Discussion Leading Edge Foundry Node advances (TSMC, Samsung Foundry, Intel)

Page 39 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

DisEnchantment

Golden Member
Mar 3, 2017
1,590
5,722
136
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.

1587737990547.png
N7 performance is more or less understood.
1587739093721.png

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

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.


1587739615344.png

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.
 

Ajay

Lifer
Jan 8, 2001
15,332
7,792
136
^Paywalled, but from elsewhere TSMC sure is making a huge investment in the USA. When all plants are completed, they will output up to 600,000 wafers/yr. I wonder if this will deep-six Intel's foundry plans.
 

DisEnchantment

Golden Member
Mar 3, 2017
1,590
5,722
136
PHOENIX, Arizona, Dec. 6, 2022 – TSMC (TWSE: 2330, NYSE: TSM) today announced that in addition to TSMC Arizona’s first fab, which is scheduled to begin production of N4 process technology in 2024, TSMC has also started the construction of a second fab which is scheduled to begin production of 3nm process technology in 2026
F21P2 added for N3, HVM 2026
 

Ajay

Lifer
Jan 8, 2001
15,332
7,792
136
Of course not. TSMC's keeping them on an N-1 node, but Apple needs the latest for their flagship chips.
And, apparently, Apple won't be moving any N-1 production for older products back from Taiwan to the USA - which makes sense when I think about it.
 

Ajay

Lifer
Jan 8, 2001
15,332
7,792
136
Also, Ian Cutress did the math for us. 600k wafers per year is only 50kwpm - so two plants will be producing about 1/2 the output of a Gigafab. Sounds a bit odd to me that a $40B investment doesn't even match the output of a Gigafab!
 
  • Like
Reactions: igor_kavinski

Doug S

Platinum Member
Feb 8, 2020
2,202
3,405
136
Interestingly, despite the fact that Tim Cook will attend the upcoming celebratory event headlined by President Biden, Apple is not expected to be using any of the Arizona based Fabs for it's SoCs.


Sure they will, Apple doesn't sell only products on with leading edge chips. They will still be selling iPhone 14 until Sept 2024, there might be an iPhone SE that uses an N5/N4 chip even longer. Ditto for Macs.

Cook said Apple would be "the biggest customer" at this TSMC fab, though that's not saying much since Apple is something crazy like 25% of TSMC's revenue.
 

maddie

Diamond Member
Jul 18, 2010
4,723
4,628
136
Also, Ian Cutress did the math for us. 600k wafers per year is only 50kwpm - so two plants will be producing about 1/2 the output of a Gigafab. Sounds a bit odd to me that a $40B investment doesn't even match the output of a Gigafab!
Maybe one of the reasons that the ex-CEO claimed Arizona product would be ~50% more costly than Taiwanese.
 
  • Wow
Reactions: Ajay

Saylick

Diamond Member
Sep 10, 2012
3,084
6,184
136
Maybe one of the reasons that the ex-CEO claimed Arizona product would be ~50% more costly than Taiwanese.
Hm... I have to guess that the price increase includes the need to ship the completed dies back to Asia where the majority of the packaging and substrate manufacturing is located. I mean, isn't one of the primary benefits of having things fabbed in Asia (specifically Taiwan) is that most, if not all, of the assembly process is within the same region? When you look at a modern AMD CPU for example, it says "Diffused in Taiwan, Assembled in Malaysia" or something like that, no? If so, it kind of defeats the point in diversifying TSMC fabs because it's just one part of the overall production line. Seems like I should look into whether or not the other key players are also opening up shop in Arizona...
 

Doug S

Platinum Member
Feb 8, 2020
2,202
3,405
136
Hm... I have to guess that the price increase includes the need to ship the completed dies back to Asia where the majority of the packaging and substrate manufacturing is located. I mean, isn't one of the primary benefits of having things fabbed in Asia (specifically Taiwan) is that most, if not all, of the assembly process is within the same region? When you look at a modern AMD CPU for example, it says "Diffused in Taiwan, Assembled in Malaysia" or something like that, no? If so, it kind of defeats the point in diversifying TSMC fabs because it's just one part of the overall production line. Seems like I should look into whether or not the other key players are also opening up shop in Arizona...

I saw some people trying to bring this to the attention of those writing the CHIPS bill. Whether money for the less sexy pre and post production stuff was included or it is all for fabs I don't know.
 

moinmoin

Diamond Member
Jun 1, 2017
4,934
7,619
136
Maybe one of the reasons that the ex-CEO claimed Arizona product would be ~50% more costly than Taiwanese.
Aside overall production line being local to Asia as mentioned, the building knowledge as well as fab configuration is likely very optimized locally as well. TSMC organically builds from R&D to Gigafabs, equipment is likely shared and reused to a degree and there is some flexibility of multiple fabs being situated close to each other. Ground up new fabs far away from there obviously can't profit of all that.
 

Ajay

Lifer
Jan 8, 2001
15,332
7,792
136
Sure they will, Apple doesn't sell only products on with leading edge chips. They will still be selling iPhone 14 until Sept 2024, there might be an iPhone SE that uses an N5/N4 chip even longer. Ditto for Macs.

Cook said Apple would be "the biggest customer" at this TSMC fab, though that's not saying much since Apple is something crazy like 25% of TSMC's revenue.
Yeah, the article I read was clearly wrong. It's looking like more and more of these n-1/2 are SoCs going into older/cheaper products that will be assembled in India. Interesting changes ahead in Apple's global manufacturing chain.
 

Ajay

Lifer
Jan 8, 2001
15,332
7,792
136
I just ran across and interesting video on TSMC in the US. One curious point, to me, is that the CHIPS act doesn't appear to have targeted the semiconductor supply chain like I thought it was supposed to (for example, in packaging). Anyway - I thought it might be a good video for those participating in this thread.

 

FlameTail

Golden Member
Dec 15, 2021
1,882
984
106
Will the transition to GAAFET transistors help SRAM scaling kick back into action?

As I understand it, current SRAM use FinFETs, which seemed to have finally hit a wall. Since GAAFET is the successor of FinFET, it should circumvent this problem right? Atleast for short-medium term?
 
  • Like
Reactions: Geddagod

moinmoin

Diamond Member
Jun 1, 2017
4,934
7,619
136
Are Intel's ambitious fab plans breaking apart again?


"The company had said construction was expected to begin in the first half of next year, with production planned to come online in 2027.

But Intel now sees a “difficult market situation” and can no longer commit to the planned start date for construction, Volksstimme reported, pointing to increased costs and a call for higher government subsidies.
"
 
  • Like
Reactions: Geddagod

Roland00Address

Platinum Member
Dec 17, 2008
2,196
260
126
Are Intel's ambitious fab plans breaking apart again?


"The company had said construction was expected to begin in the first half of next year, with production planned to come online in 2027.

But Intel now sees a “difficult market situation” and can no longer commit to the planned start date for construction, Volksstimme reported, pointing to increased costs and a call for higher government subsidies.
"
I can argue a good faith reason why this is occurring, a bad faith one, and so on and so on. Many possibilities it is unknowable but it is speculative and projective remains.

Pretty much higher interest rates makes it more expensive to build fabs, but will also lower demand for fabs, so it is a feedback loop where higher interest rates makes this part of the market less necessary. Likewise there is a demand that is psychic that people really do want more tech, more smart stuff, so on and so on, but higher interest rates delay that to a slightly farther future.
 

Exist50

Platinum Member
Aug 18, 2016
2,445
3,043
136