Intel 18A Process Node Clocks an Abysmal 10% Yield: Report

[btarunr]

In case you're wondering why Intel went with TSMC 3 nm to build the Compute tile of its "Arrow Lake" processor, and the SoC tile of "Lunar Lake," instead of Intel 3, or even Intel 20A, perhaps there's more to the recent story about Broadcom voicing its disappointment in the Intel 18A foundry node. The September 2024 report didn't specify a number to what yields on the Intel 18A node looked like to spook Broadcom, but we now have some idea as to just how bad things are. Korean publication Chosun, which tracks developments in the electronics and ICT industries, reports that yields on the Intel 18A foundry node stand at an abysmal 10%, making it unfit for mass-production. Broadcom validated Intel 18A as it was prospecting a cutting-edge node for its high-bandwidth network processors.

The report also hints that Intel's in-house foundry nodes going off the rails could be an important event leading up to the company's Board letting go of former CEO Pat Gelsinger, as huge 2nd order effects will be felt across the company's entire product stack in development. For example, company roadmaps put the company's next-generation "Clearwater Forest" server processor, slated for 2025, as being designed for the Intel 18A node. Unless Intel Foundry can pull a miracle, an effort must be underway to redesign the chip for whichever TSMC node is considered cutting-edge in 2025.



View at TechPowerUp Main Site | Source

 

[Space Lynx]

Well, now we know why the CEO change. Damn, Intel is having a hell of time. I honestly hope they make a comeback, every economy needs three competitors to thrive imo. Duopolies suck for innovation.

 

[Vayra86]

Pfoooohhh...
So in a very real, practical sense, after 10nm, Intel went into full stagnation mode, they have literally nothing yet. And that's after 10nm being plagued by delays.
Fail harder, I dare you. Its going to be a real achievement. One might say they've taken their quad core approach up in the company culture, stagnation is the golden rule, anything else an exception. Or put differently, they reap what they sow, or more accurately: Karma is a Bitch.

Six whole years of nothing. And nothing on the horizon.

Between this fiasco and Ubisoft's constant failrace I might just run out of popcorn altogether, damn. Its very enjoyable to see the world still works as it should: you underdeliver, you pay the price. Its very positive to see this happening. Failures should not be cultivated.

 

[Onasi]

I am sure this is fine. I already can feel all potential foundry clients lining up.
Jokes aside, who knows, they might still up those yield numbers. But, if 10% is what they are starting off on, that might be difficult or take so long that the node will be uncompetitive.

 

[R0H1T]

So 10A is next year, or it's already in the rearview mirror now

 

[Dristun]

Makes sense, kind of? They're trying to leapfrog all competition on planet Earth in one go after all. Hence their cash-crunch to endure until the yields are good.

 

[FoulOnWhite]

Yikes

 

[Onasi]

@Dristun
Good point. They are effectively engaged in a VERY high stakes game of Russian roulette. If they manage to win they stand to gain MASSIVELY and will have something to show to ask for more government support. But if this gamble fails… I dunno, I can actually see them selling or spinning off their foundries.

 

[jaresk]

Korean publications have also claimed Samsung foundry has better yields then tsmc. Considering how corrupt south Korea is, they might be doing the same thing now.

 

[N/A]

Nova lake on TSMC N2 could be true.

 

[TumbleGeorge]

I hope Intel succeeds with this node and then stops torturing (us) to go moar down!

 

[wNotyarD]

If 20A got canned for this, I can only wonder how bad the 20A project/yields were.

 

[Dragokar]

Well I don't like Pat, but that are just emotions.....beside that I truly believe Intel was run like Boeing in the last 10+ years..... everything was good, there was no real threat ao they slipped into do the same thing as usual and get the benefits.

You simply can't turn on a foundry switch to make good yield.....you need to let the engineers go at it but that costs a lot of money and they probably didn't want to invest that......and in the end it went down step by step.....but it was covered by still good sales and numbers...... till AMD gained market share....

 

[bug]

Pfoooohhh...
So in a very real, practical sense, after 10nm, Intel went into full stagnation mode, they have literally nothing yet. And that's after 10nm being plagued by delays.
Fail harder, I dare you. Its going to be a real achievement. One might say they've taken their quad core approach up in the company culture, stagnation is the golden rule, anything else an exception. Or put differently, they reap what they sow, or more accurately: Karma is a Bitch.

Six whole years of nothing. And nothing on the horizon.

View attachment 374497

Between this fiasco and Ubisoft's constant failrace I might just run out of popcorn altogether, damn. Its very enjoyable to see the world still works as it should: you underdeliver, you pay the price. Its very positive to see this happening. Failures should not be cultivated.

I used to say, despite various failures, Intel has enough talent to turn things around. I'm not so sure anymore. They just seem increasingly overconfident. Or clueless. Or both.

 

[john_]

fixed

 

[Daven]

View attachment 374500


fixed

Subtle. Very subtle. Lol!

 

[nguyen]

Can Intel please stop wasting sands for once

 

[Daven]

So let me get this clear in my head :

Intel 7 is a straight up name change of 10 nm
Intel 4 only used for a few laptop CPUs
Intel 3 only used for a few data center CPUs
TSMC 3 thru 6 used for data center GPUs, desktop CPUs and GPUs, mobile CPUs and GPUs and more
Intel 20A cancelled
Intel 18A horrible yields

Did I miss anything or get something wrong because the above means that nothing really came after 10 nm from Intel?

 

[Tek-Check]

Aoouch...

 

[bug]

So let me get this clear in my head :

Intel 7 is a straight up name change of 10 nm
Intel 4 only used for a few laptop CPUs
Intel 3 only used for a few data center CPUs
TSMC 3 thru 6 used for data center GPUs, desktop CPUs and GPUs, mobile CPUs and GPUs and more
Intel 20A cancelled
Intel 18A horrible yields

Did I miss anything or get something wrong because the above means that nothing really came after 10 nm from Intel?

You are correct. But something doesn't add up. When Intel cancelled 20A, they said they were doing it because 18A was doing so well, 20A isn't needed anymore.

Personally, I don't care much who builds what. Bur as long as demand is so much higher than the supply, we need all the fabs in tip-top shape.

 

[usiname]

So let me get this clear in my head :

Intel 7 is a straight up name change of 10 nm
Intel 4 only used for a few laptop CPUs
Intel 3 only used for a few data center CPUs
TSMC 3 thru 6 used for data center GPUs, desktop CPUs and GPUs, mobile CPUs and GPUs and more
Intel 20A cancelled
Intel 18A horrible yields

Did I miss anything or get something wrong because the above means that nothing really came after 10 nm from Intel?

But 14A will be nice and shinny, Just Wait©

 

[Darmok N Jalad]

Makes for a great case to take billions in taxpayer dollars. Maybe Intel is using the same planning personnel as the ones working on the California High Speed Rail to Nowhere.

 

[Daven]

You are correct. But something doesn't add up. When Intel cancelled 20A, they said they were doing it because 18A was doing so well, 20A isn't needed anymore.

Personally, I don't care much who builds what. Bur as long as demand is so much higher than the supply, we need all the fabs in tip-top shape.

I think everything coming out of Intel right now is in the 'fake it until you make it' mode. But when everything in the company is this uncertain, one can't make future plans or align strategies to take on the competition. Unfortunately, Intel will need to retreat to a previous known position. On the IFS side, this might mean backing up to 3/4 nm and further optimize. This is similar to how Glofo stopped at 12 nm. On the chip design side, Intel might need to focus on their best markets (i.e. Client solutions group) and hammer out a long-term strategy rather than go after ALL markets and potential fail in ALL of them which is where they are headed now.

 

[phints]

I had high hopes for Intel making a comeback with all their Lunar Lake/Arrow Lake and 20A/18A hype. With Lunar Lake somewhat delivering an efficiency improvement but with middling performance, Arrow Lake being a massive disappointment (despite of using the most expensive TSMC 3nm class node) and other than idle efficiency is a waste of sand. Now 18A being pushed back to who knows when... they are done. Completely done to me. I see no reason to even read or think about this company anymore.

 

[Evrsr]

I guess they didn't learn much from 10nm... They are trying to get PowerVia done, which was supposed to come with 20A. With that being canned, there still isn't an implementation on an otherwise mature node.

Having a shrink and PowerVia onto a single process may be too much to swallow at once.