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

[LittleBro]

Some things just don't add up. I'd trust information regarding Intel's 18A yields being poor but I really hope they can improve it to at least 30%.

- Broadcom statement
- Intel reserved part of TSMC's 2N capacity (https://www.techpowerup.com/318454/...anometer-process-for-nova-lake-cpu-generation)
- Intel CEO visited TSMC last month (https://www.digitimes.com/news/a20241105PD227/intel-tsmc-taiwan-ceo-production-capacity.html)
- no information about new customers signing up for Intel 18A

Intel 18A might be worth it for smaller chips, especially for CPU/GPU chiplets but I'd bet TSMC's 2N node will be better and more reliable.

 

[Wirko]

Which tech news website should I trust ?
Rumoured "abysmal 10% yield" for Intel 18A is fake news - OC3D

10% yield for some giant chip close to the 26x33 mm limit translates to a good yield for ~100 mm² chippelettes.

 

[RandallFlagg]

This article was mocked on X by multiple sources (analysts and partners) that would be in a position to know, and that mocking was backed up by none other than Pat Gelsinger.

Finally, Ian Cutress joined in and pointed out that the 8% 'yield' is 65% yield for a smartphone sized SoC - noting that TPU doesn't know what 'yield' means.

And as far as TSMC vs Intel, apples to apples tests winds up being 99% vs 60%, i.e. this:

 

[AcE]

Which tech news website should I trust ?
Rumoured "abysmal 10% yield" for Intel 18A is fake news - OC3D

Maybe generally avoid "rumours" and wait for facts, that way you're never disappointed.

And as far as TSMC vs Intel, apples to apples tests winds up being 99% vs 60%, i.e. this:

Yea I will believe that when it happens, highly highly unlikely. Intel will not overtake TSMCs massive advantage "just like that" over night.

And this is, folks, why I don't trust rumours. Someone says something extremely negative about intel, then someone else says something extremely positive 5 seconds later about the same company.

 

[RandallFlagg]

Yea I will believe that when it happens, highly highly unlikely. Intel will not overtake TSMCs massive advantage "just like that" over night.

And this is, folks, why I don't trust rumours. Someone says something extremely negative about intel, then someone else says something extremely positive 5 seconds later about the same company.

It's based on the defect rate that TPU reported. The reason this report is ridiculed is that the 'yield' is for a chip at the max optical size. In other words, some kind of monster chip.

For something more normal, albeit still large - like Panther Lake - the yield is calculated to be 60%. Yield depends on the size of the chip.

This is a pretty basic error, and one which has TPU being called the "Mainstream Media" of tech. That is not a compliment.

They really should have retracted or amended their story.

 

[AcE]

It's based on the defect rate that TPU reported. The reason this report is ridiculed is that the 'yield' is for a chip at the max optical size. In other words, some kind of monster chip.

Yea I understood that, my reply is just about the "99% vs 60%" remark. Anyway, we will see how Intel foundry turns out, rumours are a nice time sink but only get you so far.

his is a pretty basic error, and one which has TPU being called the "Mainstream Media" of tech. That is not a compliment.

They should either correct the mistake, or delete the whole news entry. I think the former is the way to go. Yea I basically agree with you, just saw the rest of your comment.

 

[Nhonho]

Which tech news website should I trust ?
Rumoured "abysmal 10% yield" for Intel 18A is fake news - OC3D

If this news from the overclock3d website is false, it is a criminal attempt to stop the fall of value of Intel shares.

If Intel's lithography yield is not too low, why did Intel contract TSMC to manufacture the compute die for its new CPUs?

 

[LittleBro]

If this news from the overclock3d website is false, it is a criminal attempt to stop the fall of value of Intel shares.

Damage control. And whole stock market is bullshit. If Elon Musk said today that Apple sucks, Apple shares would go down.
When people team up to punish shorters and start to buy a lot of low prices stock that are used for shorting, stock price raises and shorters loose millions. And that particular stock gets suspended, lol.
Look at MSFT stock and compare it to shitty quality products they deliver. Windows 11 is still bug festival 3 years after release. Stocks are not about real value of the company. See AI stocks, it's all bullshit.

If Intel's lithography yield is not too low, why did Intel contract TSMC to manufacture the compute die for its new CPUs?

Why we don't hear any news regarding good yields from Intel itself is a question in place.

 

[londiste]

If Intel's lithography yield is not too low, why did Intel contract TSMC to manufacture the compute die for its new CPUs?

Yield is only one part of manufacturing. Plus, it is more than likely that they have only a foundry or a few set up to do 18A right now. Assuming that 18A is indeed OK then Intel is in the process of reconfiguring or refitting more foundries to manufacture on 18A. Depending on what the foundries were on, this takes a long while.

Arrow Lake came out in October. As a rule of thumb the actual manufacturing of the dies used started 6 months to a year before that. Preceded by another year or so of preparations. The time frames vary by a lot but all this takes quite a long time. Pretty sure a year or two ago Intel was not sure what was going to happen to 20A or 18A so they were hedging their bets this time around.

 

[GhostRyder]

You say that, but if you look at their enterprise/server chips, I think Epyc has certainly made the gap similar to what Bulldozer was against Sandy Bridge on consumer platforms back then.

Oh for sure, I think its getting close but I still believe at the worst moment AMD was worse than Intel.

 

[Nhonho]

Oh for sure, I think its getting close but I still believe at the worst moment AMD was worse than Intel.

I think Intel secretly gave money to AMD during AMD's crisis period (I have no way of proving it, of course), because if AMD went bankrupt, Intel would become a monopoly and the US government would break it up, just like it did with AT&T.

And now AMD is not being as aggressive with Intel to it can get out of this crisis.

 

[thestryker6]

I think Intel secretly gave money to AMD during AMD's crisis period (I have no way of proving it, of course), because if AMD went bankrupt, Intel would become a monopoly and the US government would break it up, just like it did with AT&T.

Intel settled a 2005 lawsuit AMD filed by writing a check for $1.25 billion in 2009 and this settlement was finalized ~8 months after AMD spun off their fabs. There was no secret money necessary as it was out in the open.

 

[AcE]

Damage control. And whole stock market is bullshit. If Elon Musk said today that Apple sucks, Apple shares would go down.

Then you are greatly overestimating the importance of elon. If elon would talk nonsense like that, people would just laugh more at him than they do anyway nowadays, the stock market of Apple would go up and not down. And Apple and the popularity of Apple is a zillion times higher than little elons with his lies and blown up ego.

 

[Evrsr]

BSPDN was developed on a custom Intel 4 process so if it didn't pan out it wouldn't impact 20A/18A development as those were implementing GAAFET. It should have no bearing on the progress of 18A as they could have just dropped it if they couldn't get it working.

No 18A was never a 2024 node, even the branding of "5N4Y" says that: Gelsinger wasn't hired until 2021 so it was always a 2025 node. Intel 4 had a pretty big delay and Intel 3 took as long as it was originally supposed to after Intel 4, but the Intel 4 delay factors in here. In theory if Intel wasn't lying about the 20A/18A situation 20A would have been mostly on time and 18A will be, but we won't know any of this until next year or if Intel states otherwise on the record.

The plan is more like 2N4Y though. But I agree with the point.

What I will say is that no matter if BSPDN worked on Intel 4. Given the improvements, that should have probably gone into 3N if it was as ready as they said.

I also think there should be some downstream consequences from flipping the chip, when you're dealing with actual products rather than tests. Tests will run at whatever, actual chips will be at 125-500W that will be now impossible to dissipate.
Hopefully, at least 2025 will bring mobile chips in a decent shape.

 

[thestryker6]

The plan is more like 2N4Y though. But I agree with the point.

Intel 7 (10ESF), Intel 4 (was always going to be a short term node and I think it did more harm than good), Intel 3 (long term node, but seemingly hampered by using EUV capacity on Intel 4), Intel 20A (supposedly canceled at the altar of 18A, but only time will bear this out) and Intel 18A (sometime in 2025).

What I will say is that no matter if BSPDN worked on Intel 4. Given the improvements, that should have probably gone into 3N if it was as ready as they said.

It was developed on a test node with the intent of adding it to 20A/18A so I don't really think this would have been a viable path as it was developed independently of the Intel 4 and 3 nodes. It was more about risk mitigation for 20A/18A than anything else.

 

[LittleBro]

Then you are greatly overestimating the importance of elon. If elon would talk nonsense like that, people would just laugh more at him than they do anyway nowadays, the stock market of Apple would go up and not down. And Apple and the popularity of Apple is a zillion times higher than little elons with his lies and blown up ego.

Nah, Musk and Apple was something I quickly came up with, nothing serious, just example.