TSMC (TSM) - Share Chat

This was exactly why I wanted to hold some TSMC. Apple switching to ARM only added weight to that decision. I don’t hold the stock solo though, I’ve got my largest holding via Baillie Giffords Positive Change fund.

Intel need to pull their finger out.

People have been speculating about Intel moving to TSMC fab for quite a while, but the continued troubles with 10nm and 7nm yields suggested by the delays announced yesterday might now be forcing the issue.

The article below discusses what this might mean for TSMC

Well that materialised very quickly.

With AMD and Intel fighting over a limited number of wafers the ball should be firmly in TSMC’s court, it’ll be interesting to see how much profit the 2 designers can hold on to.

[Edit] - I’ve seen some speculation that this may be poor quality journalism and the Intel order is not 180,000 wafers but 180,000 chips (~2000 wafers) - I guess we’ll have to wait and see how big this news is.

AMD’s next report may help TSMC’s share price rise further.

TSMC just held their 2020 Technology Symposium, based on their roadmap and the yields they are achieving on 5N I’m increasingly confident that they can retain (or even extend) the lead they have taken over Samsung and Intel into 2021 and 2022. I think their massive installed EUV capacity is really starting to give them an edge.

No idea how this this technological advantage will impact their earnings or share price though, seems the markets are already pricing an optimistic outlook for them.

1 Like

Nice to see TSMC’s claims regarding their progress with 5N all but confirmed by Apple’s A14 announcement today (and the corresponding share price bump).

Should give confidence for AMD (Zen4, RDNA3) and Qualcomm (Snapdragon 875) who will be the other big upcoming 5N customers next year. Hopefully they can achieve similar gains on their new chips.

2 Likes

An interesting piece looking at estimating wafer costs for each of TSMCs nodes.

The bit I found interesting is when you compare a theoretical chip price across nodes.

But even at current costs it makes a great sense for makers of highly-complex chips to use TSMC’s leading-edge process because of its high transistor density as well as performance. Based on the numbers provided, it costs $238 to make a 610mm2 chip using N5 and $233 to produce the same chip using N7.

Even the staggeringly high cost of a N5 wafer is economical because of the increased density. That’s not something I fully appreciated.

Just in case anyone is interested, I wanted to review my decision 3 months in to check I’m still happy with this - not advice.

On EUV capacity

TSMC should have received / installed 50 of ASML’s EUV machines by the end of 2021 with Samsung expected to lag behind at 25.

https://www.digitimes.com.tw/tech/dt/n/shwnws.asp?cnlid=1&id=0000594851_Y6H7KHN76G4AXI0F62YVR (need to translate)

If ASML’s production capacity is as expected for 2020 and 2021 this means TSMC should retain its dominant position in small feature sizes because there simply wont be enough capacity going anywhere else.

If EUV lithography is going to be pretty much essential for any advanced process going onwards this advantage should be almost insurmountable in the near term.

On new N3 and N2 process

So while it looks like we can expect TSMC to retain the advantage on feature sizes, there is a potential bump in 2022 when they will still be using FinFET while Samsung will have moved to GAA transistors. It will be interesting to see if Samsung (and its customers) can capitalise on their GAA advantage before TSMC’s N2 node arrives in 2023 which is expected to be using GAA transistors.

Review

My biggest concerns at this point are unchanged:

  • Political risk
  • Global downturn in semiconductor demand
  • Natural disaster
  • Samsung’s lead in GAA transistors

I am now also a bit worried that TSMC is getting very expensive with a market cap of ~$400bn and a trailing p/e of ~25, hopefully it can grow into that valuation. On balance I’m happy to to hold TSMC despite the significant increase in price, it feels somewhat justified.

2 Likes

I enjoyed today’s article from the FT team :+1::

TSMC is many years and many tens of billions in investment ahead of the competition, I remain long on TSMC the 5th wave of computing has only just begun.

3 Likes

I wished that article would have mentioned ASML, who make the machines that make the machines. They also didn’t do a thorough find and replace of the words “Games Workshop” :joy:

Another interesting company would be NXP, who design chips for the car industry.

Some further background on TSM in a recent FT article.

https://arstechnica.com/gadgets/2021/03/tsmc-how-a-taiwanese-chipmaker-became-a-linchpin-of-the-global-economy/

1 Like

TSMC to Spend $100 Billion Over Three Years to Grow Capacity

https://www.bloomberg.com/news/articles/2021-04-01/tsmc-to-invest-100-billion-over-three-years-to-grow-capacity

It’s insane that one company is looking to spend about as much as the UK’s ‘once in a generation’ post-corona infrastructure investments.

5 Likes

Looks like Fundsmith has started a small position ($13M) in TSMC; their first APAC stock i believe

https://fintel.io/so/tw/2330/fundsmith-llp

How much of that 100bn do you reckon will go to ASML for DUV/EUV?

Must be 50+ units at 150 each so perhaps 7.5bn if they are 3400s

I imagine the TSMC N3 fabs might be using the next generation of high NA machine which are rumoured to be north of $200m each so maybe well over $10bn

Actually these unit number are way too low, I’ll do some maths later

1 Like

Is there a reason this is non-ISA only?

It’s not an American company and so it operates as an ADR, which makes in ineligible for ISA.

So from a quick look I think TSMC bought about 25 EUV scanners in 2020 from a total spend of about 18bn so very crudely about 20-25% capex went to ASML EUV.

So that might put a 100bn spend at around 20-25bn on EUV for perhaps 120-130 scanners. Even with ASML set to double volumes that pretty much all their capacity into 2023.

For ArF immersion DUV ASML don’t have a complete monopoly, so Nikon might take a bit.

It’s hard to know how many layers of EUV vs DUV will be used in future process nodes. I think for N5 it’s about 12 layers of EUV for about 60 layers of DUV. Although generally each EUV layer can replace 10 DUV layers so the trend has been for fewer total layers.

DUV scanners have about double the throughput of EUV so you don’t necessarily need many more units so maybe another 150 DUV units for 7.5-10bn

So maybe in total more like 30bn of this to ASML, it’s hard to see how there is going to be enough capacity with both Intel and Samsung also committed to significant expansion, it feels like someone is going to miss out.

I think it’s all priced in already as ASML has continued to surge during the semiconductor shortage.

2 Likes