Jump to content
Search In
  • More options...
Find results that contain...
Find results in...
LexiMax

Intel proposes transition to 64-bit only architecture

Recommended Posts

37 minutes ago, Halfblind said:

Microsoft released an update that year and afterwards the PC ran like trash. After that update Microsoft expects you to move away from using mechanical drives. After replacing the old drive with a newer SSD, the laptop ran as good as it did before the update. True that you can still run windows 10/11 on an old HDD but it will run like trash.

There's been no such update. One of our programmers at work still runs a mechanical drive on at least Win10 (I'm not sure if he's moved to 11 yet, that's not a requirement) and must also have an up to date system given the type of work we do. You are either experiencing confirmation bias (it was always slow but you never initially noticed) or misdiagnosed the problem (it being caused by something else that happened or just the drive aging).

There's a lot of things you tend to only notice in hindsight, like the download speeds on 7th gen consoles barely pushing 10mbps. They were always like this but when they released they were still the fastest system. Likewise we now have SSDs, of course a mechanical drive is going to look slow by comparison.

 

Now if you compare it to Windows 8, then yes OS performance has likely changed for mechanical drives because they now optimise to SSDs given that became the market standard, however this was during the transition from 8->10, it hasn't changed on 10 itself given that would be unwise to change the requirements on OEMs.

Edited by Edward850

Share this post


Link to post

@Edward850 @Murdoch

Rumor has it that Microsoft does indeed want to kill off support for HDDs as boot drives in favor of SSD's though.

https://www.techradar.com/news/is-microsoft-about-to-kill-off-the-hdd-once-and-for-all
https://www.techspot.com/news/94882-microsoft-calls-oems-kill-off-hdd-boot-drives.html
 

Also, another Windows 10 update in 2021 BSOD'd both the non-profit PC and my home PC with the same update. I was able to get both back up, thank God. Most people ended up getting the patched update that didn't get BSOD'd hot-fixed just in time. The non-profit PC updates automatically without me able to do anything about it, but I stupidly updated my home PC as soon as I got that update, so my bad. Because of that I wait at least a day before I hit that update button.

Share this post


Link to post
5 minutes ago, Halfblind said:

@Edward850 @Murdoch

Rumor has it that Microsoft does indeed want to kill off support for HDDs as boot drives in favor of SSD's though.

https://www.techradar.com/news/is-microsoft-about-to-kill-off-the-hdd-once-and-for-all
https://www.techspot.com/news/94882-microsoft-calls-oems-kill-off-hdd-boot-drives.html
 

Also, another Windows 10 update in 2021 BSOD'd both the non-profit PC and my home PC with the same update. I was able to get both back up, thank God. Most people ended up getting the patched update that didn't get BSOD'd hot-fixed just in time. The non-profit PC updates automatically without me able to do anything about it, but I stupidly updated my home PC as soon as I got that update, so my bad. Because of that I wait at least a day before I hit that update button.

 

Given how cheap SSDs are nowadays and how unless you build it yourself you are very unlikely to find a new computer with an HDD as the boot disk (here in New Zealand at least), I see no problem here. Most people don't need a lot of space and a 250GB drive is stupid cheap, so the fightback from the OEMs frankly makes no sense. I deal with Joe Average User customers on a daily basis and relatively few of them have over 100GB of stuff. Also, it's only for Windows 11 and it appears to only be a guideline for new systems, not they are suddenly going to stop supporting HDDs in as boot devices themselves so existing systems won't just stop working. As I have said many times, tech moves on. Always has, always will. It's good to support legacy stuff where possible but sometimes the cord needs to be severed.

Share this post


Link to post

So if this breaks 32-bit drivers even in 64-bit windows... I will admit that this does concern me a little. I honestly don't know what would be affected by this. Printers, virtual DVD drives, sound stuff? 

Not being able to run 32-bit operating systems with the latest hardware is less of a concern lol.

Share this post


Link to post
47 minutes ago, Danfun64 said:

So if this breaks 32-bit drivers even in 64-bit windows...

You have never been able to run 32bit drivers in 64bit Windows, which has been the case since XP 64bit edition. This has actually been part of the cause of peoples perceived problems with Vista, the driver manufacture support just wasn't there at the time for 64bit.

Share this post


Link to post

Now that Intel has now discontinued the IA-32 architecture completely, I wish to see Windows 3.1 game developers to re-release their Windows 3.1-era games on Steam or GOG via OTVDM rather than using DosBox bundled with a pirated copy of Windows 3.1.

Share this post


Link to post
1 hour ago, Wadmodder Shalton said:

Now that Intel has now discontinued the IA-32 architecture completely, I wish to see Windows 3.1 game developers to re-release their Windows 3.1-era games on Steam or GOG via OTVDM rather than using DosBox bundled with a pirated copy of Windows 3.1.

 

Given that a not insignificant number of those developers don't exist anymore, all I can say is good luck with that.
 

https://en.wikipedia.org/wiki/List_of_Windows_3.x_games

 

Many were so minor, that they do not even rate a "page does not exist" link, many have that, and most of the ones that do have a page refers to them in the past tense.

Such games only appeal to determined nostalgia buffs, and they already have ways and means around it. This decision by Intel will change nothing.

Share this post


Link to post

The only game I recognized in this list is Ecco The Dolphin. Besides Battle Chess, which I think I never played.

But I thought that Ecco The Dolphin was a Windows 95 game. At least it was released in this era. I don't see other Win95 titles from SEGA there, like Comix Zone or Earthworm Jim.

Share this post


Link to post

If you ask me, the biggest mystery here is why it took this long for such a proposal.

 

Even if we consider the need to support such old OSs it would suffice to have one range of low end CPUs to support it, not everything. What they could have done 10 years ago was to drop all this legacy support from their mainline CPUs but keep one unit in stock that retains it, but in turn costs a bit more.

Share this post


Link to post
8 hours ago, Wadmodder Shalton said:

Now that Intel has now discontinued the IA-32 architecture completely, I wish to see Windows 3.1 game developers to re-release their Windows 3.1-era games on Steam or GOG via OTVDM rather than using DosBox bundled with a pirated copy of Windows 3.1.

Processor architecture is not the thing keeping those games in the past.

Share this post


Link to post

I think I've actually got a 32-bit HP somewhere in my room, I never use it anymore because it basically runs at the speed of a german MAUS heavy tank at the slowest speed lmfao

 

My decrepit-as-fuck 4gb ram asus however... Good lord I fucking love it, has windows 10 too and gzdoom runs soooooo smoothly at the default settings!

Share this post


Link to post
13 hours ago, Halfblind said:

Rumor has it that Microsoft does indeed want to kill off support for HDDs as boot drives in favor of SSD's though.

https://www.techradar.com/news/is-microsoft-about-to-kill-off-the-hdd-once-and-for-all
https://www.techspot.com/news/94882-microsoft-calls-oems-kill-off-hdd-boot-drives.html

Another good reason for me to ditch that decade-old-yet-reliable Western Digital HDD in favor of a SSD. Not in any hurry though..

Share this post


Link to post
12 hours ago, Professor Hastig said:

If you ask me, the biggest mystery here is why it took this long for such a proposal.

People have asked Intel and AMD engineers this many times and the answer basically boils down to: Because it doesn't actually benefit things as much as people think it would.  Modern CPUs don't really execute their ISA directly, so the complexity mostly gets hidden by the decoding stage.  Now having little benefit isn't the same as having zero benefit and I guess now that we're talking about CPUs with hundreds of cores that little benefit may finally be adding up to something worthwhile.

 

As an aside: While the x86S proposal doesn't mention it (it's already a CPUID flag so nothing to propose), I do know talk of dropping MMX has been floating around (due to the weird behavior it places on registers).  I wouldn't be surprised if the future x86S CPUs also drop MMX at the same time.

 

12 hours ago, Professor Hastig said:

Even if we consider the need to support such old OSs it would suffice to have one range of low end CPUs to support it, not everything.

Given that the reason to do this would be to reduce the amount of silicon needed for a core (even if negligible for low core counts) it would make more sense to make legacy support a high end CPU feature.  Plus the people that really need it will probably happily pay extra for it anyway.

 

An interesting way to make a transitional CPU would be to pair x86S E-cores with full featured P-cores.  Then legacy operating systems could run with the E-cores disabled.  I highly doubt any such products would be planned though since it'd be needlessly complicated for the 3 people that care, and it would negate the simplified firmware benefits of the proposal.

Share this post


Link to post
19 hours ago, Blzut3 said:

As an aside: While the x86S proposal doesn't mention it (it's already a CPUID flag so nothing to propose), I do know talk of dropping MMX has been floating around (due to the weird behavior it places on registers).  I wouldn't be surprised if the future x86S CPUs also drop MMX at the same time.

 

What would this break?  I think I remember some 90's games depended on MMX (the one I remember specifically was Sega's PC port of Virtual On), but is there some other widely-deployed thing I'm not thinking of?

Share this post


Link to post
19 hours ago, Blzut3 said:

As an aside: While the x86S proposal doesn't mention it (it's already a CPUID flag so nothing to propose), I do know talk of dropping MMX has been floating around (due to the weird behavior it places on registers).  I wouldn't be surprised if the future x86S CPUs also drop MMX at the same time. 

 

 

Now THAT might pose a real problem because it would actually break existing and still working legacy software. Who knows how many older games are using MMX code somewhere?

 

Share this post


Link to post

Using MMX code and requiring MMX are two different things of course.  I'm not sure how many games are in the latter camp.  I don't know if compilers ever had the ability to auto-vectorize to MMX, but Intel has been working on getting GCC to compile MMX intrinsics to SSE (not sure off the top of my head if this got merged yet).  Clang has discussed doing similar.

 

Certainly will break something, I just don't know if there's enough software dependent on MMX to take the possibility off the table.

 

Edit: It was merged into GCC 10. https://www.phoronix.com/news/GCC-10-Emulating-MMX-With-SSE Looks like Clang's PR is in limbo? https://reviews.llvm.org/D86855

Edited by Blzut3

Share this post


Link to post

They're trying to kill off MMX? That's sus. There's an imposter among them!

 

Share this post


Link to post
8 hours ago, Blzut3 said:

Using MMX code and requiring MMX are two different things of course.

 

While that is true, there's something else to consider here: How good is software at detecting support properly?

Are they performing an actual feature check or are they just going "oh, this is newer than a {whatever the first MMX CPU was}, so I can safely use the feature" As a simple example, take a game doing both an MMX and SSE2 check and then assuming that everything that supports SSE2 also supports MMX and not providing SSE2 alternatives for all MMX code, and then using it anyway without a proper feature check for MMX itself.

 

Share this post


Link to post
3 hours ago, Wadmodder Shalton said:

Now all we need for our web browsers is the W3C and WHATWG to drop 32-bit support once AMD announces a similar x86-S architecture.

I know this is kind of your thing, but this sentence is particularly incoherent.  These groups define standards that have no relation to CPU architectures.  The end of 32-bit support in browsers will come when Google and Mozilla feel the 32-bit user base is small enough to not be worth the effort.  I have to imagine we're getting pretty close to that time.

 

This is the x86S that AMD would implement.  They might work with Intel to change a few things about the proposal, but I wouldn't expect any particular announcement from AMD until they're releasing CPUs following it.

Share this post


Link to post
26 minutes ago, Blzut3 said:

Google and Mozilla feel the 32-bit user base is small enough to not be worth the effort.  I have to imagine we're getting pretty close to that time

 

Chrome auto-updated capable people to 64 bit back in 2017. Firefox has defaulted to 64bit for some time. I could count the number of my customer's computers still running 32 bit OS installations in the last half decade on one hand and still have most of the fingers left. A Steam hardware survey in 2020 showed 32bit installs were only 0.20%. Honestly I am surprised it's even that high; obviously it will have decreased more by now.

 

30 minutes ago, Blzut3 said:

I know this is kind of your thing, but this sentence is particularly incoherent.

 

Yeah I don't know why he's so obsessed with 32bit support dying off.

Share this post


Link to post
On 5/24/2023 at 9:15 AM, HavoX said:

Another good reason for me to ditch that decade-old-yet-reliable Western Digital HDD in favor of a SSD. Not in any hurry though..

If you have a desktop, just use both. My main PC runs two SSDs (for Windows and Linux) and three HDDs (for data).

Share this post


Link to post

meh. this doesn't seem like much of an issue, i can't imagine it'd really affect much of anything. i mean, who the hell even uses 32-bit operating systems on modern hardware?

Share this post


Link to post
33 minutes ago, roadworx said:

who the hell even uses 32-bit operating systems on modern hardware?

 

Most likely people who have only just got into making computers from preexisting computer stuff.

Share this post


Link to post
2 minutes ago, Polyg0n said:

Will this affect 32-bit programs/games both on Linux and windows?

 

If the 32-bit programs/games run fine on your 64-bit OS, then this shouldn't change things afaik.

Share this post


Link to post
Just now, ReaperAA said:

 

If the 32-bit programs/games run fine on your 64-bit OS, then this shouldn't change things afaik.

good. no reason for me to loose my shit yet.

Share this post


Link to post

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
×