Intel Joins RISC-V International, Will Help With RISC-V Open-Source Software

[coder]

ARM does not make server CPUs, they make CPU design IP that could be made into a server CPU. I doubt they even have a fully designed server chip anywhere at the company.

They make the cores and the interconnect IP. And they have to validate that stuff, before delivering it to customers. So, it's likely they have a reference design they can at least run in simulation. I could believe they even do a limited run of silicon, to do further validation.

Remember: ARM is only as successful as their customers. It's in their interest to make sure all of their stuff is working as advertised and to hold customers' hands (i.e. with validated PCIe and DDR controllers, as part of their reference designs).

[coder]

Makes sense. Intel's own manufacturing capabilities are so bad, they outsourced to TSMC and now need others to fill the void in their fabs.

That was tactical, but their Foundry Services business is strategic.

[coder]

Frankly I don't get the point. That the RISC-V ISA is free doesn't change the fact that manufacturers can (and absolutely WILL) implement antifeatures in the chips.

The point is that RISC-V will be more open to competition. If your software runs on x86, you currently only have 2 real choices of CPU vendors and no option to build your own.

With RISC-V, there can be more competitive options and you could even create your own on a FPGA and run it there. Performance would be terrible, but at least you could theoretically make the hardware totally secure.

[rabcor]

Smart, can't compete with arm if you're paying them royalties can you?

Risc V is their best option, Intel's probably planning to make RISC-V based CPUs or a new RISC-V like one at least, maybe one that's backwards compatible with RISC-V so it can benefit from all the work done to support RISC V?
As for that stuff about wanting people to use their fabs for manufacturing RISC-V chips, that's possible too, but why not both?

Intel needs to go RISC, be that RISC V, ARM or their own alternative to those. RISC is the future for general purpose processing across just about every space. It's go RISC or go broke. Apple's M1 couldn't make this shit clearer. It's fairly certain AMD and Intel are both racing towards RISC architectures right now.

[coder]

My take on this is that Intel is:

1. Trying to create some FUD around ARM, in the near-term, to slow its uptake in the enterprise and datacenter markets.
2. Keeping their options open, in case they do decide to jump into RISC-V with both feet.

[coder]

Intel needs to go RISC, be that RISC V, ARM or their own alternative to those. RISC is the future for general purpose processing across just about every space. It's go RISC or go broke.

I'm not sure it has to be RISC, exactly, but it's definitely not x86.

[paulpach]

tl;dr: No, Intel is not pivoting to RISC-V.
There's some marginal improvements in reducing decoder complexity in RISC-V vs x86, but that matters mostly for low end processors for the embedded market where Intel doesn't really play.

Intel has clearly stated that they want to play in IOT. They want to provide chips, not just for desktop and server, but for all purposes.
x86 is basically dead on arrival in that space. RISCV makes perfect sense for Intel. This is not new, Intel has been working on RISCV for years. They just expanded their involvement.

Intel even tried to replace x86 (Itanium). They make x86 CPUs because that is what their customers want, not because that is what they want. Clearly, Intel is not about to replace their x86 offerings anytime soon, but RISCV lets Intel get into other markets where "intel doesn't really play" right now.

It is not like Intel has to choose between x86 and RISCV, they can absolutely do both for a long time to target different markets.

[pabloski]

Intel needs to go RISC

To be precise Intel has already gone RISC, starting with the P6 architecture. And ARM has gone partially CISC, because the ISA has been updated with complex instructions.

At the current scale of integration, with superscalar execution, multiple pipelines, etc... it has less sense to classify CPU designs as CISC or RISC. CISC exists only in the instruction sets ( microops ), while old school RISC doesn't exist at all anymore.

[coder]

Intel even tried to replace x86 (Itanium). They make x86 CPUs because that is what their customers want, not because that is what they want.

Itanium failed for a lot of reasons, one of which being that customers didn't want to lock themselves into a single vendor. With x86, they had (limited) choice. With RISC-V, they have potentially unlimited choice.

[jabl]

Intel has clearly stated that they want to play in IOT. They want to provide chips, not just for desktop and server, but for all purposes.
x86 is basically dead on arrival in that space.

They keep saying they want to get into phones, embedded, IOT and whatnot. And occasionally they even make some half-hearted moves in those directions. We'll see I guess.

RISCV makes perfect sense for Intel. This is not new, Intel has been working on RISCV for years. They just expanded their involvement.

Well, they made a relatively basic RISC-V softcore. Which is nice, but not like a huge commitment for a $multibillion company.

I guess we'll see how it plays out this time.

Intel even tried to replace x86 (Itanium). They make x86 CPUs because that is what their customers want, not because that is what they want.

Well, the situation with Itanium was pretty different. Back then R&D suggested that superscalar out of order CPU's (both x86 and RISC) would be running into scaling limits, and the solution they came up with was a very wide compiler scheduled VLIW. If this plan would have worked out, not only would Itanium have left x86 and various workstation/server RISC cpu's (which where very much alive back then) far behind performance-wise, but Itanium was protected to the hilt with patents etc. and Intel apparently imagined a world where they would enjoy monopoly profits for a very long time in the future.

Of course, in retrospect we know this didn't work out due to multiple reasons, primarily Intel's compiler team was never able to come up with the magic compiler that would make VLIW work great on general purpose code, and then AMD came out with x86-64, and the proprietary UNIX RISC market was destroyed largely by Linux and x86-64, with a focus on using clustering for HA and handling load rather than going for really really beefy and expensive servers.

RISC-V is different, as there's no protective moat like with Itanium and to a lesser extent x86-64. They may be interested in RISC-V as a way to hurt ARM, but no, I'm sure they have no intentions of helping RISC-V step on their x86 toes.

Clearly, Intel is not about to replace their x86 offerings anytime soon, but RISCV lets Intel get into other markets where "intel doesn't really play" right now.

It is not like Intel has to choose between x86 and RISCV, they can absolutely do both for a long time to target different markets.

Sure, but make no mistake about it, where there is a choice for what to push Intel would certainly choose x86 as they have a dominant position in that ecosystem with other entrants except AMD denied access, whereas in the RISC-V world they'd just be one more seller of cpu's among many others.