Discussion Leading Edge Foundry Node advances (TSMC, Samsung Foundry, Intel) - [2020 - 2025]

[DisEnchantment]

TSMC's N7 EUV is now in its second year of production and N5 is contributing to revenue for TSMC this quarter. N3 is scheduled for 2022 and I believe they have a good chance to reach that target.


N7 performance is more or less understood.


This year and next year TSMC is mainly increasing capacity to meet demands.

AnandTech Forums: Technology, Hardware, Software, and Deals

Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

www.anandtech.com

For Samsung the nodes are basically the same from 7LPP to 4 LPE, they just add incremental scaling boosters while the bulk of the tech is the same.

Samsung is already shipping 7LPP and will ship 6LPP in H2. Hopefully they fix any issues if at all.
They have two more intermediate nodes in between before going to 3GAE, most likely 5LPE will ship next year but for 4LPE it will probably be back to back with 3GAA since 3GAA is a parallel development with 7LPP enhancements.

AnandTech Forums: Technology, Hardware, Software, and Deals

Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

www.anandtech.com

Samsung's 3GAA will go for HVM in 2022 most likely, similar timeframe to TSMC's N3.
There are major differences in how the transistor will be fabricated due to the GAA but density for sure Samsung will be behind N3.
But there might be advantages for Samsung with regards to power and performance, so it may be better suited for some applications.
But for now we don't know how much of this is true and we can only rely on the marketing material.

This year there should be a lot more available wafers due to lack of demand from Smartphone vendors and increased capacity from TSMC and Samsung.
Lots of SoCs which dont need to be top end will be fabbed with N7 or 7LPP/6LPP instead of N5, so there will be lots of wafers around.

Most of the current 7nm designs are far from the advertized density from TSMC and Samsung. There is still potential for density increase compared to currently shipping products.
N5 is going to be the leading foundry node for the next couple of years.

For a lot of fabless companies out there, the processes and capacity available are quite good.

---------------------------------------------------------------------------------------------------------------------------------------------------


FEEL FREE TO CREATE A NEW THREAD FOR 2025+ OUTLOOK, I WILL LINK IT HERE

 

[LightningZ71]

It is difficult to get startup capital in China for most industries, unless the central gov't decides that they are strategic in nature. Lithography and high pressure gas turbine blades are two areas that get lots of gov't grants and support, including building vast amounts of infrastructure and gov't supported "alternative" research efforts.

 

[Win2012R2]

18A ain't getting major customer but they were saying 18A is bad which ain't just cause someone doesn't get customer doesn't mean the tech is bad look at Optane and the death it died.

And you are comparing 18A to Optane?

 

[LightningZ71]

That's a... choice.

Optane was good for it's particular uses, but it's nowhere near as impactful as a full logic node.

 

[511]

And you are comparing 18A to Optane?

Analogy
Optane had good tech -> limited customers
18A has good tech -> no external customer
Analogy applies for External Customer
T&C apply 😛

 

[Win2012R2]

Analogy
Optane had good tech -> limited customers
18A has good tech -> no customer

How about we fill in the next, final stage here:

------------
Optane had good tech -> limited customers -> DEATH
18A has good tech -> no customer -> ???
------------

Completely inappropriate analogy but you walked on your own into this pickle

 

[511]

How about we fill in the next, final stage here:

------------
Optane had good tech -> limited customers -> DEATH
18A has good tech -> no customer -> ???
------------

Completely inappropriate analogy but you walked on your own into this pickle

I wanted to say no external customer cause 18A is getting Milked one way or another as for death maybe no external customer will buy it so death for external not internal.

 

[Tigerick]

I wanted to say no external customer cause 18A is getting Milked one way or another as for death maybe no external customer will buy it so death for external not internal.

Got it.


Optane had good tech -> limited customers -> DEATH
18A has good tech -> no external customer -> DEATH

 

[511]

Got it.


Optane had good tech -> limited customers -> DEATH
18A has good tech -> no external customer -> DEATH For External

Fixed it for you

 

[Win2012R2]

I wanted to say no external customer cause 18A is getting Milked one way or another

That also means no serious external customer will risk 14A (and unlike 18A here they will have to make pre-payments)

 

[511]

That also means no serious external customer will risk 14A (and unlike 18A here they will have to make pre-payments)

That is something only time will tell as for prepayment you need to make them at the foundry there is no other way either at TSMC or at Intel or Samsung.

 

[Win2012R2]

That is something only time will tell as for prepayment you need to make them at the foundry there is no other way either at TSMC or at Intel or Samsung.

Why risk so much money with either Intel or Samsung? Much easier to go with the winner - TSMC, if they fail then nobody could have done it basically (and no competitor of theirs will do better, that's a safe bet), but yeah unknown external customers will tell Intel that their 14A PDK is "looking good!" and encourage them to continue so that they can get 10% discount at TSMC, which is what I reckon happened with 18A under Pat, but now new CEO does not want to play this losing game, the problem is he will lose then anyway.

 

[511]

Why risk so much money with either Intel or Samsung? Much easier to go with the winner - TSMC, if they fail then nobody could have done it basically (and no competitor of theirs will do better, that's a safe bet), but yeah unknown external customers will tell Intel that their 14A PDK is "looking good!" and encourage them to continue so that they can get 10% discount at TSMC, which is what I reckon happened with 18A under Pat, but now new CEO does not want to play this losing game, the problem is he will lose then anyway.

And he will beg government for funding otherwise he will say Intel is not doing this so government will fund it more.
As for TSMC it doesn't mean that if TSMC can't do it than no one can TSMC only rose through in last 12-15 years they had fair share of failure as well.

 

[Win2012R2]

And he will beg government for funding

Yeah that worked real well so far: he could have sold shares in the first place and get money without strings attached.

Having said that I reckon (if they obey to CHIPs act strings for that money) he is counting on his expensive lawyers easily arguing in 3 years that these shares were issued in violation of the CHIPs are and should be annulled.

 

[511]

Yeah that worked real well so far: he could have sold shares in the first place and get money without strings attached.

what shares?

 

[511]

Does Washington’s Investment Save Intel?

The U.S. needs Intel... but Intel needs customers.

substack.com

 

[Khato]

Good analysis. The real key point of an alignment between Intel and the US Government is the almost certainty of some manner of 'encouragement' for US companies to make use of Intel Foundry. I recall one of the articles on the deal quoted Bessent as saying something to the effect that they have no plans of pressuring other companies to use Intel... which I tend to take as confirmation that such is exactly what they plan to do.

I wouldn't have a problem with this as I don't want TSMC to reign supreme. There are plenty of products which don't require leading edge silicon. My favorite example being that NVIDIA's cash-cow products aren't on the leading edge TSMC process. Start pulling the trailing edge nodes out from under TSMC and you'll have comparable financial impact as you would with leading edge.

 

[511]

which I tend to take as confirmation that such is exactly what they plan to do.

Lmao 🤣 🤣 ofc they are gonna do that the US government don't want to rely on TSMC they will find ways to do it in the name of national security.

 

[DrMrLordX]

18A ain't getting major customer but they were saying 18A is bad which ain't just cause someone doesn't get customer doesn't mean the tech is bad look at Optane and the death it died.

That wasn't the point. The point was that Reuters hasn't run any articles bashing 14A which is in stark contrast to the treatment 18A got from the press last year. Whether or not you felt that particular article was fair is irrelevant since it proved to be true.

That also means no serious external customer will risk 14A (and unlike 18A here they will have to make pre-payments)

We don't know that yet. Nobody's really said who is looking at 14A (or even 18AP). Still waiting on that Reuters hit piece.

 

[Win2012R2]

what shares?

10% that Intel had to give US Govt in exchange for the money from CHIPs act

 

[511]

10% that Intel had to give US Govt in exchange for the money from CHIPs act

Those are new shares that are issued and getting government as a shareholder is better than selling in Open Market it boosted the share price while selling in open market wouldn't have done se

 

[regen1]

Those are new shares that are issued and getting government as a shareholder is better than selling in Open Market it boosted the share price while selling in open market wouldn't have done se

Think the point he is making is that Intel had to give up ~10% of its market share to the US Govt. for the money which it was already approved for in the Chips Act.
The thing that's supposedly changed here other than what you have mentioned is that they are getting it in lump-sum and before relevant fab-side milestones.

That wasn't the point. The point was that Reuters hasn't run any articles bashing 14A which is in stark contrast to the treatment 18A got from the press last year.

Won't it be a bit too early to go after 14A? But with those Reuters/Max Chutney guys you never know, could be around the corner.

 

[DrMrLordX]

Won't it be a bit too early to go after 14A? But with those Reuters/Max Chutney guys you never know, could be around the corner.

Probably, yes. But the Broadcom piece was from early 2024 so expect something any day now if it's going to happen at all.

 

[511]

Think the point he is making is that Intel had to give up ~10% of its market share to the US Govt. for the money which it was already approved for in the Chips Act.
The thing that's supposedly changed here other than what you have mentioned is that they are getting it in lump-sum and before relevant fab-side milestones.

Won't it be a bit too early to go after 14A? But with those Reuters/Max Chutney guys you never know, could be around the corner.

Max Chutney new Chutney flavor ?

Probably, yes. But the Broadcom piece was from early 2024 so expect something any day now if it's going to happen at all.

The same dude did QCOM buying Intel rumors

https://www.reuters.com/business/qualcomms-potential-bid-intel-turns-spotlight-their-products-2024-09-24/

 

[regen1]

Max Chutney new Chutney flavor ?

The "Exclusive" from Reuters/Max flavor

 

[DrMrLordX]

The same dude did QCOM buying Intel rumors

Then you should be glad that he hasn't written any hit pieces on Intel 14A yet.