“We have no intention to leave food on the table for others, and we want to capture every piece of business possible.”
It sums up management's stance of expanding capacity everywhere to meet demand rather than rationing it.
TSMC (TSM) · Semiconductors
Results filed 15 Oct 2026 · Quarter ended 30 Sep 2026 · Call 67 min · C.C. Wei (CEO), Wendell Huang (CFO)
Read the transcriptPress release ↗Investor relations ↗
Summaries and transcripts are machine-generated and can contain errors.
TSMC is adding three more 3-nanometer fabs and preparing to more than double 2-nanometer capacity as AI demand continues to outstrip supply.
TSMC's quarter was driven by AI and the move to 2-nanometer. High-performance computing now makes up 68% of revenue and smartphones 21%, and automotive grew fastest quarter on quarter. 2-nanometer has been in volume production since the fourth quarter of 2025 and is ramping at Hsinchu and Kaohsiung. Management says demand for leading-edge chips is still well above its ability to supply. It points to agentic AI CPUs, networking chips and AI accelerators as sources of that demand.
Management is building at what it calls an unprecedented pace, with nearly 25 fabs in progress worldwide, three more 3-nanometer fabs, and a possible Texas site beyond Arizona. The tension is cost and demand risk. Overseas fabs dilute margins, tool prices are rising, and management is being prudent on 2027 because consumer and price-sensitive markets face rising component prices, even as AI demand stays strong.
up 50.9% from a year ago
up 68.7% from a year ago
up 68.7% from a year ago
up 6.9 ppts from a year ago
up 8.1 ppts from a year ago
up 75.0% from a year ago
up from 426.83 a year ago
up from 139.38 a year ago
| Segment or metric | Value | Change |
|---|---|---|
| HPC (share of revenue) | 68% | up 22% quarter over quarter |
| Smartphone (share of revenue) | 21% | up 10% quarter over quarter |
| IoT (share of revenue) | 4% | up 5% quarter over quarter |
| Automotive (share of revenue) | 4% | up 23% quarter over quarter |
| DCE (share of revenue) | 1% | down 12% quarter over quarter |
| 2-nanometer (share of wafer revenue) | 8% | |
| 3-nanometer (share of wafer revenue) | 33% | |
| Advanced technologies (share of wafer revenue) | 79% |
Copied from the press release ↗ and other documents in the SEC filing ↗. A figure is published only if its sentence is in those documents.
| Metric | Period | Outlook |
|---|---|---|
| Revenue | fourth quarter 2026 | between US$52.3 billion and US$54.3 billion |
| Gross profit margin | fourth quarter 2026 | between 65% and 67% |
| Operating profit margin | fourth quarter 2026 | between 56.5% and 58.5% |
| Full-year revenue growth in US dollar terms | 2026 | around 44% |
| Capital budget | 2026 | between 60 and 64 billion U .S. dollars |
| Gross margin dilution from overseas fabs | the next few years | 3-4 % |
N2 entered high-volume manufacturing in the fourth quarter of 2025 and is ramping in multiple phases at Hsinchu and Kaohsiung. Management expects N2 family annual capacity to more than double in 2027, backed by smartphone and HPC AI demand.
Historically TSMC does not add capacity to a node once it hits its target, but strong multi-year demand changed that. New 3-nanometer volume production is planned in Tainan in the first half of 2027 and 2028, in Arizona in the third quarter of 2027, and in Japan in 2028.
TSMC sent teams to evaluate Texas and is in discussion with a customer building a fab called TerraFab about using actual capacity. Management has reached no conclusion, but says any such fab must be run by TSMC.
TSMC is in volume production on 5X reticle-size packaging and is working on 9X and 14X, and management says it has a back-end shortage. It has a deal with GlobalFoundries on interposers so both companies can support TSMC customers.
A14 risk production is on track for 2027 with volume production in 2028, and management expects it to be a bigger node than N2. TSMC plans to adopt High NA EUV in 2030 once it is mature and cost-reasonable, with ASML working on a 12-inch photomask by 2031.
Each quote matches the transcript word for word.
“We have no intention to leave food on the table for others, and we want to capture every piece of business possible.”
It sums up management's stance of expanding capacity everywhere to meet demand rather than rationing it.
“We expect our N2 family technologies' annual capacity to more than double in 2027, supported by strong demand from both smartphone and HPC AI applications.”
It sizes the 2nm capacity build and shows demand comes from both phones and AI.
“However, based on our assessment, we are executing well on our global capacity plan to add three new fabs to support the robust multi-year pipeline or demand for 3 nanometer technologies.”
It marks a departure from TSMC's past practice of not adding capacity to a mature node.
“In addition, the emergence of authentic AI CPUs and the continued demand for networking chips are driving further silicon demand.”
It shows AI demand broadening beyond accelerators to CPUs and networking.
“The principle of TSMC is we do not allow our customer to sublet or auction of the capacity.”
It rules out a secondary market for wafer allocation, shaping how customers can use their capacity.
“I don't believe that anybody can do that today except the TSMC. And the yield is excellent. So our customers are very satisfied with the result.”
It is management's competitive claim on large-reticle advanced packaging.
C.C. Wei says TSMC's capacity planning weighs all of those factors and still finds demand higher than supply, so it is working to narrow the gap, with demand extremely robust and no sign of slowdown. — C.C. Wei
C.C. Wei says High NA is a good tool planned for 2030, and that there is no ceiling for TSMC's profitability though a dramatic increase will not happen overnight. — C.C. Wei
C.C. Wei says demand keeps rising and the supply gap is still big so he cannot say whether it narrows, and Jeff Su says it is too early to give a capacity growth percentage. — C.C. Wei, Jeff Su
C.C. Wei says TSMC sent teams to evaluate Texas and is in discussion with a customer building TerraFab, with no conclusion yet, but any fab must be run by TSMC. — C.C. Wei
C.C. Wei says it is not a new strategy, TSMC does not allow customers to sublet or auction capacity, and only rare mask licensing collaborations occur in very small amounts. — C.C. Wei
C.C. Wei says TSMC will adopt it when it is mature and cost-reasonable, with ASML working on a 12-inch photomask by 2031, and he is confident TSMC remains the best provider of advanced technology. — C.C. Wei
C.C. Wei says TSMC treats every customer the same while considering internal capacity, will not comment on a specific customer, and expects A14 to be a bigger node than N2. — C.C. Wei
C.C. Wei says TSMC is in production at 5X reticle size, is working on 9X and 14X, and would make customers' smooth product introduction its number one priority. — C.C. Wei
Wendell Huang says it is too early to give a number, but 2026 to 2028 capex will be significantly higher than the past three years, tool inflation is real, and revenue growth should still outpace capex growth. — Wendell Huang
Wendell Huang says TSMC is accelerating everywhere including the US, and the 3 to 4 percentage point dilution applies to the later part of the five-year period. — Wendell Huang
C.C. Wei says demand from agentic AI CPUs is a rising trend, but overall AI demand is extremely robust and far above what TSMC can supply. — C.C. Wei
C.C. Wei says strong smartphone and AI HPC demand arriving together is to TSMC's advantage, and the learning curve is being handled well. — C.C. Wei
C.C. Wei says power supply is a consideration TSMC checks in the US with good results so far, and it has not changed customers' choice of backside power, which depends on the product. — C.C. Wei
C.C. Wei says CPUs are not in the narrow AI accelerator definition because TSMC cannot identify which are used for AI, though they are part of the business. — C.C. Wei
C.C. Wei says TSMC's pricing is strategic and depends on market segment, location and technology, and US-made capacity will be priced differently from Taiwan. — C.C. Wei
C.C. Wei says the deal is part of its back-end supply strategy, since TSMC has a back-end shortage and wants customers fully supported. — C.C. Wei
Yes. Management says demand is extremely robust, has shown no sign of slowdown, and remains far above what TSMC can supply today. It is adding capacity as fast as it can.
TSMC expects N2 family annual capacity to more than double in 2027. It is also adding three new 3-nanometer fabs: Tainan in the first half of 2027 and a further Tainan fab in the first half of 2028, Arizona in the third quarter of 2027, and a second Japan fab in 2028.
TSMC sent teams to evaluate Texas and is in discussion with a customer building TerraFab about using actual capacity. There is no conclusion yet, and any fab must be run by TSMC.
Management says its packaging is very competitive and that it is in production at 5X reticle size, with 9X and 14X in development. It has a shortage in back-end capacity and signed a deal with GlobalFoundries on interposers to help address it.
No. Management keeps a narrow definition of AI accelerator demand, previously described as growing at a mid to high 50s percentage CAGR from 2024 to 2029, because it cannot identify which CPUs are used for AI. It still sees agentic AI CPU demand rising.
SEC filing
Call recording on YouTube ↗. Official replay ↗. Timestamps jump to the moment.
Good afternoon everyone and welcome to TSMC's third quarter 2026 earnings conference call. This is Jeff Su, TSMC's Director of Investor Relations and your host for today. TSMC is hosting our earnings conference call via live audio webcast through the company's website at www .tsmc .com where you can also download the earnings release materials. If you're joining Joining us through the conference call, your dial-in lines are in listen-only mode. The format for today's event will be as follows.
First, TSMC's Senior Vice President, third quarter, 2026, followed by our guidance for the fourth quarter, 2026. Then Mr. Huang and TSMC's Chairman and CEO, Dr. C .C. Wei, will jointly provide the company's key messages. that will be followed. Then we will open the line for the Q &A session. As usual, I would like to remind everybody that today's discussions may contain forward-looking statements that are subject to significant risk and uncertainties, which could cause the actual results to differ materially from those contained in the forward-looking statements.
Please refer to the safe harbor notice that appears in our press release. And now I would like to turn the call over to TSMC CFO, Mr. Wendell Huang, for the summary of operations and the current quarter guidance.
Thank you, Jeff. Good afternoon, everyone. Thank you for joining us today. My presentation will start with financial highlights for the third quarter of 2026. After that, I will provide the guidance for the fourth quarter of 2026. Third quarter revenue increased 17.6 % sequentially in NT, supported by strong demand for our leading-edge process technologies.
In U .S. dollar term, revenue increased 16.2 % sequentially to $46.7 billion, above the high end of our third quarter guidance. Gross margin decreased by 1.3 percentage points sequentially to 66.4%, primarily due to margin dilution from 2nm ramp. Total operating expenses accounted for 7.8 % of net revenue. Operating margin was 58.7%. Overall, our third quarter EPS reached 29.43 NT, and ROE was 45.5%. Now let's move on to revenue by technology.
Two nanometer process technology contributed 8 % of wafer revenue in the third quarter, while 3 nanometer, 5 nanometer, and 7 nanometer accounted for 33% 28% and 10% respectively advanced technologies defined as seven nanometer and below accounted for 79% of wafer revenue moving on to revenue contribution by platform hpc increased 22% quarter over quarter to account for 68% of our third quarter revenue Smartphone increased 10 % to account for 21%. IoT increased 5 % to account for 4%.
Automotive increased 23 % to account for 4%. While DCE declined 12 % to account for 1%. Moving on to the balance sheet, we ended the third quarter with cash and marketable securities of $4 1 trillion NT, or 126 billion U .S. dollars. On the liability side, current liabilities increased by 110 billion NT quarter over quarter, mainly due to the increase of 46 billion in current portions of bond payable and bank loans and 26 billion in dividend payable.
In terms of financial ratios, accounts receivable turnover days decreased by one day to 28 days. Inventory days decreased by five days to 82 days, primarily due to the increase in wafer shipment. Regarding cash flow in CAPEX, during the third quarter, we generated about 920 billion NT dollars in cash from operations, spent $443 billion in CAPEX and distributed $156 billion for fourth quarter 2025 cash dividends. Overall, our cash balance increased $488 billion NT to $3.6 1 trillion at the end of the quarter.
In U .S. dollar terms, our third quarter capital expenditures totaled $13.8 billion. I have finished my financial summary. Now, let's turn to our current quarter guidance. Based on the current business outlook, we expect our fourth quarter revenue to be between 52.3 and 54.3 billion U .S. dollars, which represents a 14 % sequential increase or a 58 % year-over-year increase at the midpoint. Based on the exchange rate assumption of one US dollar to 31.8 NT, gross margin is expected to be between 65 and 67 percent.
Operating margin between 56.5 and 58.5 percent. This concludes my financial presentation. Now, let me turn to our key messages. I will start by talking about our third quarter 26 and fourth quarter of 26 profitability. Compared to second quarter, our third quarter gross margin decreased by 130 basis points sequentially to 66.4%, primarily due to the margin dilution from the initial ramp-up of our 2nm technology. We have just guided our fourth quarter gross margin to decrease by 40 basis points to 66 % at the midpoint, primarily due to the continued margin dilution from the steep ramp-up of our 2nm technology, as well as dilution from overseas fabs.
These impacts are expected to be partially offset by cost-improvement efforts, including productivity gains and across-no-capacity optimization. While the cost of overseas fabs remains higher, thanks to the company's overall larger scale and cost improvement efforts, we now expect the gross margin dilution from the ramp-up of our overseas fabs to be closer to 2 % more effort to support the strong structural demand from our customers. More fabs will be ramping up in the US.
Thus, we expect the dilution from our overseas fabs to widen to 3-4 % in the next few years. We will leverage our increasing size in Arizona and work on our operations to improve the cost structure. We will also continue to work closely with our customers and suppliers to manage the impact. Overall, with our fundamental competitive advantages of manufacturing technology leadership and large-scale production base, we are confident that TSMC continues to be the most cost-efficient manufacturer in every region that we operate.
Next, let me make some comments on our 2026 capital budget. At TSMC, a higher level of capital expenditures is always correlated with higher growth opportunities in the following years. Given the continued robust multi-year demand from our customers, including the emerging agentic AI market, we reiterate our full year 2026 capital budget to be between 60 and 64 billion U .S. dollars. We continue to expect our CAPEX dollars investment from 2026 to 2028 to be even more significantly higher than the past three years, as we continue to invest heavily to support our customers' growth.
Additionally, while the tool supply is tight, we always collaborate closely with our tool suppliers well in advance to prepare the capacity, just like our customers collaborate with us well in advance to plan our capacity. Even as we invest for the future growth with this level of CAPEX spending in the next few years, we remain committed to delivering profitable growth to our shareholders. We also remain committed to a sustainable and steadily increasing cash dividend per share on both an annual and quarterly basis.
Now let me turn the microphone over to CC.
Thank you, Wendell. Good afternoon, everyone. First, let me start with our near-term demand outlook. We concluded our third quarter with a revenue of US dollar 46.7 billion, higher than our guidance in US dollar terms, driven by strong demand for our leading-edge process technologies. Moving into first quarter 2026, we expect our business to be supported by continued strong demand for our leading-edge process technologies, including the steep ramp-up of our 2nm technology.
Looking ahead, we continue to observe consumer and price sensitive end market segments are being challenged due to the impact of rising component prices and macroeconomic uncertainties. As such, we are being prudent in our business planning going into 2027 while focusing on the fundamentals of our business to further strengthen our competitive position. Having said that, demand for our leading edge technologies continues to be extremely robust, supported by our strong technology differentiation and broad customer base. Therefore, we now expect our full year 2026 revenue growth to be around 44 % year-over-year in US dollar term.
Next, let me talk about AI-related demand outlook. The AI market continues to be very strong. Recent developments are very positive to AI's long-term demand outlook. We observe increasing AI demand fueled by the rising adoption of authentic AI as we transition from chatbot to AI assistance. In addition, we are observing great adoption of open source AI models. This will drive greater efficiency and help lower the barrier to future AI usage and result in more and more demand for leading-edge silicon.
Thus, AI-related demand is being supported by strong underlying demand for AI accelerators. In addition, the emergence of authentic AI CPUs and the continued demand for networking chips are driving further silicon demand. Therefore, semiconductors will remain essential in the AI mega-trend. The value of semiconductors in the global supply chain is set to continue to increase in our view. This will create more value opportunities for our customers to capture and more value opportunities for TSMC as well.
No matter which approach is taken, you will require use of the most advanced semiconductor process technology. With our technology leadership, manufacturing excellence, global capacity expansion, and customer trust, we are well positioned to capture the full multi-year structural AI-related demand. Therefore, all the AI players are working with TSMC. Next, let me talk about TSMC's capacity expansion strategy. As a frontendry, our first responsibility is to provide our customers with the best technologies and effective capacity support to unleash their innovations.
To address the structure increase in the long-term semiconductor market demand profile, TSMC collaborated closely with our customers, and our customers are customers to plan for our capacity. In addition, we are now going one step further and sending our own teams of people to check on the data center build-up process, progress, and power supply situation in the U .S. The feedback is so far so good, and the CSPs have been well prepared. Internally, TSMC employed a disciplined capacity planning system to assess the market demand from both a top-down and bottom-up approach.
This is a continuous and ongoing process. Given the fundamental complexity of leading-edge technologies and the designing and lead time involved, we already have a very good idea of customers' multi-year product roadmap and production plans. Thus, we continue to hold strong conviction in the AI mega trend. We are moving full speed ahead to increase our capacity to meet our assets demand and stepping up our capex investment to support our customers' future growth. TSMC is expanding at an unprecedented pace with near 25 worldwide in progress.
We are building 13 leading-edge and advanced packaging fabs in Taiwan over the next several years, and another five to six fabs overseas. Our new node, N2, has already entered high-volume manufacturing in the fourth quarter of 2025 with good year. N2 is ramping successfully in multiple phases at both Hsinchu and Kaohsiung side. We expect our N2 family technologies' annual capacity to more than double in 2027, supported by strong demand from both smartphone and HPC AI applications.
We are also leveraging our manufacturing excellence to generate more N2 wafer output, further increasing our capacity support to customers. In addition, we have already begun preparing the capacity for our A14 technology. Risk production is well on track for 2027, and volume production will start in 2028. Thus, TSMC's semiconductor technologies and manufacturing We are continuing to play a pivotal role in supporting the global semiconductor industry while unleashing our customers' innovation. Now, let me talk about TSMC's global N3 capacity expansion update.
As we said before, historically, we do not add additional capacity to a loan once it reaches its targeted capacity. However, based on our assessment, we are executing well on our global capacity plan to add three new fabs to support the robust multi-year pipeline or demand for 3 nanometer technologies. In Taiwan, volume production for our new 3 nanometer fab in Tainan Science Park is well We are on track to begin the first half of 2027. In addition, driven by the strong demand from our customer, we have decided to add one addition of three nanometer fives in Tainan.
Volume production is scheduled for the first half of 2028. In Arizona, we have completed the construction of our second fives and are in the process of tour moving and installation, volume production will begin in third quarter 27, slightly ahead of our previous schedule. In Japan, we are processing well to our plan to utilize 3 nanometers in our second fab, and volume production is scheduled in 2028. In addition to all the new FABs, we continue to convert 5nm tools to support 3nm capacity in Taiwan.
We are also leveraging our manufacturing excellence to create productivity across our FAB in all locations. To generate more wafer output, we are also focusing on capacity optimization across nodes, which include flexible capacity support among N7, N5 and N3 nodes. Thus, we are using multiple levers to do everything we can, whenever we can, as fast as we can to maximize the support to our customers. We have no intention to leave food on the table for others, and we want to capture every piece of business possible.
This concludes our key message, and thank you for your attention.
Thank you, CC. This concludes our prepared statements. Before we begin the question and answer session, I would like to remind everybody to please limit your questions to two at a time to allow as many participants as possible the opportunity to ask their questions. should you should you wish to raise your question in chinese i will translate it to english before our management answers your question for those of you on the call if you would like to ask a question please press the please press the star then one on your telephone keypad now if at any time you'd like to remove yourself from the questioning queue please press star two
now let's begin the q a session. Operator, can we proceed with the first caller on the line, please?
Yes. First one, Laura. James, David. Go ahead, please.
Thank you very much. Good afternoon. And I appreciate Cici for sharing with the TSMC capacity expansion strategy. So may I follow up more on that? Given the increasing aggressive AI semiconductor demand. How does TSMC now see the AI revenue growth? And how does TSMC reconcile customers' capacity requirements with some of the potential concerns on the data center power availability and also maybe funding recently and other critical components such as HBM, APS, advanced packaging, and also in the networking side, we see the tightness of laser as well.
So if these constraints create a mismatch between semiconductor supply and also actual AI development, how would CSMC adjust your capacity, minimize the potential overbuilding in the supply chain? That's my first question.
OK, Laura. Laura has a very involved first question. I think she's asking, how do we see the AI demand outlook? The AI demand, as Cece said, seems to be very strong, but she's worried or concerned. How do we reconcile the strong demand of chips, AI chips, versus risks related to whether it be power supply?
I think also like different bottlenecks within the supply chain, et cetera. How do we make our assessment of what is the appropriate capex and capacity we need to prepare? Is that correct, Laura? Yes, thank you.
That's a long, long question. Laura, let me share with you. We have a very sophisticated capacity plan process. And we also have a very disciplined on our consideration. And let me share with you, we put every possibility we can, talking about the CSPs, CAPEX, data centers, build out financial, power supply, and also other component of supply, et cetera, into our consideration.
And so we make our own judgment. And then based on our assessment, the demand, we still think the demand is higher than the supply and we are working very hard to narrow the gap. Did I answer your question, Laura?
Sure, thank you. Thank you very much. So, so far from our comprehensive starting supply chain, we are still quite positive on the demand in the longer term run. So in that case, can you also give us more color about that?
Any update on your AI CAGR growth? I recall previously we were talking about like a 50%.
We, sorry, sorry. Yeah, so part of Laura's question is then based on our assessment, we had previously guided our AI CAGR, AI accelerator CAGR, which is a narrow definition to go mid to high 50s percentage from 2024 to 2029. So Laura wants to know, do we have any update to that number today?
Laura, all I can say right now is the demand is very extremely robust.
Okay. And we did not see any sign of slowdown. Whether it's a high 50 KG, it doesn't matter. it probably will be higher. Okay.
Yes. Thank you very much. Yeah. And my second question is also about like the, given the very strong demand, the CapEx cycles, we see that definitely this round will be much higher than before. And TSMC also announced that earlier, you will adapt a high NAEUV around 2030. So should we think about that?
that will be further improved productivity, that will be more meaningful for further accelerating revenue growth in the longer term. And also given that the capex intensity are also getting much higher, so how should we think about that, the pricing power and also the long-term growth margin, the capital efficiency return?
Okay. That's actually two questions, but okay, we'll just very quickly. So Laura's part of it is on with the strong demand on our CAPEX cycle, how do we view the adoption of high NA, which we said we will, you know, in 2030.
And then her question is on profitability, long-term outlook. Yeah, thank you.
Oh, okay. Let me answer that question. Yes, for the long-term productivity, we continue to improve it. And high-end definitely is a very good tool. So far, I think our plan is to use it in 2030. Again, our capx will be much higher than today. And so you're talking about the gross margin or the profitability can be increased or not. Let me say that.
There's no ceiling for TSMC's profitability. But a dramatic increase will not happen overnight either. So our philosophy or our strategy always are working with our customer. and we want to make them be successful and then we can get our value.
Thank you. Very helpful. Appreciate it. And congrats for the great result. Thank you. Thank you.
Operator, can we move on to the next participant, please?
Next one, Goku Harihalan. JP Morgan. Go ahead, please.
Hi. Good afternoon, JT, Wendell, and Jeff.
My first question, again, on capacity addition, looks like the short-term. Just wanted to hear, Cici, your view on this. And I think it's been quite some time since we've seen TSMC add capacity at more than 10 % every year. I think if you look at our annual report, I think it's been like about 10 years that you've added capacity at more than 10%. Given this accelerated build out, can we start getting wafer capacity growth closer to 10 % or even higher than 10 % in the next couple of years?
Any comment on what is your planned with a capacity growth in the next couple of years? That's my first question.
Okay, so Gokul's first question is regarding the capacity addition and the demand. So his question first is, you know, the tightness in the wafer supply, do we think that will be even more or greater in 2027 than 2026? And then along with that, how do we see our overall capacity addition? Although you say 10%, I think every year we disclose in the annual report, it's been about mid to high single digit and sometimes even lower.
But anyways, regardless, those are the two questions. So first on just the demand outlook and then the supply, will next year be even more tight than this year?
I just mentioned the demand is extremely robust and getting higher and higher. Of course, we have our own judgment and we look at every angle. As I said, we talk to customers. customers are customers, and then we do our own one more step to check the data centers are built out.
All the results we put into our judgment and our assessment, the result is the demand is very, very high. And so the gap between demand and supply are still quite big and that's why we put all our strength to increase the capacity as I just mentioned. Whether it's bigger or the gap becomes bigger or the gap becomes narrow, we are working very hard. I cannot answer this question right now because my customers continue to increase their demand.
Okay. Goku, I think it's a bit too early to talk about capital capacity increase percentage going forward. But let me just remind you that the increase every year, although the number is, even though the number may be small, but the increase in advanced packaging, sorry, advanced notes are really the main driver for the capacity increase.
Okay. Fair enough. Okay. Secondly, I think you alluded to more aggressive over these expansion plans in the U .S. especially. And I think recently there's been a lot of discussion with potentially a new fab cluster in Texas, as well as I think one of your customers also talked about discussions with TSMC as well.
So Sisi, could you talk a little bit about your overseas plan, like the more accelerated kind of capacity build out and also any engagement with SpaceX or Tesla, given your customer did call you out last week in terms of some engagements here. Like, what are the broad contours of this kind of engagement likely to be? Is it similar to what you've done with some of the customers in Japan and ESMC as well?
Okay. Thank you, Gokul.
So Gokul's second question, he notes that CC said, we're expanding at an unprecedented place, doing everything we can as fast as we can to increase the capacity. So his question is particularly to the U .S., what is the status or update? there's been reports that we are looking to expand in Texas. There's been reports that one of our customers we're engaging in discussions. He wants to know what kind of framework or parameters of any cooperation may be or any comment on either of those two things.
Okay, Goku. We work very hard to expand our capacity in the U .S. This is necessary because of all our customers, most of the large customers in the U .S., they request TSMC to increase the capacity. And also U .S. government are very helpful. So that's a place we want to expand. So recently, we did expand our capacities consideration to see outside Arizona whether we can get another chance to expand our capacity. And Texas is the one.
So we send the teams to Texas to do all the evaluation. And we also talked to one of our customers who tried to build a fab, TerraFab, let me be more specific. And we discussed with them, see whether we can be utilizing the actual capacity also. So we're still in discussion. If we have any solid conclusion we'll share with you. But today, and that's it. Yes. From Arizona to Texas, yes.
Okay. But all of these are going to be DSM-C run FAPs, right?
I think you're not becoming part of the TerraFAP project or anything like that. Can you just help us understand what are the principles that you're using here? Is there a similar principle that you use for JSM or ESMC?
Okay. So Gokul's question is not, you know, in particular to the potential cooperation with this customer. Would it be TSMC-owned? What's the principles that we have, if anything?
We are still in discussion. Let me say that. It's not, you know, either way it would be possible.
But the fab got to be run by TSMC.
Operator, can we move on to the next participant, please?
Yes, next one, Kreej Sankar, TD Cohen. Go ahead, please.
Yeah, thanks for taking my question. I told them, first one, some customers talk about a secondary vapor market where one customer passes their vapor allocation to another. Is this like a very small portion of the business? Is it customer specific? Or do you see that as an emerging trend where customers who demand weak pass some of their allocation to stronger customers?
I'm gonna add a quick follow-up.
All right, Krish, thank you. So Krish's first question is asking, some of our customers are talking about a reseller, a secondary wafer market where they can pass their allocation on. So is this a very small portion of our business or is this more of a trend for us to do? And is this something that we allow our customers to do?
Well, let me answer that. To say whether this is a new strategy, new methodology, the answer is no.
The principle of TSMC is we do not allow our customer to sublet or auction of the capacity. But in certain cases, customers can collaborate to a mask licensing agreement, but this is not our common practice. And the amount is very, very small. Chris, did I answer your question?
Yes, thank you very much for that. And then a quick follow up. You did touch upon it, the INA agreement with ASML from 2030 onwards. I'm just curious to know historically TSMC was more ahead of the curve with the EUV adoption.
It looks like by the time you adopt high NA EUV, it will be 2030, 2031, but some of your competitors are already been using it by then. So do you see that as a disadvantage or you're not worried about competition getting ahead with high NA EUV?
Okay. So Krish's second question is looking at high NA. He notes some of our competitors are both using already adopted or will adopt high NA earlier than TSMC. So his question, does this put us at a competitive disadvantage or risk compared to our competitors in terms of technology leadership?
Chris, let me answer that. But we always evaluate a new tool. And if we want to put it into the mass production, it got to be the best technology with a very reasonable cost. And most important is the best maturity. So we say in 2030 we'll start to use it because we also think at the time high NA will be mature and we're also working very closely with ASM to establish a 12 inch photo mask by 2031.
And at that time, we think that high-end cost will be reasonable and for TSMC to make it into the mass production. And this is a very good tool. Let me emphasize that. And in the smaller geometry, it definitely has some advantage. Whether that our technology, because we are using it later than our competitors, so disadvantage to TSMC? No. Because we have confidence that we are still the best company to offer the most advanced technology to our customers.
Chris, does that answer your question? Thank you very much.
Operator, can we move on to the next participant, please?
Yes. Charlie Chen, Morgan Stanley, go ahead, please.
Hi, gentlemen. Thanks for taking my question. And also, I'm here from AI, semi-demand, amid some recent market correction. So my first question is really about your image capacity extension strategy. As far as I know, one of your IDM customers is expanding capacity aggressively. So there are some concerns about this customer may in source more and also to you.
So first of all, I'm not sure how you're going to deal with this. a further business opportunity. And secondly, I remember last time CCU mentioned about A14 could be a bigger node than N2. May I know whether that assume the big IDM company will outsource more CPU, like server CPU at A14? Thank you.
Okay. So Charlie's first question is around IDM competitor. He wants to know very specifically if we see this competitor moving to in-source more.
So his question is then our prior comment of A14 being a larger, bigger node than N2 has even been does this include the server CPU from this IDM competitor? Which?
Well, Charlie, we treat every customer So whether it's IDM or purely fabulous, we treat them the same. But of course, we take into consideration of the internal capacity. So we calculate our support, we calculate our capacity build up according to our own judgment. Actually, in our A14, we have seen more and more customers actually higher than the 2nm activities so far.
And we expect that A14 will be a bigger node than our current N2 technologies.
I see. So it doesn't need to consider this idea. We can still have to be a bigger note than N2. Is that the right understanding?
Well, again, Charlie wants to know very specifically this customer's outsourcing plan, which you probably may need to ask them. but
yeah well charlie as i said um to be frank with you i don't come in on a specific customer but let me guarantee you that we take everything into consideration okay okay okay do you have a second question thank you
oh yes i do uh thanks jess um so um the signature question is a kind of follow-up about your
advanced packaging uh from what we're seeing that the residual size is getting much bigger. Sissy, you said last time or even two quarters ago about your response, you know, SOW 3.5D. So lots of initiative is going on. So my question is that do you think those are kind of a corpus SOW 3.5D can really catch customers are very rapid developments of their larger record size chip like 14 reticle. And secondly, if some customer they initially choose to use Intel's EMT and for some reason cannot get to sufficient demand, are you okay to bail out or support customers for the co-aspiration or capacity?
Thank you.
All right. So Charlie's second question is focus on the advanced packaging. I guess, Charlie, if I understand your question, summarize it, it is, of course, customers are moving to larger and larger reticle size, 14X. So he wants to know TSMC's roadmap and solutions and how competitive we can be against, I think he mentioned specifically, our competitors, EMIB-T. So how do we view our solution? Is it as competitive as our competitor? That's the first part of this question.
All right, Charlie. Actually, I still think that TSMC's advance assembly or packaging technology are very, very competitive. But the most important thing is we're using that one to help us, help our customers' product can be introduced in the market smoothly. That's our number one priority. And so we have to offer them the best technology with enough capacity. So far, let me share with you today's situation. We are in the real world production of 5X radical size.
I don't believe that anybody can do that today except the TSMC. And the yield is excellent. So our customers are very satisfied with the result. And then we are working on the 9X radical size and moving into the 14X radical size. Of course, we are also working on the cold pass and to make sure that in the future we can offer a reasonable cost reduction. But so far, the cold wars is still the number one choice for our customer.
As time goes by, yes, our competitors offer a very competitive solution. still I want to say that today's situation, we are the only one that, you know, make sure that the advanced packaging is a 5X radical size. And still, we are very competitive. I want to emphasize that. And then Charlie's question, if the competitor cannot deliver on their solution, Charlie, I think you were saying, will he be willing to be something like a backup plan? I try to avoid the answer directly, but let me say that.
Yes, we will be number one priority. I already indicate that TSMC's customers' product can be introduced into the market smoothly. That's a number one priority. Okay. And that gives you enough hint. Yes, indeed. Indeed.
So, thank you for all the answers. I hope you don't give any food to others, including the advanced page. Thank you. Thank you very much. Thank you.
All right, operator, let's move on to the next participant, please.
Next one, Sunny Lin, UBS.
Go ahead, please.
Good afternoon. Thank you very much for taking my question. So my first question is to have a bit more color on how we should think about KPEX beyond 2026. Recall a few quarters ago you mentioned if you are doing your job well then you may be able to drive KPEX to grow a bit less than top line. But now obviously the overall cloud demand has further strengthened and you're also willing to support your clients with more substantial expansion and therefore wonder at this point when you look at maybe the coming maybe in line with sales or even that factor just like for 2026 and then meanwhile obviously the tool price is also increasing given a very tight
supply and therefore going to 2027 will the higher tool price also be a driver for high key pack as well. Thank you.
Higher, sorry. What was the last part of your question, Sunny? A tool, higher tool price, higher what?
Well, so yeah, I wonder when we think about 2027 key pack, should we also consider tool price inflation to be a meaningful variable for a lot higher key packs compared with this? Understood. Okay, sorry. Okay, thank you.
So, Sunny's first question is about our CAPEX beyond 26, really, she's asking 27, 28, the next two years. With the, you know, strong, extremely robust demand that CC indicated, question is around our capital intensity the next two years. How do we see the capital intensity trending revenue growth versus a CAPEX growth, part one? And then also the impact of tool inflation on the CapEx outlook as well.
Okay, Sunny, this is Wendell. It's a bit too early to share the CapEx number beyond this year.
But as we said in the prepared remarks, we have very strong conviction in the AI megatrend. Therefore, our CapEx from 2026 to 2028 is expected to be significantly higher than the past three years. And this is a step up of our investment to support the customer's growth. We have no intention to leave any food on the table for others. The tool price inflation is real. The tool price, material, labor costs, et cetera. However, the majority of the CAPEX is always related to the long-term demand.
So that will be the composition going forward. And capital intensity wise, as we always said, if I do our job right, the increase in revenue will outpace the increase in CAPEX. And that is still the case from what we are looking at.
Got it. Thank you very much. If I may have a very quick follow up on the CAPEX side. earlier, Wendell, you raised the margin dilution from overseas to, I think, 6 % to 4 % starting from 2027, given faster overseas expansion.
But for the US, you're still so far focused on Arizona. So does that imply for Arizona, the expansion for the second phase will happen very quickly, or you are actually accelerating the expansion for the third phase or even the other phases as well, is there for the result for higher margin dilution for next year?
Okay, so Sunny's question is kind of about the puts and takes to next year's margin. She does note, as Wendell said in our prepared remarks, that the overseas fabs will widen a bit to 3 % to 4 % in the next few years.
So she wonders what's driving that wider dilution. Cece did say we are pulling in the schedule for the second fab for 3 nanometer. Her question is, are we also pulling in third fab? What's leading to this wider dilution? And maybe, yeah.
Okay, Sunny, as we mentioned, we are doing everything we can, wherever we can, to increase our capacity to meet the strong demand from the customers. And we are also accelerating everywhere, including in the U .S.
We did mention before that the later part of the five-year period, the dilution, margin dilution from overseas fat will be three to four percentage points. And that still remains the case. So it's three to four percentage points in the second part of the five-year period.
I see. Thank you very much for the clarification. My second question, I want to follow up on the poll.
That was two questions, Sunny. So that's already two questions. Can we, sorry, but we probably just to accommodate as many as we can, we're going to move on.
Sure. Thank you for your understanding.
Operator, can we move to the next participant, please?
Next one, House Liu, Bank of America. Go ahead, please.
Yes, hi, C .P., Wendell, and Jeff. Thanks for getting my questions and congrats on the green results. My first question would be on your view regarding the potential shift in AI demand mix. You mentioned in the prepared remarks that the AI accelerator chipset demand is proliferating into authentic AI demand, including CPUs and networking switches.
We understand the overall AI investment is still pretty strong and even turning stronger across the board, but can we get more clarity that are you seeing your customers' demand shifting toward more efficient compute solutions, especially monetization gets more important, and how do we think about the incremental opportunity for your manufacturing and packaging business.
So Haas's first question is on the AI-related demand. His question is, are we seeing a shift in the AI-related demand more and more to CPU or networking switches and away from AI accelerators?
Or any type of comment?
Well, all I can say is AI's demand is extremely robust. Whether it comes from accelerator or more from the CPU side, it's not for me to make a comment. Although we did see the demand from authentic AI CPUs continue to increase, yes, that is a trend today. But other than that, everything is very robust. And I want to emphasize again, in our judgment, we consider all the parameters and the demand actually still far above our capability today to supply.
and we are working very hard to narrow the gap that's all I can say
thank you that's very clear and my second question is just regarding the your learning curve going for for your dual technology tracks strategy going forward so historically I think you have been ramping up the yield of a new node to producing the trip sets with smaller design and simpler design to improve the learning curve of a new node. But for the future node, families are supporting both a conventional power delivery for the applications like mobile and consumer applications, and superpower rail technology for HPC and AI chipsets.
Operating tool process architectures at the same thing make the learning curve more challenging, or increase the manufacturing complexity. And how should we think about the implication to your gross margins from the yield and also cost control perspective. Thank you.
Okay. So Haas's second question is, his question is really about ramping a new node, the learning curve of yield and cost, and therefore the implication to gross margin, really, I think is what he's driving at. And his question is that in the past, a new node is mostly ramped, you know, with smaller die size and simpler design.
But, you know, so, you know, today we have, you know, both the smartphone and also with the AI and HPC applications coming in. Does that increase the complexity of the node process architecture, make the learning curve more challenging or difficult, and therefore the implication to the margin?
Well, today's situation actually is a little bit different from previous years. Yes, we now have a very strong demand from smartphones and also AI HPC applications. And so the smaller size per se or the big size in the accelerator, they happen at the same time.
And this is to TSMC's advantage, let me say that. Because we are good. We have been prepared for this kind of thing. And we're happy to work with us. And so the learning curve, yes. But we are doing a reasonable job. And so the customers are very happy today.
Okay, thank you so much.
Okay, thank you. Operator, let's move on to the next participant, please.
Out the line, McQuarrie, go ahead, please.
Hi. I really appreciate my name taking my question.
I have a question on the power as a getting vector. So I want to understand how great constraint led you to reduce your customer forecast in history. And also could this tighter and tighter power grid supplies make your customer actually assess the adoption of your superpower rail products for the better performance store?
Okay, so Arthur's first question is about the power supply or electricity supply. Is it a gating factor, and have we seen that lead to a reduction in customers' forecasts?
And with the tighter, on the other side, or on the other hand, with the tighter power electricity supply, have we seen more customers' adoption of superpower rail technologies? DEAN YANG - Well, let me update that. Definitely the power supply situation is one of the consideration factor that we are looking into. And that's why also we do one more step to double check whether the power supply situation in the U .S. And, you know, we have done all the study and so far so good, let me say that.
And whether this kind of situation change our customers' behavior and move more into the backside power rail, the answer is no. you know, whether you use the backside power, those kind of metallization scheme is based on your need for your product. Okay, so some of the accelerator might be using it, some of the product like mobile phones, they don't use it. So it depends on customers or product. Did I answer the question?
Yes, thank you, CC.
Do you have a second question?
Yeah, thanks, Jess. So just recall the previous discussion on the agentic AI demand on the CPU. So I wonder, you know, according to our previous framework, demand focus framework, the middle heightens 50 growth, including the CPU or not, and if we heard our clients increase the CPU forecast significantly. When is the next time we refresh our AI demand growth numbers?
Okay. So Arthur's second question is on, his question is really on our, we already stated, but anyway, our AI accelerator definition.
He wants to know what the emergence of agentic AI AI and CPU, does that AI accelerator definition include CPUs or not? And if, when, we will provide inclusion for him.
Let me answer the question. It's not. It's not including in our AI demands calculation. The reason is very simple, because we cannot identify which CPU user in the conventional And also that it depends on our customers' application and their own location also. So, no? So far today, we stick to our initial very narrow definition on the AI's demand.
I understand. And so if we have seen a very strong CPU-driven demand, when we will revise up the AI?
Arthur, I think we already said very clearly it's not part of the AI accelerator definition. It's a narrow definition but it is certainly part of our business okay yeah sure okay operator in the interest of time can we take question from the last participant on the line please
yes go ahead please
Hello, thank you for the question.
You know, into 2027 you have n3 turning a creative and now that's crossed over and to still in its diluted brand and the overseas dilution is widening. So on the off side, there have been reports of significant price increases across your mature and advanced nodes um could you provide some color into into perhaps some of that conversation and where pricing does move does it typically apply across the existing contracted book or is it mainly being applied to incremental capacity above what customers have already committed to yeah
okay so uh uh the first question is talking about i guess you're asking about the potential puts and takes for next year?
Or you're asking specifically on pricing, sorry.
Pricing. Specifically on pricing.
Okay. So it's a very specific question that there's reports that we are increasing the price. He wants to know, does it apply across the existing contracted commitments from customers, or is it mainly for the incremental capacity?
Well, we have discussed this kind of thing before, that our pricing strategy is very strategic. It also depends on the market segment, depends on the location, and depends on the technology and something.
Once we offer the flexibility of the physical location in the U .S., it will be a little bit different from the one we produce in Taiwan. So no matter what, we try to get our value to be recognized and we will give our shareholders the maximum value of our technologies.
Understood. For the second question, something you said in your prepared remarks that stuck out to me was we will not be leaving food on the table. And earlier, I think it was maybe one or two weeks ago, there was the announcement via Global Foundries that they would be providing silicon interposers to TSMC.
And so with Lux and Color into, you know, how are you thinking through these strategic decisions in terms of procuring from some of the specialty foundries versus providing in-house?
Okay. So his second question is, again, he notes that recently we had an announcement with Global Foundry on Interposer. So how do we view these partnerships and what is our approach strategically?
Let me answer the question. We designed a deal with global fundry and this is part of our overall strategy in the back end supply.
Actually our number one priority is to help our customer to be successful in their market. We want them to be fully supported by all the supply including the front end and back end. Definitely we have a shortage on the back end. And we like to work with our so-called partner with global fundries so that we can both TSMC and the Global Fundry can support TSMC's customer for their product to be introduced with enough support on the capacity.
Understood. Thank you very much for my questions.
Okay. Thank you for your questions. Thank you for all the participants. This does conclude our question and answer session. Before we conclude today's conference, please be advised that the replay of the conference will be accessible within 30 minutes from now. The transcript will become available 24 hours from now, and you can find both will be available through TSMC's website at www .tsmc .com. So again, thank you for joining us today.
We hope everyone continues to stay well. Have a good autumn, and we hope you will join us again next quarter. Thank you, and have a good day.