Intel's death throes

Intel's death throes.
With Windows for ARM now running x86 applications, Intel's monopoly is as good as dead. Desperate, Intel try to scare Microsoft away with patent threats. Of course, it won't work. Meanwhile, the RISCV consortium keeps growing.
arstechnica.com/information-technology/2017/06/intel-fires-warning-shots-at-microsoft-claims-x86-emulation-is-a-patent-minefield/

Other urls found in this thread:

anidb.net/perl-bin/animedb.pl?show=character&charid=87224
twitter.com/SFWRedditVideos

Good. Add in AMD's Epyc obsoleting Intel's entire line of Xeons overnight, and you've got a recipe for multiple shoahs.

Too bad there's one (potentially two) capable ARM cores on the market.

Funny enough, AMD is part of the RISCV consortium. They do, no doubt, have a performant implementation on the works.
Keep in mind x86 is performant despite the ISA, not because of the ISA. Even Intel CPUs do simply translate the x86 ISA to a different internal ISA.
Furthermore, as demonstrated in studies the presentations of you can find in recent RISC-V workshop minutes, code density of RISC-V is on par with amd64.

RISC-V is unfinished uncompetitive meme. Well, at least for now.

BOOM, Rocket and pulp are already competitive in the markets they target. The non-privileged base isa is also finished and stable. Test ASICs are made regularly. Toolchain support (including both llvm and gcc), Linux and other systems (sel4, freebsd included) already have ports. It's moving crazy fast. It's already been used in cameras sold to the public.
With the big names working on RISC-V, success is as good as assured. The market will be flooded with them sooner than we probably expect.

Good, x86 needs to die. Intel will die with it, since switching the ISA is not that difficult for AMD, they did it at least once moving from TeraScale to GCN.

Switching the ISA shouldn't be hard for Intel either.
If they were any smart, they'd be doing it. But all points to them sticking to x86 and missing that train.
They will probably not disappear, they'll show up, late, with an implementation of their own. But they'll just be yet another player in the market by then. Bye bye monopoly.

Considering Intel already tried dropping x86 with Itanic, eh, I think it'll be hard for Intel.

It would be hilarious if AMD joined OpenPOWER and made some efficient PowerISA implementation.

It could have gone somewhere, except they had two insurmountable problems:
- They were trying to do VLIW in a general purpose CPU. Not an easy job for compilers.
- amd64 happened, which forced microsoft to em64t, in turn being read by everybody as intel giving up on Itanium.

More likely IBM will show up with a good RISCV implementation.
IBM is a RISCV member, too.

IBM cares not about efficiency or yields. They like their chips thicc and power hungry, yet delivering maximum performance. It's up to AMD/Google/whatever to develop efficient RISC-V implementation.

Being the fastest is one form of good, too. There's different targets to hit. IBM is not trying to make SoCs for smartphones like qualcomm.

And Qualcomm is not trying to make x86 substitute. Anyway, we'll see. As for now x86 takes the crown in desktop/server space.

>And Qualcomm is not trying to make x86 substitute.
Except for the Microsoft Surface powered with Qualcomm SoC the OP link talks about.
>server space
ISA less relevant there. As long as Linux, nginx, docker and all that crap do run, which do.
There, Intel is scared shitless of AMD's EPYC anyway. AMD had good reputation on the server space with their athlon64 hardware, which will no doubt facilitate a choice of EPYC.

That Qualcomm SoC is underpowered piece of shit. As I said, there's one (maybe two) powerful ARM cores on market.
>EPYC
On June 20th, the shoah comes. The tears of millions of jews will fuel me.

I refuse to take Qualcomm seriously until they can emulate x86 and match an A8-7600 in 45W TDP mode in cinebench.

>A8-7600
That's junk. Zen APUs are coming though.

You don't seem to get it.
Emulation is just to bridge the transition. Expect binaries in whatever new fancy ISA for popular windows software.
They're doing the same apple did when they moved from 68k to PPC, and from PPC to x86. For all that's worth, apple's also shown signs of abandoning x86.

But can it play 2hu?

>2hu
Probably. They're not particularly CPU intensive. I've played most of these (up to what was available in the time, probably th10 or 11) on a year 2000 PC.

You make some fine points but how about you stop avaterfagging with that sterile bitch?

Satori is love, Satori is life. How can you not like Satori?
Anyway, introducing some variety.

Crazy and infertile. How can anyone like her? Anyway, how strong are intels patent claims regarding x86 emulation? If its done without any code from intel then shouldn't it be perfectly legal?

>For all that's worth, apple's also shown signs of abandoning x86
They switch architectures about every 10 years, it's about time they do it again

>Anyway, how strong are intels patent claims regarding x86 emulation?
They're just scaremongering. They tried to pull this sort of shit on other x86 implementators back in the pentium/ppro/p2 era, and lost.
Microsoft aren't gonna be scared by some threats from Intel. If a lawsuit comes to be, they can and will fight.
By merely showing they're worried, Intel has already lost.

RISC IS THE FUTURE

Well x86 is Risc in a way

The ISA sure isn't. It's so bad intel has to translate it into some sort of private RISC internally.

>On June 20th, the shoah comes. The tears of millions of jews will fuel me.
Have they shown any pricing or clock speeds yet? Are they really going to make us wait until it's in shops to find out?

Who is this semen demon?

anidb.net/perl-bin/animedb.pl?show=character&charid=87224

>Your waifu is 6% trash.
What did he mean by this?

Their new waifu system. Chars can be waifu-ed or trashed. That % is the ratio.

64 bit AMD ARM servers? Lovely.