This is a most excellent place for technology news and articles.
I sincerely hope that if they come up with a 128bit instruction set they call it "x80" to maintain backwards compatibility with previous set names and be deliberately confusing to everyone.
x87.
Because that would be unambiguous what it means.
But 0x80 is how you'd normally express 128 as hex. So it's relevant. But deliberately confusing.
enemies are joining forces against a common enemy: ARM
ARM has a high probability of blowing a tire.
They have a complex relationship with their licensees which may try to cause self-sabotage trying to pull more of the money home. See the various licensing fights.
If you don't want or need x86, what does ARM have to offer-- in the long term-- over RISC-V, which is much less coupled to a single firm's caprice? We can assume the gap in performance will continue to shrink ovrr time.
You can buy an ARM laptop right now.
And you can't run Linux on it.
You can run Asahi Linux on M1 MacBooks right now. If you didn't see the news, they've even been able to run some relatively modern AAA games with decent enough frame rates. Granted that's only the M1 macs, but there's at least one relatively modern ARM laptop that you can run Linux on.
I'll totally concede that the new Snapdragon laptops aren't running Linux yet. It seems like Qualcomm is taking Linux support seriously, but I'm a bit skeptical as someone who has been absolutely fucked by a shitty Qualcomm kernel module.
Go RISC-V!
"When we step back and look at it, we see that it is a de facto open ecosystem. Open ecosystems benefit from having consortiums with stakeholders, all with a voice in driving the ecosystem forward"
Lol if the x86 ISA is an open ecosystem, CocaCola is an open source beverage.
Trying to read between the lines though, is Norrod implying that they are considering moving toward a more open x86?
The impression I'm getting is that they're thinking about sacrificing various legacy features (32bit being the most obvs one) to bring performance per Watt closer to the competition. And then put in its place a more standardized successor? RIS86? They will likely aim to steer clear of formal (published) standards because they want their trade secrets. However, they will need to simultaneously ensure not to violate antitrust rules.
Might be easier for them to just forfeit trade secrets and embrace a different model altogether.
Aren't they legally required to be, or face monopoly litigation
With the rise of ARM, I don't think they'd need to, since ARM gives enough competition.
At this point, I think they could get away with it. It’s no longer a long shot to argue that x86 has serious competitors.
You're looking for "collusion".
I would imagine that any x64 binary compiled to work with both chipsets would only use the instructions that are common between them. This would mean there's not much gain in developing a new insutruction unless both companies support it.
I'm sure there's some super-optimised stuff out there that targets extra instructions when they're available, but it's very rare.
For some reason Sweeney and Torvalds are working together, that I find the most shocking
AMD and Intel are well-matched competitors right now except for AMD having some money.
