TSMC Q323 Earnings Call Full Transcript: Question 1 to Question 5

---

---

Question 1. Gokul Hariharan from JP Morgan

Operator: Gokul Hariharan from JP Morgan.

Gokul: Yeah, hi, congratulations on the great result and thanks for the details on N3 and N2. My first question is on the technology leadership.

Gokul: Given we are hearing a lot of competitive messaging from your US IDM competitors slash customer in the last few months, Intel seems to think that they will be getting into technology or process technology leadership in 2025. Just wanted to hear what does TSMC think of Intel’s claim?

Gokul: And when TSMC thinks about customer engagement, do you feel you will lose a little bit of market share to Intel when it comes to the N2 or the first generation of nanosheet transistors? Or you think your very high market share in N3 will continue into N2? That’s my first question.

Host: Okay, thank you, Gokul. Please allow me to summarize your first question. So Gokul’s first question is about technology leadership and competition with, I think, particularly IDM. He notes this IDM is very messaging about taking over process technology leadership from TSMC.

Host: And so Gokul’s question is, how do we see this? Are we concerned that we will lose market share by nanosheet or N2 to this IDM, given their claims of regaining process technology leadership? Is that correct, Gokul?

Gokul: Yeah, that’s right. Thanks.

C.C.: Well, Gokul, this is C.C. Wei.

C.C.: Let me answer your question with a very simple answer, say no. But, well, I will stay a little bit long. Actually, we do not underestimate any of our competitors or tax them lightly.

C.C.: Having said that, our internal assessments show that our N3P, now I repeat again, N3P technology demonstrated comparable PPA to 18A, my competitors’ technology. But with an earlier time to market, better technology maturity and much better cost.

C.C.: In fact, let me repeat again, our two nanometer technology without backside power is more advanced than both N3P and 18A. And what the semiconductor industries are most advanced technology when it is introduced in 2025. Does that answer your question, Gokul?

Gokul: Okay. That’s quite clear. Thank you very much. Could you also talk about, my second question is, could we talk a little bit about the AI related demand?

Gokul: You’ve seen a pretty strong demand on the data center side and you’ve talked about AI being about 6% of revenues this year, mostly on the data center side.

Gokul: Are we starting to see more engagement on AI demand on edge devices based on TSMC’s expectations? Is this going to be a big road driver in the next one to two years for TSMC’s leading edge AI devices, sorry, AI on edge devices? Do you think that that is a bigger driver for you?

Host: Okay. Thank you, Gokul. So Gokul’s second question is on AI related demand. He notes, of course, that we have talked last time also about our AI related demand outlook and particularly focused on what we call server AI processors or Gokul referring to data centers.

Host: So his question is really more about on the edge devices. Are we starting to see AI related demand for edge devices? Do we expect this to be a big growth driver in the next one to two years for our leading edge technologies as well?

C.C.: Well, the answer is also very simple. Yes.

C.C.: We do see some activities from customers who add AI capability in end devices such as a smartphone and PCs through neural engine and AI on PC or whatever it is. And we certainly hope that this one will add to the course, help TSMC (become) more strengthened under our AI business.

Gokul: So do you see that happen in the next one year or is it something that will happen in a more longer term horizon? Just wanted to understand when to think about the cadence of this growth.

C.C.: Okay. Let me answer briefly. It started right now. And we were expecting that the more and more customers will put that AI capability into the end devices, into their product.

Gokul: Got it. Yeah. Thank you very much.

Host: Okay. Thank you, Gokul. Operator, can we move on to the next participant, please?

Question 2. Charlie Zhang from Morgan Stanley

Operator: Next one to ask question, Charlie Zhang from Morgan Stanley.

Charlie: Hi, good afternoon, C.C., Mark, and Jeff. So my first question is about the cycle recovery. So much appreciation about your comments about in demand.

Charlie: So my question is about when do you expect there will be an uptick of the fab utilization, assuming demand is stabilizing and also inventory get back to the healthy level.

Charlie: Because it’s just very hard to believe your fab utilization outside of the leading edge will stay at only 70%, 80%. So first question is about, do you see that overall fab utilization to pick up at any time soon? Thank you.

Host: Okay. So Charlie, I think your first question is sort of, if I’m correct, around the cycle and in terms of how do we see the cycle and the recovery? When do we see the fab utilization picking up, which really is a function of, I guess, do we see the cycle bottoming out and when do we expect the recovery? Is that correct?

Charlie: Yes. Thank you.

C.C.: Well, Charlie, this is C.C. Wei again. Let me answer your question. As I said, we do observe some early signs of demand stabilization in PCs and smartphones and market. Those two segments are the biggest segment for TSMC’s business. We want to say that 2024 will be a very healthy growth. But right now, did we see the bottom? Very close, very close.

C.C.: We want to, I cannot give you a number, but it’s because it’s too early to call it a sharp rebound. But even with a macro environment remain uncertain, we weigh in customers inventory control in first half of 2024. Having said that, we already say that we are strong technology leadership and the broad customer base. 

C.C.: And those two are unique and specific to TSMC and never our customer to win business in market and TSMC continues to deliver healthy growth. And that’s why we can do better than overall industry. And that’s why we have  confidence that we will have a healthy growth next year.

Charlie: Okay, okay, thank you. Yeah, so just want to kind of verify because we do see some green shoots and some rush orders to wafer foundry sector. Thank you. But second question is also about AI.

Charlie: My question is about, over the past three months, do you see any upper revision of a forecast and amend from the GPU or does that cost and chip?

Charlie: And I know it’s a very, very recent, right? Just two days ago, the US put up some additional export control on the AI shipment to China. Do you think that that is going to have any kind of near-term or long-term impact to your AI semi-revenue growth assumption? Thanks.

Host: Okay, thank you, Charlie. So Charlie’s second question is related to AI. Two-parter, first, do we see over the last three months, I think his words were an upper revision in the demand from AI in the last three months.

Host: And then he wants to know, given the recent additional regulations announced, what would the impact to the AI demand be to TSMC, both for the short-term and the long-term.

C.C.: Charlie, the AI demand continues to grow stronger and stronger. So from TSMC point of view, now we have a capacity limitation to support them, to support the demand. We are working hard to increase the capacity to meet their demand. That’s for one thing.

C.C.: Now, US government put a new regulation and for some of the product cannot be shipped to mainland China. However, it is, you know, just for a couple of days, we are still in evaluation, we are still doing our assessment. But so far we can tell you that the impact to TSMC is limited and manageable. At least for the short-term, for the long-term we are still evaluating what is the consequence. At least for the short-term, for the long-term we are still evaluating what is the consequence.

Charlie: Got it. So actually there was one embedded question about the custom chip. So may I know your perspective about this custom chip long-term may kind of gain share or outgrow GPU products. So what’s your assumption for this custom chip in terms of the real contribution within AI or to TSMC? May I ask this follow-up?

C.C.: Okay, let me answer it. Where the customer develop the CPU, GPU, AI accelerator or AC for all the type of AI applications. The commonalities that they all require usage of a leading edge technology with stable yield delivery to support large or die size and a strong boundary design ecosystem.

C.C.: All of those are TSMC’s strengths. So we are able to address and capture a major portion of the market in terms of a semiconductor component in AI.

Charlie: Okay, okay, thank you.

Host: Okay, thank you, Charlie. Operator, can we move on to the next participant, please?

Question 3. Bruce Liu from Goldman Sachs

Operator: Next one to ask question, Bruce Liu from Goldman Sachs.

Bruce: Okay, thank you for taking my question. I think the question is try to ask about the outlook for next two years. I do recall that management mentioned about like 15 to 20% revenue kickers from 21 to 26. The smartphone business supposed to be in line or slightly below the corporate average in terms of the gross rate.

Bruce: But the smartphone business was down meaningfully in 2023. Do we expect to see a sharp rebound for the smartphone business to get back to the corporate average in terms of gross rate? Also that the management also mentioned that the backend business will grow in line with the corporate average. Will we see a much stronger growth from the backend business in the coming two years?

Host: Okay, so Bruce’s question really is looking at, we have indeed, you’re right, Bruce. We had said we will grow between 15 to 20% revenue kicker in US dollar terms from 2021 to 2026. So Bruce’s question wants to break down the components here to look at smartphone in particular, given that it has been a slower going market these last few years.

Host: How do we see the growth of the smartphone market the next one to two years in the context of this CAGR? And then also I think Bruce is also asking about the backend growth. We have said previously we’ll grow slightly faster than the corporate average.What is the current expectation now also for the next one to two years? Is that correct, Bruce?

Bruce: Yes, thank you. Right, Bruce, this is Wendell.

Wendell: Yeah, in the next two to three years, backing up to 21 to 26, we still expect that the smartphone as a whole will grow slightly slower than the corporate average. We still think that HPC will be the strongest one and will be the major growth contributor to our multi-year growth. This is your first question.

Wendell: As to the growth of backend business, we still expect that the backend business as a whole will grow slightly faster than the corporate average in the five-year time period.

Bruce: I see, thank you. So the next question is for the technology traders. I mean, for TSMC, for revenue contribution for 5nm, we start to see the minimal revenue contribution for 5nm starting at third quarter 2020, and 3nm is third quarter 2023. So the cadence, it looks like, is longer for 3nm versus 5nm and 7nm.

Bruce: What is the technology cadence moving forward? Are we able to see a minimal revenue contribution of 2nm in two-year time frame or three-year time frame? I think the technology migration cadence is an important indicator as well.

Host: Okay, thank you, Bruce. So Bruce’s second question, again, as he said, is around the technology cadence. He notes that 5nm started contributing revenue in 3Q20, but I think you’re saying 3nm revenue is a contribution in 3Q23. So the cadence is growing longer to kind of three years between five and three.

Host: So what will be the technology cadence in the future? And thus for 2nm, when should we expect meaningful revenue as well? Is that correct, Bruce?

Bruce: That’s correct.

Host: Okay, thank you.

C.C.: Okay, Bruce, let me answer the question. I think that we develop the technology to meet the customer’s demand. That’s a first priority to us. But then different customer may have a different product schedule consideration. And as time goes by, the technology complexity actually become more and more complicated.

C.C.: And our customer design their product and react to the market situation. So let me answer the question that’s in a very simple way. TSMC’s technology cadence remain constant to support our customers’ growth. But whether we got the same amount or same percentage of the revenue, it will depend on customers’ product schedule.

Bruce: But the follow-up question is that if the customer doesn’t really need the advanced technology in as fast as before, maybe we slow it down a bit, which will make a better returns, right?

Host: So Bruce’s follow-up is, so if the customers do not need the needed node as fast or as soon as before, then slow down the cadence. Does that mean we will see better returns?

C.C.: Well, we don’t slow down our technology development per se. We might slow down our capacity expansion according to customers’ demand. Did I answer your question, Bruce? That’s it. That’s what we are doing right now.

Bruce: I understand. Thank you.

Host: Okay. Thank you. Operator, can we move on to the next participant, please?

Question 4. Laura Chang from Citi

Operator: Yes, the next one, we have Laura Chang from Citi.

Laura : Hello. Hi, good afternoon. Thank you for taking my question. I think my question is also similar to what Bruce mentioned about the capacity expansion plan.

Laura: As we see that healthier inventory situation right now, and at the same time, the most advanced process now depends on customers’ demand. So I just want to get your feeling about the overall, the capex outlook or capacity outlook into the next two years.

Laura: Do you feel that will be better to resume the year-on-year capex next year or later? Considering there’s still quite a strong demand on N3, N2 ramp-up at the same time, the most advanced now seems you will see maybe two times more expensive on N2 versus N3. So my question is just about the future capacity expansion plan. Yeah, thank you.

Host: Okay, so Laura’s first question, again, is on the capacity expansion plan. She notes that, of course, the inventory, as we said, is becoming healthier, but also with N3 and N2. So she really wants to know what is the capex and capacity outlook plan for the next one to two years?

Wendell: Right, hi, Laura, it’s Wendell. C.C. just mentioned our capacity plan will really depends on the customer’s product plan. Now, in terms of capex, what we can see now is that we, in the past few years, have invested very heavily to capture the growth in the next few years.

Wendell: And as we begin to harvest those investments, we expect the increase of our capex to be leveling off in the next few years. That doesn’t mean the dollar amount is going to reduce, but the capital intensity is expected to decline in the next few years. That’s what we can see at this moment.

Laura: Thank you, Wendell. I recall that previously we mentioned about, like we target longer term to back to like mid-30s for our capex intensity. So is any possibility we see we achieve that target next year?

Wendell: It’s, first of all, it’s not a target. It’s a forecast based on the customer’s demand. And secondly, it’s like three or five years out. It’s not the immediate next year.

Laura: Okay, thank you. Very clear. And also my second question is about N2 progress. I appreciate C.C.’s previous sharing on the progress and also the backside power timeline. I’m just wondering that what would be the most challenging part if we migrate to like a backside power? And also since the transistor density will become a totally different analogy. So I’m just wondering that the client’s migration or intention to adapt the most advanced node into our next two, three years. And what will be the most challenging in technology perspective? Thank you.

Host: Okay, so Laura, let me make sure I got this right. So your second question is about N2 and you want to know with the adoption of backside power rail, what is the most challenging part from a technology perspective? And then how do we see the customer adoption?

Laura: Yes, correct. Thank you.

C.C.: Laura, I answer that. You know, as a technology moving to more and more advanced node, the challenge is always there. Technology complexity increase dramatically, but we can do it. No doubt about it. And we are still remain the technology leadership in this industry.

C.C.: If you ask me, what is the most challenging part? I will say its cost. I mean, that’s how you look at it today. Inflation, everything and the tool have become more and more expensive.

C.C.: Although we can do it on time to meet customers’ requirement, our challenge right now, actually, I would say, number one, cost. I want to reduce the cost so more customer can afford it.

C.C.: But even with that, actually, we have a lot of customer interested and engaged with the TSMC today. Actually, it’s probably higher than the N3 at a similar stage. Okay. Did I answer your question?

Laura: Yes, very clear. Thank you very much, CC.

Host: Okay, thank you, Laura. Operator, can we move on to the next participant, please?

Question 5. Brett Simpson from LFA Research

Operator: Right now, we have Brett Simpson from LFA Research.

Brett: Thanks very much. I had a question for CC regarding the hyperscalers.

Brett: Major US hyperscalers are hiring a lot of chip designers at the moment and looking to make their own AI silicon, going direct to TSMC, much like Apple has done over the years. Are you, is TSMC generally supportive of this trend or not?

Brett: And can you give us your perspective as to whether hyperscalers have the in-house IP and skills to cut out the ASIC suppliers or not and go direct to TSMC? Thank you.

Host: Okay, Brett, thank you. So Brett’s first question is looking at his observation, US hyperscale companies are hiring a lot of people to do AI custom chips silicon and working directly or coming directly to work with TSMC. So his question is, is TSMC support such efforts? And how do we see such type of customers, I guess? Is that your question, Brett?

Brett: Yeah, and to generally just share your perspective on how you see this part of the business developing at TSMC. Thank you.

C.C.: Hi, Brett. Okay, those are hyperscalers. I don’t comment on the specific customer, but all we know or our fundamental rule is where the customer develop the CPU, GPU, AI accelerator or ASIC for their own application or for any purpose in the AI area, we will support them, actually.

C.C.: And because of our technology leadership and our good manufacturing, so we are able to address and capture a major portion of the market. And so you are asking whether we support it or not, we support every customer all over the world.

Brett: Okay, thanks, that’s very clear. And maybe just as a follow up, I wanted to ask about some of the areas in the quarter that were weaker than expected, namely seven nanometer, I think it’s half year on year, and also automotive that saw a decline sequentially, a meaningful decline sequentially.

Brett: Can you just help us understand how you build back up seven nanometer, what led to the weakness and what’s happening in automotive and how do you assess prospects for automotive over the next six, nine months or so? Thank you.

Host: Okay, so two parts to Brett’s question, maybe I’ll start with the second, but anyway, looking at automotive demand and the weakness in the past quarter, how do we see the automotive demand?

Host: And then also his question is about seven nanometer, also year on year sharp revenue decrease. So how do we see the outlook for seven nanometer as well?

C.C.: Well, let me answer the question on automotive demand. In fact, in the past three years, automotive demand has very strong and we deliver whatever they asked.

C.C.: And today I think it’s automotive demand already entered the inventory adjustment mode in the second half of 2023. However, we still expect the automotive demand-wide increase again in 2024, because the more and more EV, more and more functionality being added to automotive and that’s what we saw.

C.C.: Now talking about the N7, the seven nanometers technology, why we have such a low utilization or the revenue decrease? It’s a goal beyond our initial, our original plan because we expect the N7 to be very fully utilized even now, but it is not.

C.C.: Let me answer the question because we suddenly have in 10 years, the smartphone demand dropped dramatically from about 1.4 billion units to about 1.1 billion now.

C.C.: So that exactly in this timeframe, the N7 utilization has been impacted and followed by one major customer who delays their product introduction. And so that’s why we have a low utilization.

C.C.: But seeing that, we are confident to backfill our seven, six nanometer capacity with additional wave of specialty demand from a consumer, RF, connectivity and other applications and will return to a healthy level of utilization over the next several years.

C.C.: This is very similar to a situation that we have a 28 nanometer back in 2018 and 2019 timeframe. Okay. In the beginning it was underutilized for a period of time and we work hard with our customer and then for developing some specialty technologies and then now we have to expand 28 nanometers of specialty capacity. That’s the same kind of a story. Brett, did I answer your question?

Brett: Thank you. Okay, thanks a lot. Yep. Thanks very much.

Host: All right, Brett. Thank you. Operator, can we move on to the next participant please?

*The entire content of the Transcript was assisted by AI tools and edited by human correction for the purpose of industry information dissemination and reference only. It should not be used as an investment guideline, and official announcements should be the primary source for detailed information.

(Photo Credit: TSMC)