News Intel Foundry and Arm to Collaborate on 1.8nm Mobile SoCs

[Admin]

Intel and Arm to cooperate on design technology co-optimization for Intel 18A node.

Intel Foundry and Arm to Collaborate on 1.8nm Mobile SoCs : Read more

 

[Giroro]

Intel's 18A node still isn't 1.8 nm nor "1.8 nm class". It's not expected to have any features that are even remotely close to 1.8nm lithography.

It's extremely disappointing that their "cynical size-agnostic rebrand to confuse people into to thinking their transistors are smaller and therefore better" actually worked as effectively as it has.
Are we also going to forget that their little efficiency cores are not full-sized fully-featured cores?

 

[rluker5]

Neither are Zen cores or Skylake cores compared to p cores.

 

[JamesJones44]

Intel's 18A node still isn't 1.8 nm nor "1.8 nm class". It's not expected to have any features that are even remotely close to 1.8nm lithography.

It's extremely disappointing that their "cynical size-agnostic rebrand to confuse people into to thinking their transistors are smaller and therefore better" actually worked as effectively as it has.
Are we also going to forget that their little efficiency cores are not full-sized fully-featured cores?

This is an industry standard these days. TSMC's 4nm node is actually 5nm, but it's enhanced so they give it a new number. Advertised NM in the "chip" world became obsolete as a unit of comparison long ago.

 

[cyrusfox]

If you can't beat them (ARM) Join them (invite them to use your latest node). Interesting times ahead.
Lets see how these multiple nodes roll out.

 

[jkflipflop98]

Intel's 18A node still isn't 1.8 nm nor "1.8 nm class". It's not expected to have any features that are even remotely close to 1.8nm lithography.

It's extremely disappointing that their "cynical size-agnostic rebrand to confuse people into to thinking their transistors are smaller and therefore better" actually worked as effectively as it has.
Are we also going to forget that their little efficiency cores are not full-sized fully-featured cores?

Right. Right.

Now point to the structure on TSMC's 3nm node that's actually 3nm. We'll wait.

 

[kjfatl]

Intel's 18A node still isn't 1.8 nm nor "1.8 nm class". It's not expected to have any features that are even remotely close to 1.8nm lithography.

It's extremely disappointing that their "cynical size-agnostic rebrand to confuse people into to thinking their transistors are smaller and therefore better" actually worked as effectively as it has.
Are we also going to forget that their little efficiency cores are not full-sized fully-featured cores?

For decades Intel reported actual geometry numbers while their competition used numbers that were 30% smaller than actual. They did this specifically to make uninformed investors think they were somehow better than Intel and it worked. Since Intel was unable to end this deceptive practice done by TSMC, Samsung, Global Foundries and others., they began reporting using the same yardstick. Does this suddenly make their product significantly better and turn them back into a technology leader. The answer in the eyes of most investors is YES! They are suddenly a process generation ahead of where they were under the old naming convention.

In a year or two, you can expect TSMC and Samsung to announce their 10A process based on 1/3 wavelength numbering. Actual features will be 30A

 

[bit_user]

If you can't beat them (ARM) Join them (invite them to use your latest node).

This isn't entirely new. Intel previously announced plans to fab 3rd party SoCs, including some with ARM cores, like 5 years ago.

 

[TerryLaze]

This isn't entirely new. Intel previously announced plans to fab 3rd party SoCs, including some with ARM cores, like 5 years ago.

Intel has been making their own RISC/ARM CPUs since all the way back in the 80' ,that's a few years more than just 5.
https://en.wikipedia.org/wiki/Intel_i960

If you can't beat them (ARM) Join them (invite them to use your latest node). Interesting times ahead.
Lets see how these multiple nodes roll out.

The last time ARM could realistically beat intel was back in the 90' where amiga apple and other micros all moved to risc CPUs, that ship has sailed now.
ARM caters to a completely different market now from what intel is.

 

[bit_user]

Intel has been making their own RISC/ARM CPUs since all the way back in the 80' ,that's a few years more than just 5.
https://en.wikipedia.org/wiki/Intel_i960

That's misleading.

First, don't confuse RISC with ARM. Not to get in a debate over whether ARM is truly RISC, but it just doesn't benefit anyone's understanding to muddle them together. Second, the i960 has been discontinued for about 15 years, now.

They also inherited a line of ARM-based microcontrollers from DEC, called StrongARM, which they followed with XScale and sold off to Marvell, in 2006.

More importantly, this isn't Intel designing ARM cores (unlike the above examples), but rather partnering with ARM to help them optimize their IP for Intel's 18A process node.

The last time ARM could realistically beat intel was back in the 90'

Cool story. Let's wait and see how Nvidia's Grace compares to Sapphire Rapids.

On a per-vCPU basis, Amazon's Graviton 3 already trashed their Ice Lake SP instances (as well as Milan).

ARM caters to a completely different market now from what intel is.

No, but you saying that tells me Intel is definitely worried about ARM.

Sure, ARM plays more in the embedded space than Intel. ARM also dominates phones, unlike Intel. However, ARM is gaining penetration into the laptop segment and on pace to overtake x86 in the cloud, in just a few years.

 

[JamesJones44]

Cool story. Let's wait and see how Nvidia's Grace compares to Sapphire Rapids.

On a per-vCPU basis, Amazon's Graviton 3 already trashed their Ice Lake SP instances (as well as Milan).

Adding to that, Apple Silicon holds it's own against Intel/AMDs laptop parts while using less power.

ARM is very capable. Vertical scale has always been the question for RISC based processors, for now I personally think this is still an open question, but it has been slowly catching up to its CISC based counterparts now the power consumption has become a concern at all levels of processing.

 

[TerryLaze]

That's misleading.

"If you can't beat them, join them. "
Was what you responded to, I just pointed out that it's not about intel beating or joining anybody, they have been making ARM/RISC since ages.

Cool story. Let's wait and see how Nvidia's Grace compares to Sapphire Rapids.

Yeah we waited 40 years, let's wait another 40 to see what will happen.

No, but you saying that tells me Intel is definitely worried about ARM.

If intel where worried they would buy out ARM or one of the patents they need to be useful to people.
They wouldn't groom them to become a big customer of their foundry service.
Intel knows that they can make a buttload of money without having to fear any competition.

 

[bit_user]

Vertical scale has always been the question for RISC based processors,

What do you mean by "Vertical scale"?

 

[bit_user]

If intel where worried they would buy out ARM or one of the patents they need to be useful to people.

Given Nvidia's acquisition of ARM was thrown out on antitrust grounds, there's 0% chance Intel would be allowed to acquire them, and I'm sure they know that.

As for patents, I don't understand what you're proposing, or how that would be any different than the situation that currently exists. Don't tell me Intel isn't pursuing it's usual, aggressive patent strategy. And yet, that hasn't been enough to keep ARM down.

Somehow, no matter what situation Intel finds itself in, you always manage to come out swinging. Even if you have to whip up some nonsense.

They wouldn't groom them to become a big customer of their foundry service.

IFS is too desperate for customers, and quite a lot of the chips being made on leading nodes contain ARM cores. Intel really has no choice but to hold its nose with one hand, while it grasps ARM's hand with the other.

 

[jkflipflop98]

I mean, if everything turns completely upside down there's nothing stopping Intel from making their own ARM parts. Maybe even ARM parts that can also execute x86 instructions. Who knows?

 

[bit_user]

I mean, if everything turns completely upside down there's nothing stopping Intel from making their own ARM parts.

I've been saying this for years. People have this idea that Intel and AMD will be x86 until the end of time, but I'm sure they'll jump ship if/when they deem it advantageous to do so. That has nothing to do with helping ARM optimize its IP, however.

It will be one of the most closely-guarded secrets, since customers getting wind of such plans might get spooked and shy away from continuing to build out x86-based infrastructure. You don't want to cannibalize the sales of your existing products, too much before the new product even hits the market!

If we're talking specifically about ARM, then it'll be key to watch how the ARM vs. Qualcomm/Nuvia litigation plays out. Essentially, ARM is trying to force Qualcomm to buy an architectural license that forces all of their SoC customers to pay royalties. If ARM wins (and what I've read suggests it likely won't), then that should push any prospective entrants more towards RISC-V than ARM. Although, while the ARM server market is now quite mature, the same cannot be said for RISC-V.

Maybe even ARM parts that can also execute x86 instructions. Who knows?

Apple has demonstrated good enough support at JIT execution of x86 that I think it probably doesn't warrant the additional silicon. But, who knows?

There are some semantic differences between the ISAs, which I'd assume would be somewhat complicated to architect the backend around, but I've been (indirectly) contradicted by Jim Keller, who said they launched the (mothballed) K12 project when they considered how much commonality there was behind the decoder.

 

[JamesJones44]

What do you mean by "Vertical scale"?

Performance on an IPC biases

 

[JamesJones44]

I've been saying this for years. People have this idea that Intel and AMD will be x86 until the end of time, but I'm sure they'll jump ship if/when they deem it advantageous to do so. That has nothing to do with helping ARM optimize its IP, however.

Intel technically already did it once with Itanium. Had Intel done a better job with their compiler support for Itanium it's possible that they could have stopped producing x86 altogether and just came up with a transition plan.

 

[bit_user]

Had Intel done a better job with their compiler support for Itanium it's possible that they could have stopped producing x86 altogether and just came up with a transition plan.

Not sure about that. I once heard a compelling argument that business customers (aside from those already inclined towards mainframes and the like) were reluctant to adopt IA64, for fear of vendor lock-in. With x86, we had at least 3 vendors, at the time. Intel built a patent fortress around IA64 that eliminated the possibility of any clones.

The cool thing about ARM is that it doesn't make its own CPUs and it's very protective of ISA compatibility between implementations. So, you're virtually guaranteed always to have multiple vendors to choose from.

 

[TerryLaze]

Given Nvidia's acquisition of ARM was thrown out on antitrust grounds, there's 0% chance Intel would be allowed to acquire them, and I'm sure they know that.

So you think that intel has only started to fear ARM in the last few years?! After nvidia tried to buy ARM?!

As for patents, I don't understand what you're proposing, or how that would be any different than the situation that currently exists. Don't tell me Intel isn't pursuing it's usual, aggressive patent strategy. And yet, that hasn't been enough to keep ARM down.

Intel isn't trying to keep ARM down that's the point of my post, intel needs ARM for their own products like their FPGAs.

IFS is too desperate for customers, and quite a lot of the chips being made on leading nodes contain ARM cores. Intel really has no choice but to hold its nose with one hand, while it grasps ARM's hand with the other.

They have 70 bil in cash laying around, they have all the choice they want.
It's not like they they are struggling and have to take every customer just to get by.
A better ARM infrastructure will help intel make their own products better as well.

 

[JamesJones44]

Not sure about that. I once heard a compelling argument that business customers (aside from those already inclined towards mainframes and the like) were reluctant to adopt IA64, for fear of vendor lock-in. With x86, we had at least 3 vendors, at the time. Intel built a patent fortress around IA64 that eliminated the possibility of any clones.

The cool thing about ARM is that it doesn't make its own CPUs and it's very protective of ISA compatibility between implementations. So, you're virtually guaranteed always to have multiple vendors to choose from.

Good point on the lock in.

My primary reason for pointing to it was Intel was willing to attempt a change/support more than one ISA once before, there is no reason to think they wouldn't do it again if they felt it would help their business.

 

[bit_user]

They have 70 bil in cash laying around, they have all the choice they want.

Intel had to cut their dividend to fund the buildout of future fabs. That should put things in perspective.

It's not like they they are struggling and have to take every customer just to get by.

The simple fact is that Intel doesn't have enough volume from its own products to support the fabs, any longer. They need additional volume, in order to afford the continuing investments needed to stay competitive with TSMC and Samsung.

 

[TerryLaze]

Intel had to cut their dividend to fund the buildout of future fabs. That should put things in perspective.


The simple fact is that Intel doesn't have enough volume from its own products to support the fabs, any longer. They need additional volume, in order to afford the continuing investments needed to stay competitive with TSMC and Samsung.

They did cut their dividend to not touch their reserves, it's a huge gigantic difference.

The simple fact is that any company does market research before doing anything, I know you think that absolutely everybody at intel is a bumbling fool but I'm sure they know what they are doing, they don't have enough volume from their own products to support the fabs, any longer" ??
They are building UP volume, many times over, from a point where they had maybe close to full capacity, of course they don't have enough volume that's the point of expansion, to increase the volume.
They are building up their GPU market, they are starting to put buttloads of cache into their server CPU/GPUs, they are going to need a lot more volume going forward, and of course they will also want to have external customers for any down/slow time but I doubt it's anyway near to essential for them.

 

[bit_user]

They did cut their dividend to not touch their reserves, it's a huge gigantic difference.

They wouldn't dare cut their dividend, if they didn't have to. I also don't know where you get that $70B figure. From what I see, they only had $23.3B in cash and short-term investments, last Dec. And when you're making multi-$B losses, one quarter after another, you can't afford to burn through all of your savings.

Considering their total liabilities are $78.8B, losing that cushion of $23.3B in cash would destroy their credit, meaning they'd have to borrow on extremely unfavorable terms. And $23.3B doesn't go nearly far enough, when we're talking about an an annual CapEx of ~$20B.

I know you think that absolutely everybody at intel is a bumbling fool

When did I ever say that, or anything even close?

I'm sure they know what they are doing,

Yes. They know the cost of new nodes is increasing exponentially, and they're facing stiff competition in their core markets, meaning they can't make up the difference by either charging more or selling more. The economics of leading-edge semiconductor manufacturing are very harsh and unforgiving. They're not opening their fabs to external customers to be nice, but rather because it's the only way to keep their fabs competitive.

of course they will also want to have external customers for any down/slow time but I doubt it's anyway near to essential for them.

Not only is it essential now, but when IFS is spun out into a fully independent company, it will need a large and diverse customer base to support its continued success.

Believe me: if there's anything I trust people at Intel can do, it's math.

 

[TerryLaze]

They wouldn't dare cut their dividend, if they didn't have to. I also don't know where you get that $70B figure. From what I see, they only had $23.3B in cash and short-term investments, last Dec. And when you're making multi-$B losses, one quarter after another, you can't afford to burn through all of your savings.

Considering their total liabilities are $78.8B, losing that cushion of $23.3B in cash would destroy their credit, meaning they'd have to borrow on extremely unfavorable terms. And $23.3B doesn't go nearly far enough, when we're talking about an an annual CapEx of ~$20B.

That 23bil is their spending money, the amount they made or are investing/spending short-term.
Retained earnings is what they saved up since they opened up.

Intel Reports Fourth-Quarter and Full-Year 2022 Financial Results

www.intc.com

Retained earnings 70,405

When did I ever say that, or anything even close?

Indirectly, by everything you say.
You think that they made plans to build up that many FABs without having any idea of what they are going to do with them...