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

 

[Josh128]

Apple Has Reportedly Secured More Than Half Of TSMC’s Entire 2nm Capacity To Starve Its Rivals, Including Booking An Entire Plant

A couple of Apple’s competitors are expected to adopt TSMC’s 2nm technology, but the iPhone maker has reportedly kept its rivals at bay by controlling more than half of the initial supply

wccftech.com

 

[511]

Good Luck AMD to get capacity

 

[fastandfurious6]

yeah apple M5/M6 2nm (whichever they do for 2nm) will be beast

but.... what is the real benefit? m3/m4 already is beast enough for every use case

 

[Joe NYC]

Oops sorry @511, I mean to reply to @Joe NYC for this

I think Mercury is good at macro trends (overall revenue, overall units, market shares) but they don't have any tools to assess unit sales per generation of products. Unless they go to Intel and Intel tells them (unlikely).

Other than that, it is just an estimate.

IDC conducts actual surveys, and they many have data on this. But their reports are proprietary / paywalled / expensive.

 

[Joe NYC]

Good Luck AMD to get capacity

I have not seen any comments on TSMC impact of NVidia / Intel deal.

IMO, TSMC has an incentive to ensure that from now on, AMD never has a capacity constraints from TSMC side.

 

[511]

IMO, TSMC has an incentive to ensure that from now on, AMD never has a capacity constraints from TSMC side.

The problem is who will they choose 20% or 10% to prioritize?

I have not seen any comments on TSMC impact of NVidia / Intel deal.

It's mostly a bigger plus for IFS cause the packing is Intel and Custom Xeons are Intel nodes NVIDIA GPU Tiles are from TSMC.

 

[Antey]

Is Zen 6 (mobile/desktop) on N2P? ... or N2P is HPC only?

 

[Josh128]

Is Zen 6 (mobile/desktop) on N2P? ... or N2P is HPC only?

Current word on the street is that mobile is N3P, desktop and enterprise N2 or N2P depending on the product.

 

[Joe NYC]

The problem is who will they choose 20% or 10% to prioritize?

It's mostly a bigger plus for IFS cause the packing is Intel and Custom Xeons are Intel nodes NVIDIA GPU Tiles are from TSMC.

TSMC will prioritize what will produce long term benefit to TSMC.
A customer in hegemonic position with unlimited cash (NVidia), who can go to, or will subsidize competition (Intel) does not benefit TSMC.

 

[Antey]

Thats very curious. N2 seems like a half node in terms of density and speed at the same power but a full node at power savings. That power saving would be really good on mobile.

N3E vs N5 V1.0 is +15% perf, -25-30% power, x1.7 logic density.
N3P vs N3E is +5% perf, -5-10% power, x1.04 logic density.
N2 vs N3E is +10-15% perf, -25-30% power, x1.15 logic density.
N2P vs N3E is 18% perf, -36% power, x1,2 logic density.

Now that we are here.

A16 vs N2P is 8-10% perf, -15-20% power, x1,07-1.10 logic density.
A14 vs N2P is 15% perf, -30% power, x1.2 logic density.

Cannot wait for the allmighty A7 with +3-5% perf, -10% power, x1.1 logic density over A14. And that spending 100+ billion dollars and using new crazy tech to reduce voltage leakage and other new problems.

 

[Joe NYC]

Current word on the street is that mobile is N3P, desktop and enterprise N2 or N2P depending on the product.

BTW, AMD has put a lot of effort into N3P, and based on my deep dive (or actually, jumping in a rabbit hole), AMD has the first and only product on N3P - Mi350.

It is going to become a bigger node for TSMC, when NVidia puts Rubin on it and AMD puts a lot of client silicon on it.

I don't know about Apple, I don't think Apple has used or is planning to use N3P

 

[Antey]

isn't A19 using N3P right now?

 

[poke01]

Apple Has Reportedly Secured More Than Half Of TSMC’s Entire 2nm Capacity To Starve Its Rivals, Including Booking An Entire Plant

A couple of Apple’s competitors are expected to adopt TSMC’s 2nm technology, but the iPhone maker has reportedly kept its rivals at bay by controlling more than half of the initial supply

wccftech.com

View attachment 130620

View attachment 130619

Not suprising. Apples got the iPhone fold next year.

 

[poke01]

isn't A19 using N3P right now?

Yes it is

 

[adroc_thurston]

Apple Has Reportedly Secured More Than Half Of TSMC’s Entire 2nm Capacity To Starve Its Rivals, Including Booking An Entire Plant

A couple of Apple’s competitors are expected to adopt TSMC’s 2nm technology, but the iPhone maker has reportedly kept its rivals at bay by controlling more than half of the initial supply

wccftech.com

View attachment 130620

View attachment 130619

More FUD. Boring.

 

[poke01]

Apple Has Reportedly Secured More Than Half Of TSMC’s Entire 2nm Capacity To Starve Its Rivals, Including Booking An Entire Plant

A couple of Apple’s competitors are expected to adopt TSMC’s 2nm technology, but the iPhone maker has reportedly kept its rivals at bay by controlling more than half of the initial supply

wccftech.com

View attachment 130620

View attachment 130619

Wccftech is so stupid. Apple has the demand to book entire plants, they don’t do it to starve other companies.
The iPhone moves a tons of volume, more than any consumer product.

 

[poke01]

More FUD. Boring.

Let me guess, you believe that AMD brings in more money for TSMC. Tell me what is AMDs volume compared to Apples?

 

[adroc_thurston]

Let me guess, you believe that AMD brings in more money for TSMC.

No, it's just that there's enough N2 for everyone.
But also yes, between GPGPU slop and eating Intel's share, every AMD wafer shipped is a net win for TSM.

Tell me what is AMDs volume compared to Apples?

Doesn't matter, AMD and NV volumes grow TSMC wafer run rates, Apple doesn't. Apple lives in stale bread markets.

 

[Doug S]

Not suprising. Apples got the iPhone fold next year.

I doubt that will increase iPhone unit sales, or is intended to. Every fold sold will be one less Pro Max sold. Higher ASP, but same number of A20P SoCs needed.

 

[poke01]

I doubt that will increase iPhone unit sales, or is intended to. Every fold sold will be one less Pro Max sold. Higher ASP, but same number of A20P SoCs needed.

Kuo is anticipating 10 million iPhone folds to be sold. That’s half of the total number of S24 ultra sales numbers.

 

[Doug S]

Kuo is anticipating 10 million iPhone folds to be sold. That’s half of the total number of S24 series sales numbers.

Yes but unless those buyers are coming from Android it is just shifting sales from other iPhone models, so it doesn't imply an increase in the number of A20P SoCs needed.

 

[poke01]

Yes but unless those buyers are coming from Android it is just shifting sales from other iPhone models, so it doesn't imply an increase in the number of A20P SoCs needed.

People buy multiple phones. I see that with Samsung users, they buy the Fold and the S25 Ultra. It will be more apparent with Apples user base.

 

[adroc_thurston]

People buy multiple phones

they, in fact, do not.

 

[Joe NYC]

isn't A19 using N3P right now?

It actually does. As I said in my post, I did not check Apple.

 

[Joe NYC]

they, in fact, do not.

Also, with rate of improvement in phones slowing, there is a lot less reason to upgrade, unless the phone is stolen or broken.