News Intel terminates x86S initiative — unilateral quest to de-bloat x86 instruction set comes to an end

The only way to make x86 not entirely suck, cancelled.

Intel just told the world "go ARM or RISC-V", x86 will never be clean, it will never be the future.

Let's drop this 50 years old tech and let's move on.
Can you read? They have a very good reason for cancelling it and it doesn't mean that it won't happen. They cancelled their solo project for something better, collaboration between Intel, AMD and other companies to help decide what happens with x86.

Move on to what? At the moment there is no other processors that could replace x86, yes there may be desktop ARM CPUs coming but as of right now there is no alternative.
 
The issue is, backward compatibility is aways used as an argument against moving to other architectures.
yes, but again Apple has shown that a proper translation layer can be used for that.

You could make pure x64 and use translation layers for anything that needs x32.

heck you could likely make a physical add in card specifically for 32bit stuff as last resort.

arm is closing in on x64 because x86 is dragging it down (all for sake of backwards compat) from making some further strides....that can't last and sooner its done the better for everyone.
 
Can you read? They have a very good reason for cancelling it and it doesn't mean that it won't happen. They cancelled their solo project for something better, collaboration between Intel, AMD and other companies to help decide what happens with x86.
Agree, and as written in the article, why spend money on something that can be done dividing the expenses.

Move on to what? At the moment there is no other processors that could replace x86, yes there may be desktop ARM CPUs coming but as of right now there is no alternative.
There are a lot of alternatives if you do not need compatibility with x86.
 
yes, but again Apple has shown that a proper translation layer can be used for that.

You could make pure x64 and use translation layers for anything that needs x32.

heck you could likely make a physical add in card specifically for 32bit stuff as last resort.

arm is closing in on x64 because x86 is dragging it down (all for sake of backwards compat) from making some further strides....that can't last and sooner its done the better for everyone.
100% agree. Apple has clearly shown that with some effort translation can work well. It's not perfect, but for 99% of the use cases it works well enough.
 
  • Like
Reactions: artk2219 and King_V
"While we have pivoted away from the x86S initiative, our focus remains on driving innovation and collaboration within the x86 ecosystem." - Intel spokesperson to Tom's Hardware."

Ohhhhhhh reeeeaaaaaallllly?

Got to love contradictory statements.
 
Can you read? They have a very good reason for cancelling it and it doesn't mean that it won't happen. They cancelled their solo project for something better, collaboration between Intel, AMD and other companies to help decide what happens with x86.

Move on to what? At the moment there is no other processors that could replace x86, yes there may be desktop ARM CPUs coming but as of right now there is no alternative.
Move to an ISA that is not LOCKED between 2 companies !!!
Also, it's very possible to emulate x86 binaries on a different ISA with not that much loss (Rosetta, Box64, etc...)

So yes, we deserve this bloated/closed/obsolete crap to go away and enjoy newer/better standards.
 
  • Like
Reactions: artk2219
it doesn't mean that it won't happen. They cancelled their solo project for something better, collaboration between Intel, AMD and other companies to help decide what happens with x86.
It does suggest that either one of the Ecosystem Advisory Group members had a big problem with part of the proposal, or that there's a different, competing proposal that maybe goes even further (but isn't a strict superset). Otherwise, I'd have expected Intel just to transfer its x86S proposal as-is to the new Advisory Group and finalize it under that body.

yes, but again Apple has shown that a proper translation layer can be used for that.
You're missing what this is all about. The stuff proposed for removal in x86S is mostly obsolete OS-level or boot-time stuff. So, the whole question of emulation isn't even relevant.

What it does affect is the potential to boot legacy operating systems on bare metal. However, I have to wonder how many of those ancient OS's would even work on modern systems, due to a variety of other hardware subsystems in modern PCs that they don't support.
 
I honestly don't think that X86S would have saved much silicon.
It would only have alienated a group of users, albeit small.
With today's multi-core CPUs why couldn't they just have one or two legacy or L-cores and the other cores can just be the newer instruction set only, sitting on the same die.
 
With today's multi-core CPUs why couldn't they just have one or two legacy or L-cores and the other cores can just be the newer instruction set only, sitting on the same die.
That wouldn't help, because the point is to simplify the functional view of the system from the OS' perspective. The OS already has to run 16-bit apps in emulation, so it wouldn't help there. As mentioned above, 32-bit apps can still run natively, under X86S.

About the only thing that "legacy cores" would enable you to do is boot a legacy 16-bit or 32-bit OS on just those cores. However, these operating systems seemingly haven't worked on Intel CPUs sold for the past 4 years and I haven't heard anyone crying "bloody murder!" So, the value of that seems dubious.
 
Compatibility is always going to be an excuse, we've been here before with the move to Win64 and how long Win32 kept this back and the mess we now have in Windows. X86 group/Microsoft should say frag it and are say in 2030 next version of Windows will be x64 only. You need compatibility, then thats Windows 11 until 2035 and then EOL. Apple has twice shifted cpu's, because they control the hardware and software, forced everybody to play ball and I think most would agree everybody has benefitted one way or another.
 
I feel canceling it altogether is strange, why they didn't move it first to the x86 Advisory Group and let the group decide whether to cancel it or not?

Maybe intel didn't make a progress in x86s to justify "moving it" in the first place.
 
1624854527_untitled-design-1.jpg