4 x86-64 Platform:
View attachment 345162

IBM-Compatible PC x86-64 Platform:
View attachment 345163


Look at how different both of those platforms are to one another - they are both x86-64, yet are totally different, and thus have different platforms, firmware, and bootloaders.
This is what we are talking about in terms of ARM platforms needing to stop being different/proprietary from one another, and be more standardized like the IBM-compatible PC platform; it doesn't mean we want it to exclusively use UEFI itself, just for the platform/firmware/bootloader to be standardized.

This is the point we are trying to make to you that you keep missing.
The CPU and CPU ISA are just one part of this equation, and it happens to be that we want this for the ARM ISA.

I don't need to watch that, I know exactly what you're talking about and it can't be done on a platform as diverse as ARM, not across 'all' devices. Having said that, there's nothing stopping development of one desktop platform, as stated, we already have one.

As stated, what you perceive as an issue isn't going to stop manufacturers using ARM processors as they actually like the idea of creating their own ecosystems.

The only platform that's going to struggle is Windows, and Windows needs to evolve. We don't have to cave to the whim of Microsoft.
 
Last edited:
  • Like
Reactions: ChadD
like this
I don't need to watch that, I know exactly what you're talking abut and it can't be done on a platform as diverse as ARM, not across 'all' devices. Having said that, there's nothing stopping development of one desktop platform, as stated, we already have one.
We aren't talking about the zillions of mobile devices with ARM-based platforms.
As for the desktop platform, the Raspberry Pi is very solid, and if everything else going forward could be based on it, that would be fantastic.
As stated, what you perceive as an issue isn't going to stop manufacturer susing ARM processors as they actually like the idea of creating their own ecosystems.
Nope, and this is why it will take ARM that much longer to over take x86-64 systems, and is going to make computing environments feel just like they were in the 1960s through the early 1980s with each company having its own proprietary platform, proprietary OS, and proprietary software.
It might be better than that because of licensing, but only time will tell.

This is what we are hoping to get away from, which is direly an issue for end-users like us.
The only platform that's going to struggle is Windows, and Windows needs to evolve. We don't have to cave to the whim of Microsoft.
Windows is an OS, not a platform, but I get what you are saying, and Microsoft is definitely going to struggle.
haha, no one here has said anyone needs to cave to the whim of Microsoft, and doing so would be bad for virtually everyone except Microsoft.
 
Well the fact is: That's not how ARM works. It may have worked for x86, x86 isn't licensed like ARM is.
Perhaps, though if and when NVIDIA owns ARM, we will see how that changes.
He actually didn't.


I just watched that yesterday, and while as a whole having a single system platform makes it easier to hammer away at, it doesn't stop a ton of proprietary platforms running different versions of the same OS from being heavily exploited.

DukenukemX is correct about what he stated, and I have witnessed that exact thing happen in enterprise numerous times.
At least with Apple, applying the updates for severe CVE risks is relatively painless, and I definitely can't say that about a ton of various Android devices, OEMs, and varying degrees of support.
As for your comments regarding 'DD', they're obviously nothing more than semantics.
https://en.wikipedia.org/wiki/Dd_(Unix)
No, the program is literally called dd, not "DD" - that isn't semantics, that's an incorrect statement on your part.
 
We aren't talking about the zillions of mobile devices with ARM-based platforms.
Right...So the goal posts just shifted? To quote an earlier post:
We're not just talking about the Raspberry Pi. Most ARM SoC's are in phones and tablets
And this post:
As someone who puts custom roms on his phones and tablets, I can tell you that it sucks massive ass. Sucks royal massive ass. Sucks even more ass when it's a no name brand Chinese device where you can't ever get a software update. I have a car stereo in my car that runs Android and it can't connect to the internet because apparently Google likes to phone home to verify internet connection, and my stereo is set to China. I need to root it which I haven't found a method to root it to fix it. If this was x86 I could just format and reinstall the OS but not ARM based devices.
As for your semantics, all they're highlighting is a back against the wall inability to form a decent rebuttal. No one cares abut your pointless corrections, you master of the written word. :rolleyes:
 
Perhaps, though if and when NVIDIA owns ARM, we will see how that changes.
Read my earlier post on how ARM licencing works, because you really seem to be struggling with this concept while nit picking on the use of capitalization and terminology.

Naturally when i use DD I don't capitalize it in terminal due to the fact that unlike Windows, Linux is case sensitive.

EDIT: For your convenience:

https://hardforum.com/threads/new-arm-v9-announced.2009193/post-1044977921
 
Last edited:
Right...So the goal posts just shifted? To quote an earlier post:

And this post:

As for your semantics, all they're highlighting is a back against the wall inability to form a decent rebuttal. No one cares abut your pointless corrections, you master of the written word. :rolleyes:
We are still talking about ARM for desktop, laptops, workstations, servers, etc., which is what we want standardization for.
If that happens for mobile devices, great, but I wouldn't hold my breath on that one.

DukenukemX was just pointing out that the majority of ARM devices exist in mobile devices, and how segmented said platforms are, which is correct.
That isn't us changing the goal posts.
Read my earlier post on how ARM licencing works, because you really seem to be struggling with this concept while nit picking on the use of capitalization and terminology.

Naturally when i use DD I don't capitalize it in terminal due to the fact that unlike Windows, Linux is case sensitive.
I get how ARM licensing works, and that anyone can do anything with it.
The point is is that it is just going to needlessly create a bunch of proprietary platform silos that are only going to benefit the corporations and not the end-users, and it will also take ARM that much longer to overcome x86 by doing so, thus in all reality, will hurt the corporations and end-users (and security) in the long run and will make things more needlessly difficult for a quick profit in the short-run.

Again, the program is literally called dd, not "DD"; DD isn't a thing.
Actually, you can spell it however you want, I really don't care. :)
 
We are still talking about ARM for desktop, laptops, workstations, servers, etc., which is what we want standardization for.
Well that's not highlighted in the posts quoted above. Basically, you're all over the place here.

I get how ARM licensing works, and that anyone can do anything with it.
The point is is that it is just going to needlessly create a bunch of proprietary platform silos that are only going to benefit the corporations and not the end-users, and it will also take ARM that much longer to overcome x86 by doing so, thus in all reality, will hurt the corporations and end-users (and security) in the long run and will make things more needlessly difficult for a quick profit in the short-run.
And manufacturers like this, increasing the popularity of ARM differently to what you may be used to under x86. That's not to state that ARM cannot remain open on certain platforms like the Raspberry Pi, which is basically the defacto standard for a desktop/hobbiest platform. Although I will state there is a issue regrading the proprietary nature of a certain component on the SoC chosen for the Raspberry Pi. However, the Pi is still essentially an open platform with a boot process that could be standardized if integration is really the issue you believe it to be - Which I don't believe it is.

Again, the program is literally called dd, not "DD"; DD isn't a thing.
Actually, you can spell it however you want, I really don't care.
Well, if we want to get into semantics...I did spell the command right.
 
I've heard of those rumors. I hope Microsoft does it, if just to spur Qualcomm into action. Qualcomm is notoriously unable to devote its attention to any consumer platform that isn't for phones. Want something for PCs and smartwatches? Be glad if we release one chip every two or three years! A microkernel switch might help, at least so long as Microsoft can ensure backward compatibility.

Apple does so well in certain categories in part because it's one of the few companies that bothers to show up consistently. It rules smartwatches and tablets because it has at least one new model each year, every year, and often with significant (if sometimes overdue) software improvements. Qualcomm, Google and Microsoft often come across as entitled kids who skip classes in a course, but then wonder why they failed the final exam — they want all the glory without putting in the work. If QC and Microsoft aren't careful, they could see Apple walk away with the lead in ARM-based PCs simply by delivering the constant iteration it's known for.
I would say so far ya Apple is in the lead. Unless Qcomm MS Nvidia Google or someone pops up with a true cutting edge Arm desktop design. Apple is going to walk away with consumer grade performance. I know that still sounds crazy... I mean M1 is cool but its not that. However M1 is their first build... aimed at Macbook AIR and Mini hardware. The real laptop class chip is still coming... never mind what they have in store for Mac Pros. At this point it seems like the non apple world is a couple years behind already.
 
The fact is:

- ARM devices cannot be 'standardized' as that's not how ARM works.
- Manufacturers like this as they can create their own silicon and therefore their own ecosystems.
- If Windows struggles with this concept, Windows needs to evolve - As the concept is not going away.
- It's actually an advantage from an attack vector perspective.
- Raspberry Pi OS works across all Raspberry Pi's, so there's a desktop platform with as much standardization as you're going to get. I see no reason why there couldn't be an installer, but I also don't see an issue with the current method of installing an OS - At least I don't have to download the OS, extract the ISO and convert it to bootable media, install the OS and then install the drivers. I just download the tarball, open the imaging software (or use DD), create my SD card, pop the SD card into the machine. Furthermore, I'm far from limited, there's a number of operating systems to choose from - I believe there's even a Windows variant.
Of course it can be standardized.

If Nvidia completes their ARM purchase .... and they really want to kill x86. Yes they will need to standardize a ARM socket, and platform. Does that mean they still can't license ARM to Apple or Rasberry or Samsung or some Chinese company to use in some super locked down priority thing. Of course not. However if they want to go after the Desktop / Gaming market at some point yes they need a official ARM socket.... and ideally a platform on which to sell MOBOs. If Dell wants to build their own ARM systems and get chips from whoever so be it... but they will also need the option to buy a socketed version. So they can do the standard... buy this model with X or Y, a bit more ram a bit less, a 8 core arm or a 12 core arm. ECT ECT.

If they want to replace the x86 PC industry... they are going to have to have a option that looks a lot more like the current system. Yes they need a standardized platform AS well. It doesn't have to be exclusive. They are going to continue selling ARM to companies like Fujitsu who will make continue to make massive 48+ core chips that will need 8000 pins. However they will have to standardize something more consumer focused where as consumers we could hopefully buy a Nvidia chip or a Qcomm chip or a 16 core Fujitsu home version. There is a place for the super specialized locked down stuff... but if they want to go after the market for real they will need a standardized platform.

If Nvidia isn't completely stupid that is exactly what they will do. If I was them I would have the standards already completed on paper and ready to throw in regulators faces when they say; "Qcomm says you are going to be bad stewards, and are in general evil." Nvidias response should be no no we are going to open ARM like never before make CPUs inter changeable and open the ARM consumer platform to the mother board manufacturer industry ect.

What Nvidia and ARM need to create is a situation where in 2 years you could buy a "Asus ROG Nvidia Socket 1" + "64GB of Corsair DDR5" + "AMD RX7700" + "Samsung 980 Pro" + "Nvidia Tegra V9 3.6ghz 12 core Socket 1 CPU" {or any other ARM license holder that wants to build their CPUs in socket 1 format]
Then you have something that could really put the boots to x86. In combo with that if Nvidia is pushing their own Linux distro as Intel does with clear. And Microsoft puts out a proper version of Windows ARM (imo if it ran on Linux that would be the best solution for MS and the market) where the newest kernel had drivers for basically anyones ARM chips. At that point Intel who.... AMD would unmothball K1 pretty fast and I think ARM could basically completely replace x86 faster then even I would dream of.

This notion of backwards compatibility at this point is a pretty laughable road block... anything written in the last 5 years is either already ported or easily done. Apple has had no issue getting Adobe to port things like photoshop. The outliers mostly run in things like wine under linux, I really really really doubt MS would have any issues building a rosette like bit for Windows. MS already has their own internal wine like software they run on their cloud stuff which is all running Linux.... they are also building their own ARM server chips (that isn't a rumor its admitted) no doubt they are building those tools right now if there not already done.
 
Last edited:
Back
Top