Taiwan Semiconductor Manufacturing Firm Restricted (NYSE:TSM) This autumn 2023 Earnings Name Transcript January 18, 2024 1:00 AM ET
Firm Members
Jeff Su – Director, Investor Relations
Wendell Huang – Vice President and Chief Monetary Officer
C. C. Wei – Vice Chairman and Chief Govt Officer
Mark Liu – Chairman
Convention Name Members
Charlie Chan – Morgan Stanley
Bruce Lu – Goldman Sachs
Gokul Hariharan – JP Morgan
Randy Abrams – UBS
Laura Chen – Citi
Brad Lin – Financial institution of America
Krish Sankar – TD Cowen
Nicholas Barrett – Macquarie
Jeff Su
Good afternoon, everybody, and welcome to TSMC’s Fourth Quarter 2023 Earnings Convention and Convention Name. It is nice to see everybody in individual as soon as once more. That is Jeff Su, TSMC’s Director of Investor Relations and your host for right now. Right now’s occasion is being webcast dwell by means of TSMC’s web site at www.tsmc.com, or you can too obtain the earnings launch supplies. [Operator Instructions]. The format for right now’s occasion will likely be as follows.
First, TSMC’s Vice President and CFO, Mr. Wendell Huang, we’ll summarize our operations within the fourth quarter 2023 and full yr of 2023, adopted by our steering for the primary quarter 2024. Afterwards, Mr. Huang, TSMC’s CEO, Dr. C. C. Wei, and TSMC’s Chairman, Dr. Mark Lu, will collectively present the corporate’s key messages. Then TSMC’s Chairman, Dr. Mark Lu, will host the Q&A session, the place all three of our executives will take your questions.
As normal, I might prefer to remind everyone that right now’s discussions might comprise forward-looking statements which can be topic to important dangers and uncertainties, which may trigger precise outcomes to vary materially from these contained within the forward-looking statements. Please consult with the Protected Harbor discover that seems on our press launch.
And now I want to flip the microphone over to TSMC’s CFO, Mr. Wendell Huang for the abstract of operations and the present quarter steering.
Wendell Huang
Thanks, Jeff. Comfortable New 12 months, everybody. Thanks for becoming a member of us right now. My presentation had been begin with monetary highlights for the fourth quarter and a recap of full yr 2023. After that, I’ll present the steering for the primary quarter 2024. Fourth quarter income elevated 14.4% sequentially in NT greenback or 13.6% in U.S. {dollars}. as our fourth quarter enterprise was supported by the continued robust ramp of our {industry} main 3-nanometer know-how. Gross margin decreased 1.3 proportion factors sequentially to 53%, primarily attributable to margin dilution from 3-nanometer ramp.
Working margin decreased 0.1 proportion factors sequentially to 41.6%, barely forward of our steering, primarily attributable to working leverage on increased income. Total, our fourth quarter EPS was NT$9.21 and ROE was 28.1%.
Now let me transfer on to income by know-how. 3-nanometer course of know-how contributed 15% of wafer income within the fourth quarter, whereas 5-nanometer and 7-nanometer accounted for 35% and 17%, respectively. Superior Applied sciences outlined as 7-nanometer and under, accounted for 67% of wafer income.
On a full-year foundation, 3-nanometer income contribution got here in at 6% of 2023 wafer income. 5-nanometer was 33% and 7-nanometer was 19%. Superior Applied sciences accounted for 58% of complete wafer income, up from 53% in 2022.
Shifting on to income contribution by platform. HPC elevated 17% quarter-over-quarter to account for 43% of our fourth quarter income. Smartphone elevated 27% to account for 43%. IoT decreased 29% to account for five%. Automotive elevated 13% to account for five% and DCE decreased 35% to account for two%. On a full-year foundation, Smartphone, IoT, DCE decreased 8%, 17%, and 16% respectively. HPC remained flat whereas automotive elevated 15% in 2023. Total, HPC accounted for 43% of our 2023 income. Smartphone, 38%, IoT, 8%, and automotive, 6%.
Shifting on to the stability sheet. We ended the fourth quarter with money and marketable securities of NT$1.7 trillion or US$55 billion. On the liabilities aspect, present liabilities decreased by NT$56 billion primarily as a result of lower in accounts payable. On monetary ratios, accounts receivable days decreased 4 days to 31 days, whereas days of stock additionally declined 11 days to 85 days, primarily attributable to a better 3-nanometer wafer cargo.
Concerning money circulate and CapEx, throughout the fourth quarter, we generated about NT$395 billion in money from operations, spent NT$170 billion in CapEx, and distributed NT$78 billion for the primary quarter ’23 money dividend. Total, our money stability elevated NT$154 billion to NT$1.47 trillion on the finish of the quarter. In U.S. greenback phrases, our fourth quarter capital expenditures totaled NT$5.24 billion.
Now let’s take a look at the recap of our efficiency in 2023. 2023 was a difficult yr for the worldwide semiconductor {industry}, however our know-how management enabled TSMC to outperform the foundry {industry}. Our income decreased 8.7% in U.S. greenback phrases to US$69 billion or decreased 4.5% in NT phrases to NT$2.16 trillion. Gross margin decreased 5.2 proportion factors to 54.4%, primarily reflecting decrease total capability utilization and 3-nanometer ramp, partially offset by a extra favorable overseas change charge.
To increase our know-how management, we proceed to develop our R&D funding in 3-nanometer and 2-nanometer growth, regardless of a decrease income base in 2023. Thus, working margin decreased 6.9 proportion factors to 42.6%. Total, full-year EPS declined 17.5% to NT$32.34 and ROE was 26.2%. On money circulate, we spent US$30.45 billion or NT$950 billion in CapEx, whereas producing NT$1.7 trillion in working money circulate and NT$292 billion in free money circulate. We additionally paid NT$292 billion in money dividends in 2023. I’ve completed my monetary abstract.
Now let’s flip to our present quarter steering. We count on our enterprise within the first quarter to be impacted by smartphone seasonality, partially offset by continued HPC-related demand. Based mostly on the present enterprise outlook, we count on our first quarter income to be between US$18 billion, and US$18.8 billion, which represents a 6.2% sequential decline on the midpoint. Based mostly on the change charge assumption of US$1 to NT$31.1, gross margin is predicted to be between 52%, and 54%. Working margin between 40%, and 42%. This concludes my monetary presentation.
Now let me flip to our key messages. I’ll begin by making some feedback on our fourth quarter ’23 and first quarter ’24 profitability. In comparison with third quarter, our fourth quarter gross margin decreased by 130 foundation factors sequentially to 53%, primarily as a result of margin dilution from the continued ramp-up of our 3-nanometer know-how. We’ve simply guided our first quarter gross margin to be flat sequentially at 53% on the midpoint, primarily as a much less favorable overseas change charge assumption is offset by product combine adjustments attributable to smartphone seasonality.
Taking a look at full yr 2024, given the six components that decide our profitability, there are a number of places and takes I want to share. On the plus aspect, we count on our utilization charge to rise in 2024 as our enterprise recovers. Nevertheless, as we transfer, as we’ve got mentioned earlier than, and three is predicted to dilute our gross margin by about 3 proportion factors to 4 proportion factors for the total yr of 2024 because the income contribution will likely be a lot increased than in 2023.
As well as, we’ve got a technique in order that a few of our N3 capability might be supported by N5 instruments, given the robust multiyear demand. Such a plan will allow increased capital effectivity within the mid to long-term however requires value and energy within the near-term.
Most of this conversion will happen in second half of 2024 and we count on it to dilute our gross margin by about 1 proportion level to 2 proportion factors in second half of 2024. Lastly, we’ve got no management over the overseas change charge, however which may be one other think about 2024. Lengthy-term, excluding the impression of overseas change charge and contemplating our world manufacturing footprint enlargement plans, we proceed to forecast a long-term gross margin of 53% and better is achievable.
Subsequent, let me discuss our 2024 capital finances and depreciation. Yearly, our CapEx is spent in anticipation of the expansion that may observe in future years. In 2023, we spent US$30.4 billion decrease than our prior steering of roughly US$32 billion as we proceed to tighten up our capital spending the place applicable given the near-term uncertainties.
In 2024, our capital finances is predicted to be between US$28 billion and US$32 billion as we proceed to take a position to assist clients’ development. Out of the US$28 billion to US$32 billion CapEx for 2024, between 70% and 80% of the capital finances will likely be allotted for the superior course of applied sciences. About 10% to twenty% will likely be spent for specialty applied sciences and about 10% will likely be spent for superior packaging, testing, masks making, and others. Our depreciation expense is predicted to extend near 30% year-over-year in 2024, primarily as we ramp up our 3-nanometer applied sciences.
Lastly, let me make some feedback on our long-term CapEx and money dividend distribution coverage. At TSMC, a better stage of capital expenditures is all the time correlated with increased development alternatives within the following years. Previously few years, we’ve got sharply elevated our CapEx spending in preparation to seize and harvest the expansion alternatives from HPC, AI, and 5G megatrends.
Regardless of a difficult 2023, our income stays properly on monitor to develop between 15% and 20% CAGR over the following a number of years in U.S. greenback phrases, which is the goal we communicated again in January 2022 investor convention. With our 2024 CapEx steering of US$28 billion to US$32 billion, the speed of enhance of our capital spending has begun to level-off as we seize and harvest the expansion.
The goals of TSMC’s capital administration are to fund the corporate’s development organically, generate good profitability, protect monetary flexibility, and distribute a sustainable and steadily growing money dividend to shareholders. Because of our rigorous capital administration, in November, TSMC’s Board of Administrators accredited the distribution of a NT$3.5 per share money dividend for the third quarter of 2023, up from NT$3 beforehand. It will grow to be the brand new minimal quarterly dividend stage going ahead.
Third quarter ’23 money dividend will likely be distributed in April 2024. In 2023, TSMC’s shareholders obtained a complete of NT$11.25 money dividend per share, and they’re going to obtain at the least NT$13.5 per share money dividend for 2024. Within the subsequent few years, we count on the main focus of our money dividend coverage to proceed to shift from a sustainable to a steadily growing money dividend per share.
Now let me flip the microphone over to C.C.
C. C. Wei
Thanks, Wendell. Good afternoon, everybody. First, let me begin with our 2024 outlook. 2023 was a difficult yr for the worldwide semiconductor {industry}, however we additionally witnessed the rising emergency of generative AI associated functions with TSMC as a key enabler.
In 2023, weakening world macroeconomic circumstances and excessive inflation and rate of interest exaggerate and extend the worldwide semiconductor stock adjustment cycle. Concluding 2023, the semiconductor {industry} excluding reminiscence {industry} declined about 2%, whereas foundry {industry} declined about 13% year-over-year. TSMC’s income declined 8.7% year-over-year in U.S. greenback time period. Regardless of the near-term challenges, our know-how management allow TSMC to outperform the foundry {industry} in 2023 whereas we’re positioning us to seize the long run AI and high-performance computing associated development alternatives.
Getting into 2024, we forecast fabless semiconductor stock to have returned to a better stage exceeding 2023. Nevertheless, macroeconomic weak spot and geopolitical uncertainties persist, probably additional weighing on client sentiment and the top market demand.
Having mentioned that, our enterprise has buttoned out on a year-over-year foundation and we count on 2024 to be a wholesome development yr for TSMC, supported by continued robust ramp of our industry-leading 3-nanometer applied sciences, robust demand for the 5-nanometer applied sciences and sturdy AI-related demand. Coming off the steep stock correction and low base of 2023. For the total yr of 2024, we forecast the general semiconductor market, excluding reminiscence to extend by greater than 10% year-over-year.
Whereas foundry {industry} development is forecast to be roughly 20%. For TSMC, supported by our know-how management, a broader buyer base, we’re assured to outperform the foundry {industry} development. We count on our enterprise to develop quarter-over-quarters all through 2024 and our full-year income count on to extend by low-to-mid 20% in U.S. knowledge phrases.
Subsequent, let me discuss our N3 and N3E ramp-up and progress. Our 3-nanometer know-how are the most-advanced semiconductor know-how in each PPA and transistor know-how. Because of this, virtually all of the world’s smartphone and HPC innovators working with TSMC on 3-nanometer applied sciences. Our N3 efficiently entered quantity manufacturing and revel in a robust ramp in second half ’23, accounting for six% of our complete wafer income in 2023. N3E additional leveraged the robust basis of N3 to increase our N3 household with enhanced efficiency, energy, and yield. N3E has already into quantity manufacturing within the fourth quarter of 2023.
Supported by sturdy demand from clients in each smartphone and HPC functions, we count on income from our 3-nanometer know-how to greater than triple in 2024 and account for mid-teens proportion of our complete wafer income. We additionally proceed to offer additional enhancement of our N3 know-how, together with N3P and the N3X. With our technique of steady enhancements of our 3-nanometer course of applied sciences, we count on robust multiyear demand from our clients and are assured that our 3-nanometer household will likely be one other massive and long-lasting node for TSMC.
Now I’ll discuss concerning the AI-related demand and our N2 standing. The surge in AI-related demand in 2023 helps our already robust conviction that the structural demand for energy-efficient computing will speed up in an clever and linked world. TSMC is a key enabler of AI functions. Regardless of which method is taken, AI know-how is evolving to make use of extra complicated AI fashions, as the quantity of computation required for coaching and affect is growing. Because of this, AI mannequin should be supported by extra highly effective semiconductor {hardware}, which requires use of essentially the most superior semiconductor course of applied sciences.
Thus, the worth of TSMC know-how place is growing and we’re well-positioned to seize the main portion of the market when it comes to semiconductor element in AI. To deal with unassessable AI-related demand for energy-efficient computing energy, clients depend on TSMC to offer essentially the most modern processing know-how at scale with a reliable and predictable cadence of know-how providing.
On the identical time, as course of know-how complexity elevated the engagement lead time with buyer additionally began a lot earlier. There’s virtually all of the AI innovators are working with TSMC and we’re observing a a lot increased stage of buyer curiosity and engagement at N2 as in contrast with N3 at an analogous stage from each HPC and the smartphone functions.
Our 2-nanometer know-how will undertake slim sheet transistor construction and be essentially the most superior semiconductor know-how within the {industry} in each density and vitality environment friendly when it’s launched in 2025. Our N2 know-how growth is progressing properly with system efficiency and yield on monitor or forward of plan. N2 is on monitor for quantity manufacturing in 2025 with the ramp of much like N3.
As a part of our N2 know-how platform, we additionally developed the N2 with bottom energy rail resolution, which is best fitted to particular HPC functions primarily based on efficiency, course, and maturity concerns and we count on at energy rail will likely be out there within the second half of 2025 to clients with manufacturing in 2026. With our know-how of steady enhancement, N2 its spinoff will additional prolong our know-how management place and enabled TSMC to seize the AI-related course alternatives going to the long run.
Lastly, let me discuss our specialty know-how methods at mature mode. For TSMC right now round 70% of our complete income is 16-manometer and extra superior node, which ranging contribution from 3-nanometer and 2-nanometer applied sciences within the subsequent a number of years. This quantity will solely enhance. Our mature node publicity is a run 20% of our complete income.
TSMC’s technique at mature node is to what we closure with strategic accomplice to develop specialty know-how options to mid-customers the requirement and create differentiated and long-lasting the place you to clients. Now focus is to pure increased capability for specialty applied sciences relatively than only a nominal capability. To the event of the differentiated specialty applied sciences, the profitability of our mature node might be round our company common gross margin.
Trying forward, we forecast 28-nanometer would be the candy spot for our embedded reminiscence functions and we count on our long-term structural demand at 28-nanometer to be supported by a number of sorts of specialty applied sciences. Thus, we’re increasing our 28-nanometer specialty manufacturing capability abroad to assist the long-term structural market demand. We consider demand for the differentiated specialty know-how will stay regular regardless of the potential {industry} capability enhance and our utilization charge and structural profitability, and however you all know might be properly protected sooner or later.
This concludes my current ready remarks. And now, let me flip the microphone over to Mark.
Mark Liu
Thanks, C.C. Good afternoon, everybody. First, let me discuss our world manufacturing footprint replace. TSMC’s mission is to be the trusted know-how and capability supplier for the worldwide logic IC {industry} for years to return. In right now’s fractured globalization surroundings, our technique is to develop our world manufacturing footprint to extend our buyer belief, develop our future development potential, and attain for extra world expertise. Our abroad determination are primarily based on our clients’ wants and a vital stage of presidency subsidy or assist. That is to maximise the worth for our shareholders.
Firstly, in Japan, we’re constructing a specialty know-how fab in Kumamoto, which is able to make the most of 12-nanometer, and 16-nanometer, and 22-nanometer, and 28-nanometer course of applied sciences. We are going to maintain a gap ceremony for this fab on February ’24, subsequent month. And quantity manufacturing is on monitor for the fourth quarter of 2024.
In Arizona, we’re in shut and fixed communication with U.S. authorities on incentive and tax credit score assist and making robust progress in facility provide chain infrastructure, utility provide, and gear set up for our first fab. We proceed to work intently and develop robust relationships with our native union and commerce companions in Arizona, together with not too long ago signed an settlement with Arizona Constructing and Development Trades Council our new framework for cooperation. This settlement extends our collaboration throughout enhanced workforce coaching and growth, shared dedication to web site security, hiring native employees, and establishing common communication. It’s a win-win for all events.
We’re properly on monitor for quantity manufacturing of N4, or 4-nanometer course of know-how in first half of ’25 and are assured that when we start operations, we can ship the identical stage of producing high quality and reliability in Arizona as from our fabs in Taiwan.
In Europe, we plan to construct a specialty know-how fab spend money on Germany, specializing in automotive and industrial functions with our three way partnership companions. We proceed to be in shut communication with the German federal, state, and metropolis governments and their dedication to this challenge stays robust and unchanged. Fab development is scheduled to start in This autumn 2024 this yr.
In Taiwan, after all, we proceed to spend money on and develop our superior know-how capacities to assist our clients’ wants and their development. Given the sturdy multiyear demand for our 3-nanometer applied sciences, we’re increasing our 3-nanometer capability in Taiwan Science Park. We’re additionally getting ready our N2 quantity manufacturing beginning in 2025. We plan to construct a number of fabs or a number of phases of 2-nanometer applied sciences in each Hsinchu and Kaohsiung Science Parks to assist a robust structural demand from our buyer C.C. simply talked about.
In Kaohsiung Science Park, the federal government approval course of is ongoing and can also be on monitor. Whereas the preliminary value of abroad fab, I beforehand talked about are increased than TSMC’s fab in Taiwan. We’re assured to handle and decrease the fee hole and stay dedicated to ship worthwhile development and maximize the worth for our shareholders.
Now let me discuss my retirement. On December nineteenth final yr, I introduced that, I’ve determined to not search nomination of Board members for the following time period and can retire from the corporate after the 2024 Annual Shareholders Assembly in June. Enable me to say this, over the previous 30 years, I’ve been extremely lucky to have the ability to work at and contribute to TSMC.
I began at TSMC 30 years in the past as a frontrunner of a small four-person fab development crew. It has been my privilege to function Chairman of TSMC and after our legendary Founder, Dr. Morris Chang, over the past six years. Throughout this time, we’ve got reaffirmed our dedication to our mission, to be the trusted know-how and capability supplier to the worldwide logic IC {industry} for years to return, whereas adhering to our core values of integrity, dedication, innovation, and buyer belief.
TSMC’s success relies on offering the {industry}’s most modern processing know-how at scale, in a most effective and cost-effective method, to allow all of the innovators to efficiently provide their greatest merchandise to the world. We collectively have labored diligently to boost our deal with our know-how management, competitiveness, world manufacturing footprint, digital excellence, sustainability, and company governance to maximise the worth for our clients and our shareholders. The previous 30 years with TSMC has been a rare journey for me and I need to prolong my sincerest due to our unbelievable, proficient crew and all our TSMC’s colleagues whose diligence, dedication, and can-do spirit have made the corporate into what it’s right now.
Now TSMC’s nomination, company governance and sustainability committee of the board has beneficial Dr. C. C. Wei to succeed as the corporate’s subsequent Chairman, topic to the election of the incoming board in June 2024. If Dr. Wei is elected to be Chairman, he also needs to proceed in his present function as CEO. Supported by a deep and skilled crew of Senior Executives, a lot of whom have been with TSMC for a lot of, a few years.
As I look forward to spend extra time with my household and beginning the following chapter of my life after my — our AGM in June, I stay totally assured in TSMC’s technique, management, and execution and firmly consider TSMC will proceed to carry out outstandingly within the years forward.
Thanks to your belief in TSMC, and the perfect is but to return for the corporate and its shareholders. This conclude my messages and our key messages collectively. Thanks to your consideration.
Jeff Su
Thanks, Chairman. This concludes our ready statements. So, earlier than we start the Q&A session, I want to remind everyone to please once more restrict your questions to 2 at a time, so we will permit all of the contributors a possibility to ask their questions. Questions will likely be taken each from the ground and in addition from the decision on-line. Must you want to increase your questions in Chinese language, I’ll translate it to English earlier than our administration solutions your query. [Operator Instructions].
So now let’s start the Q&A session. Once more, our Chairman, Dr. Mark Lui would be the host. Let’s take the primary two questions from the ground, please.
Query-and-Reply Session
A – Jeff Su
Okay. Our first query comes from Charlie Chan from Morgan Stanley.
Charlie Chan
It is nice to see you once more in individual, Comfortable New 12 months. Enable me to stay seated. I’ve some lengthy query to you. So, first query is to C. C. I’m very interested by your feedback concerning the know-how management, as a result of your competitor is aware of all buyer intel, states that their PPA is forward of your 2-nanometer, even the fee is decrease. So, I need to seek the advice of your opinion, why there is a totally different story and the way can we choose. And given these debates, how TSMC goes to plan the long run capability for this buyer and in addition competitor, we need to seize this chance, but in addition keep away from any overexpansion?
Jeff Su
Okay. Thanks, Charlie. Simply please permit me for the good thing about the viewers right here in individual and on-line to summarize your query. So, Charlie, first query is round form of the know-how management and in addition our relationship, I suppose, or capability planning with a particular IDM. So, the primary a part of his query is on the know-how half. He notes this IDM. It says their PPA is forward of TSMC’s 2-nanometer and the fee might be decrease. We mentioned our know-how is {industry} main, so how can we reconcile the distinction? And in addition, how can we plan the long run capability planning for such kind of buyer?
C. C. Wei
Charlie, you named my buyer’s title, that is my buyer and my competitor. Let me repeat the final time once I are available on their know-how. The remark stays the identical, in order that they’re new is the know-how could be very comparable or equal to TSMC’s N3P. We additional examine once more with all of the specs or the attainable revealed in know-how, transistor know-how and every little thing. My remark stays the identical with an enormous benefit within the know-how maturity. Due to, in 2025, after they say that their latest know-how will likely be go on manufacturing. For TSMC, that would be the third yr with a really high-volume manufacturing within the fabs.
So once more, I do not need to make an excessive amount of of a touch upon my clients’ declare, however let me guarantee you. We proceed to have a know-how management and we proceed to have a broader base of buyer and virtually everyone, virtually. They’re working with TSMC.
Charlie Chan
In that case, would you aggressively spend the capability as a result of outsourcing is extra possible?
C. C. Wei
Actually. We’re increasing our capability with US$28 billion to US$32 billion. That is an enormous cash. That what we used for 3-nanometer and 2-nanometers at capability.
Mark Liu
Let me add some coloration to this. I believe C. C. has been very modest. I believe he is claimed that N3P is akin to their 18A. We nonetheless affirm our assertion. However I would love you to take a look at a special perspective. And what C. C., what the opposite aspect, the declare is likely to be proper, however it’s solely to their very own product. And IDN usually their know-how for their very own product. We’ve foundry us, we optimize our know-how for our clients’ product. So, that is an enormous distinction. What you utilize for the high-power server may very well be very totally different than what used was the sketches in your fingers, smartphone and even the massive knowledge edge AI processors. So, it’s best to have a look at this, I believe the time in contrast with PPA, we nonetheless affirm our assertion, however I believe simply have a look at our clients’ motion that simply inform us all of the tales.
Charlie Chan
Thanks, Mark. So, Jeff, can I’m going to the second? Sure. So, Mark, so to start with, I actually respect your management. I consider our world traders respect your previous six-year, create numerous shareholders worth. Thanks. My query is concerning the content material of your speech in November. The speech was concerning the TSMC within the period for AI. You talked about some, very attention-grabbing knowledge factors. You used AI know-how to enhance the protection clarification, additionally the EUV throughput by, for instance, 10%.
Now it is the generative AI might be very huge breakthrough when it comes to know-how. Do you assume Samsung or Intel, by leveraging the generative AI, can actually break by means of and catch up your know-how? And in addition, earlier than your retirement, any sort of an enormous unfinished aim or targets for TSMC? Thanks.
Jeff Su
Thanks, Charlie. Charlie’s second query is directed to Chairman. He famous in November, Chairman gave a speech the place he shared how TSMC has all the time been using huge knowledge, machine studying and AI to enhance our operational efficiencies. His query is, whether or not now with generative AI, will this allow or permit our rivals to do the identical factor and catch-up and slim the hole?
Mark Liu
Thanks, Charlie. The discuss I gave in final November was the viewers is the Taiwan’s {industry} firms. The aim I need to give that’s, I see synthetic intelligence is usually a nice alternative for the {industry} in Taiwan. Similar to Taiwan is an enormous nation for semiconductor, it may be huge nation for synthetic intelligence sooner or later, that is how I encourage them. So far as whether or not our competitor are utilizing AI, after all, they use AI. Simply have a look at the all the corporate AI firm in Silicon Valley or in US, that is not a secret. However however, AI is just in its nascent stage. Solely final November, the primary massive language knowledge is introduced, ChatGPT introduced. We solely see the tip of the iceberg.
I need to give a {industry} a optimistic be aware that, although 1-nanometer or sub 1-nanometer may very well be difficult, however we’ve got a brand new know-how functionality utilizing AI to speed up the innovation in science and that’s our half. And we’ve got been engaged on that for a few years already. So, after all, it is a truthful competitors. It is no secret.
Jeff Su
You have got one other query about one?
Mark Liu
Sure. Certainly. I’ll retire in June and from now to June is a very long time. Within the firm, a number of factor can occur. I hope we positively execute to C.C.’s forecast of this yr. I believe by the center of this yr, I believe we’re fairly positive, we will accomplish that. After all, C.C. simply talked about our know-how growth is on the slew of success. By June this yr, we are going to know what we’re going to fare in 2025. I give our govt a milestone. I do not need to share with you, however it should be very thrilling for TSMC.
And naturally, any longer, I merely need to encourage our folks in TSMC that the world has modified. Similar to you talked about, we’ve got to make use of synthetic intelligence for future know-how. So, we are going to go into world, we strive our world footprint, and we are attempting our digital excellence. By digital excellence, you imply we won’t depend on the hardworking of Taiwan engineer.
Solely we’ve got to recreate our job to faucet their abilities and elevate up the semiconductor know-how engineering to a special stage base on what we have already got. And naturally, the company governance is one factor I all the time in my coronary heart. Throughout this transition, I need each govt and our board to stick to the sound company governance in order that we be certain all the method steps is abided by our moral governance guidelines. Thanks.
Jeff Su
Okay. Thanks, Chairman. Let’s take the, transfer on to the second individual from the ground. I believe in entrance, Bruce Lu of Goldman Sachs, please.
Bruce Lu
Thanks. Once more, the query is unquestionably coming from the AI for positive. I believe as C. C. talked about, virtually each AI chip is working with TSMC. Nevertheless, the traders’ concern is all the time just like the greenback content material as a proportion of shoppers’ value for AI is quite a bit decrease than smartphone or different chips. In order additionally, C.C. talked about that, you promote a wafer to your buyer, however once you purchase it again, it is much more costly. So, can we count on that the greenback generated by TSMC from AI, might be elevated within the coming years, whether or not it is by means of like no migration or superior packaging or something we will count on or what sort of charge we will count on for that?
Jeff Su
Let me summarize Bruce’s first query. I believe, once more, he’s round AI associated. He notes that, mainly, virtually all of the innovators are working with TSMC at AI, however the worth per chip that we appear to be capturing is decrease than for a smartphone or PC. So, his query is that, I believe can we count on the greenback worth captured by TSMC to extend within the subsequent few years? And can this be, this extra worth extra come from the front-end course of node wafer manufacturing or will it’s by means of the superior packaging options?
C. C. Wei
Nicely, let me reply the best one first. The income come from the entrance finish and again finish collectively. To seize the worth, sure, we’re engaged on it, positively. However first, let me say that I am very pleased that my buyer has been very profitable within the AI space and we’re a key enabler for the AI functions. Thus far right now, every little thing you noticed on the AI, it is come from TSMC.
Now, right here come the questions, how we’ll seize the worth. We’re engaged on it, and truly, we see right now, they are saying that the element of complete worth within the AI, the entire AI knowledge middle is a really small proportion quantity. If we slim down the AI element or the semiconductor’s worth in the entire system, sure, it is a small proportion. However for TSMC, we have a look at our gross sales. The AI’s CAGR, that is the gross charge yearly is about 50%, and we’re assured that we will seize extra alternatives sooner or later. So, that is what we mentioned that as much as 2027, we’re going to have excessive teenagers of the income from very slim, we outline the AI utility processor, to not point out concerning the networking, to not point out about all others.
And to additional prolong our price, truly, all the sting system together with smartphone, together with the PC, they begin to put the AI’s utility aspect, they’ve some sort of a neural processor, for instance. So, the silicon content material will likely be drastically elevated. Though the unit it is truly a low single digit in CAGR, however the silicon content material is extra necessary. So, put all collectively, if we run among the AI associated utility, truly, it is fairly an enormous quantity for TSMC to develop.
Jeff Su
Okay. Bruce, do you might have a second query, please.
Bruce Lu
The second query is extra for the know-how management. I imply, as we’re shifting into the nanosheet or superior node, we see one other know-how outlined DeFi, these days. For instance, like a high-NA, EUV instruments, TSMC appears to have a special view with different friends. I imply, prior to now, like, 20 years, 30 years, there are a number of know-how DeFi that TSMC all the time select the best choices, proper? So, are you able to inform us that why you select your present route in comparison with your friends? What’s the execs and cons? What is the benefit? And the way assured that TSMC can leverage that to be the important thing sectional issue for the vanguard?
Jeff Su
Okay. Bruce’s second query is with regard, I believe, to know-how growth and decision-making, mainly. He notices that, right now there’s divergence or his phrases DeFi between totally different firms’ know-how determination whether or not to undertake nanosheet transistor construction, whether or not to undertake high-NA instruments, he is aware of prior to now that this has all the time occurred in our {industry}, however TSMC has in some way managed to make the best determination. So, he’s asking particularly how, what can we have a look at or consider in our decision-making course of? What are the professionals and cons and benefits and possibly most significantly, how assured are we about our know-how choices going ahead whether or not nanosheet or high-NA given our rivals’ actions?
C. C. Wei
Bruce, you requested a really technical query, I am not very positive everyone know the high-NA or is nanosheet or that every one run, however let me reply the query. We all the time make the best determination and our monitor document present that. Is that sufficient? Okay. Let me elaborate somewhat bit extra due to know-how itself isn’t any worth, just one can serve your buyer. So, we all the time work with our buyer to present them the perfect transistor know-how and the perfect power-efficient know-how and at an affordable value.
Extra importantly, the know-how maturity that within the high-volume manufacturing, that is all necessary. The whole lot comes collectively, so each time we all know that there are some new construction, new instruments similar to excessive NA UV, we have a look at it rigorously, have a look at the maturity of the instruments, have a look at the price of the instruments and have a look at the schedule of that, easy methods to obtain it. We all the time make the best determination on the proper second to serve our clients. And up to now, all our clients are proud of TSMC’s progress. Did I reply your query? Virtually everyone work with TSMC on 2-nanometer, besides one.
Jeff Su
All proper. Let’s go to the web. Take the following two questions from the contributors, who’re dialing in by way of the convention name, please. Operator, may you please state the title and firm.
Operator
Our query from the road is Gokul Hariharan from JPMorgan.
Gokul Hariharan
Sure. [indiscernible] I merely talked about that the AI proposition [indiscernible].
Jeff Su
Okay. Gokul, I would like you to decelerate a bit as a result of the road shouldn’t be that clear. I do assume I acquired his query, which is, he needs to substantiate, C. C., you talked about that, we’ve got a really slim definition, we name server AI processor contribution and that you simply mentioned it may be excessive teenagers in 5 years’ time as a result of the final time we mentioned low teenagers.
C. C. Wei
The demand is instantly being elevated since final, I believe final yr, the Q1, as much as March or April, when ChatGPT grew to become fashionable. So, buyer reply rapidly and ask TSMC to arrange the capability each in front-end and the back-end and that is why we’ve got confidence that this AI’s income will enhance. We solely slim all the way down to the AI utility course of, by the best way. So, we have a look at ourselves that, we put together the know-how and the capability in each our front-end and in addition our back-end. And so, it is within the early stage up to now right now. We already see the rise the momentum and we count on, for those who guys proceed to trace this one, the quantity will enhance. I’ve confidence to say that, though I do not understand how a lot.
Jeff Su
So excessive teenagers, you verify?
C. C. Wei
Or increased.
Jeff Su
Okay. Gokul, hopefully that clarifies that first query.
Gokul Hariharan
Thanks a lot. The second is about gross margin. Within the downturn, we’re forming up at a lot increased [indiscernible] than earlier than. discuss somewhat bit about what occurs once we get again to shut to full utilization? I believe we’re nonetheless working at properly under full utilization in 2023. And will you additionally clarify the gross margin dilution that you simply’re anticipating in second half ’23, due to this capability conversion. What precisely results in that gross margin dilution? And is that like a onetime sort of dilution that lasts for somewhat little bit of time and sort of ranges off in 2025?
Jeff Su
Okay. Thanks, Gokul. So, If I heard accurately, Gokul’s second query is round gross margin. So, two components to it, possibly the second half first, which is he’s asking I consider about this gross margin within the second half of this yr, notably with what Wendell had described, they’re plan to transform among the capability. The gross margin impression right here and is that this a onetime factor? Is that this higher capital? What does this imply within the mid- to long-term profitability? That is the primary half. After which I will go to the second half.
Wendell Huang
Proper. Second half, as I mentioned, there are two destructive components affecting our gross margin this yr. The primary one is the N3 dilution. N3 quantity will likely be a lot larger within the second half than within the first half. So the second half impression from N3 dilution will likely be between 3 proportion factors to 4 proportion factors. And in addition, the N5 capability transformed to N3 that may largely happen within the second half as properly. In order that will likely be 1 to 2 proportion level. That is for this yr.
For the long term, for those who have a look at these two components, our N3 dilution will steadily scale back, as a result of the profitability will proceed to enhance or enhance within the subsequent a number of years. And N5 transformed to N3, it is a onetime short-term impression on profitability, which is able to convey capital effectivity to us within the center to long-term. And the advantages collectively could be a lot larger than the onetime hit within the short-term.
So, for those who’re speaking concerning the longer-term profitability, together with these two components plus we’re promoting our price, our know-how worth, as C. C. talked about. We proceed to drive down the fee. We construct our capability primarily based on the long-term market profile and never the short-term cyclicality and subsequently allow us to have a reasonably good utilization. The one factor we’re not in a position to management is overseas change charge. So, for those who put all these collectively, we nonetheless consider that 53% and better long-term gross margin is achievable.
Jeff Su
Gokul, does that reply each components of your query?
Gokul Hariharan
Sure. So simply to make clear, so given we’re at a a lot decrease utilization than regular, what you counsel Wendell is that gross margin ought to get again to the mid to high-50s as soon as the up cycle begins to realize extra momentum, identical to what we noticed in 2022. Is {that a} affordable expectation?
Jeff Su
Okay, thanks. So Gokul actually, he is asking 53% and better, can it’s increased? Due to course he seems to be eventually yr the utilization was decrease and we nonetheless managed to ship. So, he is questioning as soon as utilization goes again to 4, can it get to mid to high-50s?
C. C. Wei
We’re engaged on it. Actually, we put together our capability in accordance with clients’ demand. Final yr could be very difficult as a result of everyone missed their forecast and so did TSMC and we noticed the utilization charge is fairly dangerous. And I consider everyone acquired extra expertise within the subsequent few years and so TSMC’s utilization charge will proceed to extend, I assure that.
Gokul Hariharan
So, the query is, we’re engaged on it, it may be.
Jeff Su
Okay. Thanks, Gokul. Thanks. Operator, let’s transfer on to the take the query from the second participant on the decision.
Operator
The second to ask query is Randy Abrams from UBS.
Randy Abrams
Okay. Sure. And good luck to each Mark and C.C. as you undergo the upcoming transition. I needed to ask, going again to the query on the IDM, I believe earlier you conceded that your rivals’ course of is definitely fairly good for optimizing to their very own merchandise. May you discuss your view on sustainability of the ramp of that IDM outsourcing with your personal merchandise in HPC? When you look out over the following two to a few years. When you see that persevering with to develop or reversed the place there may very well be a little bit of a cooling off from among the alternative you might have proper in entrance of you now?
Jeff Su
Okay. Thanks, Randy. So, Randy’s first query goes again to the IDM. His query is with IDM saying their know-how is fairly good. What’s the danger or how can we see the sustainability of this IDM’s outsourcing enterprise to TSMC within the subsequent two to a few years? Can this proceed to develop? Or will this reverse and return in home to the IDM? And the way can we handle our plan for this?
C. C. Wei
Randy, that is query. Truly, we’ve got taken under consideration all of the concerns, together with the IDM, can do it by MCO. We pull that one into consideration. We truly in our capability pending. Truly, we took very conservative technique to prevail our capability in this type of a state of affairs. I can not converse extra due to that is our technique.
Randy Abrams
Okay. If I can ask a follow-up truly simply by means of the CapEx, the place I believe earlier you acknowledged charge of enhance would slowdown, however I believe nonetheless in finance ought to enhance over time as you develop. When you may focus on the CapEx that you simply guided was flat, ought to we consider it as a pause the place as you begin to transfer into 2-nanometer, there must be one other wave of enhance? And second half considerably associated, however curious concerning the geographic enlargement. There’s been a number of press about new fabs in Japan, second fab and potential third superior fab. And it appears like the primary fab went easily. So, are you beginning to redirect or assume extra enlargement to Japan relatively than U.S. or probably each? So, you might have each choices as you progress to 3-nanometer?
Jeff Su
Okay. Randy, that is a number of questions. So, I will take that as your second query, okay, mainly. So, the primary a part of it’s concerning the CapEx. He notes that, Wendell mentioned, the speed of enhance is starting to stage off. Randy’s query is for this yr and take the midpoint, NT$30 billion is mainly flat. So, is that this only a non permanent pause within the CapEx? And with 2-nanometer within the upcoming years, ought to we count on the greenback quantity to return up, that is the primary half.
Wendell Huang
Okay, Randy, the CapEx greenback quantity yearly, might range. It relies on the totally different state of affairs. The speed of enhance positively is slower than the previous three years. When you have a look at, I believe the opposite means of that’s the capital depth. Previously three years, the very best level is 2021. It was over 50%, after which adopted by 47% and 43%. And this yr, for those who do the maths, it should be mid-30s. We count on within the subsequent a number of years, it would stay round mid-thirty proportion capital depth.
Jeff Su
After which the opposite a part of Randy’s query is on the geographic enlargement. He notes a number of reviews saying, we might construct a second fab in Japan and that we even might construct a 3rd. His query is admittedly, are we redirecting our abroad enlargement focus extra to Japan or has it modified something within the U.S.? Randy, I believe that is what you are making an attempt to ask, proper?
Randy Abrams
Sure, that is what I am making an attempt to ask.
Mark Liu
Are you able to repeat the query once more?
Jeff Su
Randy is saying, look, he is aware of there’s a number of discuss going to construct a second fab in Japan, possibly three. He simply needs to know, are we shifting our abroad enlargement focus to Japan from the U.S. or is there any huge important change?
Mark Liu
No, no. I believe, the second fab in Japan is in critical analysis stage. We have not introduced to the general public but and we’re nonetheless discussing with the Japan authorities. Though they’re very cooperative, so that you is likely to be ready for that. However that know-how will nonetheless be both 7 or 16, 12 applied sciences. And bear in mind, the our Kaohsiung fab, the primary fab was 28-nanometer or 7-nanometer. Now it is changing into 2-nanometer. That’s the shift, if there’s a second fab in Japan, that is our present plan.
Randy Abrams
That is useful. Rapidly on 3-nanometer, as a result of 5-nanometer was slightly-delayed. Would 3-nanometers nonetheless come two years after the brand new plan for 5-nanometer in Arizona? Thanks.
Jeff Su
In Arizona, Randy needs to know, that we’ve got 5-nanometer in first half ’25. What is the plan for the second fab with 3-nanometer?
Mark Liu
Sure. The second fab shale is underneath development. However what know-how in that shell continues to be in underneath dialogue. I believe that additionally has to do with how a lot incentives that fab, the U. S. Authorities can present. And sure, there will likely be a niche. At the least, present planning is ’27 or ’28, that will likely be time-frame. To be sincere, a lot of the fab in abroad, what truly be loaded, what know-how is being arrange, actually, it is a determination of shoppers’ demand in that space at that timing. So, nothing is definitive, however we are attempting to optimize worth for the abroad fab for TSMC.
Jeff Su
Thanks, Chairman. Thanks, Randy. Within the curiosity of time, we are going to take the following two from the ground. I believe there’s one right here first Laura Chen from Citibank.
Laura Chen
I believe we acquired a number of dialogue concerning the main place in essentially the most superior node. So, I simply have a query about, what’s your view on the mature node dynamic. In notably, we’re seeing that globally, thought-about geopolitical pressure, so we’re seeing that the fab over the place on the earth, so do you see that, in the long term, any concern on {industry} clever overcapacity? So, what’s TSMC’s technique? And in addition, what’s your view in your mature nostril profitability as properly? That is my first half.
Jeff Su
Okay. Thanks, Laura. So, Laura’s first query is on mature node technique and profitability. She notes with the geopolitical dynamics that there is a number of capability being constructed on the mature nodes, so her query is, can we see or count on an {industry} large oversupply? And possibly extra importantly, what’s the impression to TSMC’s mature node technique and profitability?
C. C. Wei
I believe your commentary is true. There is likely to be an excessive amount of of a capability being constructed proper now for mature nodes. So, the priority on overcapacity is legitimate. Now let’s discuss TSMC. As I mentioned, TSMC elevated the mature node capability for specialty know-how differentiated with others. We work with buyer and that sort of capability, truly efficient capability as we title it, is with dedication from clients loading and for the way forward for enterprise, as a result of we provide the worth for our buyer to design their product.
So, we consider that they’ll retain their merchandise worth even the capability is flattered within the {industry}. And as long as our buyer is doing properly, TSMC is doing properly. And so, the profitability, as I mentioned in my assertion, it will likely be across the company common, so we do not have concern.
Mark Liu
We converse for TSMC, okay? It may very well be {industry} points.
Laura Chen
That is very useful. Thanks. And in addition, my second query is again to AI associated. As we all know that a number of traders care quite a bit about your superior packaging progress, we additionally know that TSMC acquired an excellent progress on 3D IC SoIC. So, are you able to share with us your progress growth past CoWoS what’s your plan on the 3D IC and what is the schedule and capability you’re aiming for within the subsequent two, three years?
Jeff Su
Thanks, Laura. So, Laura’s second query is on superior packaging. She notes once more the robust demand for AI associated functions, so superior packaging, the progress, after all, CoWoS demand could be very robust. Her query is admittedly, I believe, past CoWoS into true 3D IC or integration options similar to SoIC. What’s the progress that we see the engagement from clients, the capability and mainly the outlook for these segments of the enterprise?
C. C. Wei
The demand truly could be very robust. Right now’s state of affairs that we can not provide sufficient capability to assist our buyer. And that situation will proceed in all probability all the best way to subsequent yr. Though, we’re working very exhausting to extend the capability, for instance, this yr, we’re doubling our output and nonetheless not sufficient, and so we proceed to extend for the following yr. The progress, up to now so good as a result of we invested on the superior packaging know-how for greater than 10 years already. So, we count on the expansion charge for CoWoS for 3D IC or for SoIC per se, it will likely be greater than 50% CAGR within the subsequent few years, at the least. And so, we’re assured that the demand is there. It is TSMC’s functionality to supply sufficient capability to assist our buyer.
Laura Chen
For CoWoS, you can be doubling and what is the thought concerning the subsequent yr? Do you might have any preliminary thought?
C. C. Wei
I’ll discuss to you subsequent yr.
Jeff Su
We’ve a query right here from Brad Lin from Financial institution of America Merrill Lynch. I believe within the curiosity of time, we’ll take one query sorry, from your self, after which we’ll take yet one more from the road after which yet one more in individual, if there’s any.
Brad Lin
Thanks, for taking my query. So, my query will likely be nonetheless round N3 and in addition IDM. So, as we perceive, the demand is unsure, however we will positively enhance our enterprise certainty by gaining market share. So, can we count on, some extra contribution or market share achieve particularly from IDM aspect or any extra contribution from PC aspect? Perhaps, properly, by the top of the yr or any time quickly?
Jeff Su
Okay. So, Brad’s first query is about IDM outsourcing. I believe, once more, given the know-how management that we’ve got. He needs to know, can we count on extra enterprise or outsourcing from the IDM by the top of this yr? And the way can we see it? Or is it unsure going ahead?
C. C. Wei
That’s too particular. You say the IDM outsourcing, I do know whom you discuss, so I higher to not make any remark. I state what I mentioned, we take every little thing into consideration, we welcome the enterprise, however we ready our capability enlargement.
Brad Lin
Obtained it. Thanks, very a lot.
Jeff Su
Your second query?
Brad Lin
Sure. So, it is on the superior packaging. So, we all know that the CoWoS S proper now’s a mainstream, so have the administration seen the purchasers changing to both CoWoS R, CoWoS L? After which what is the implication to income and margin profile?
Jeff Su
So, it is also a really particular, too particular query. However once more, Brad needs to know, CoWoS S appears to be the mainstream right now to see clients switching to CoWoS L or CoWoS R and what is the margin implication?
C. C. Wei
Nicely, let’s make a joke, I even did not know what is known as CoWoS R, CoWoS L. However anyway, we’re working with buyer to assist them with enough capability, though it is not 100% now however we do our greatest. And we’re creating that subsequent technology, CoWoS A or one thing like that for our buyer, and it is overwhelming that it welcomed by all my clients, so we’re getting ready the capability for you.
Brad Lin
Obtained it. That is the final one, however not the least, not the query.
Jeff Su
That is 2 questions.
Brad Lin
No, not a query. So mainly, I need to say, properly thanks, Mark, to your management, contribution and endeavor for the previous 30 years, not only for not simply TSMC, but in addition for Taiwan. We want you a cheerful retirement and in addition the brand new chapter of life. Thanks.
Jeff Su
We are going to take the ultimate questions from the final two contributors. Let’s go browsing first, then we’ve got one closing in-person, okay? Operator?
Operator
Subsequent from the decision is Krish Sankar from TD Cowen.
Krish Sankar
Thanks for taking my questions. I’ve two of them. First one, I believe Wendell, you spoke about income development for the yr and once more from gross margin steering. I am simply making an attempt to marvel how to consider gross margin for the total yr within the context of the truth that, TSMC goes to develop within the low to mid-20%, how do you consider development margin for full yr 2024? After which I had a follow-up.
Jeff Su
Okay, Krish, once more, sorry, we couldn’t hear you that clearly on-line. However I believe his query is, appropriate me if I am fallacious, with income outlook that we gave, low to mid-20s development in U.S. greenback time period, what’s the outlook for the total yr gross margin? Is that what you’re asking?
Krish Sankar
That is proper, sure.
Wendell Huang
I simply talked about a few places and takes on the gross margin of this yr. And I additionally mentioned that, the second half, we can have a better dilution from two components. However we’re not prepared to present out a full yr steering on gross margin but. We are going to discuss that as time goes by. However let me say this, longer-term, with all of the components collectively, nonetheless 53% or increased is unquestionably we’re very assured that, we are going to obtain that.
Jeff Su
Sorry, Krish, why do not you go forward?
Krish Sankar
Sure. Only a fast follow-up and I simply need to say thanks quite a bit for all assist. Simply to follow-up, when it comes to the income development for this yr, December quarter you exited HPC and smartphone, roughly 43% of income. What is going on to drive the expansion this yr? Is HPC or smartphone, which goes to be higher this yr to get to the low to mid-20%?
Jeff Su
Okay. Krish, Sorry. Once more, we couldn’t hear you that clearly, however I believe I acquired the gist simply to your query. Perhaps the best way Krish’s query is what the elements that is driving the income development this yr, possibly we will share with him by the 4 development platforms?
Wendell Huang
Chris, the HPC can have the very best development, truly a lot increased than the company. The opposite three platforms will all develop, though slower than the company.
Krish Sankar
Thanks very a lot. Thanks, C. C. Thanks, Wendell.
Jeff Su
Sure, no downside, Krish. Then we are going to take the ultimate query from the ground. The primary row right here. Nicholas Barrett of Macquarie. The microphone is on. You’ll be able to ask.
Nicholas Barrett
Very fast query. Thanks very a lot, Jeff. Is it attainable or would you count on that a few of your Arizona clients may very well be solely clients in Arizona, however some U.S. clients solely need to purchase wafers made within the U.S.?
Jeff Su
Sorry. Your query is, will clients in Arizona solely be U.S. clients?
Nicholas Barrett
Is it attainable that some U.S. clients solely need U.S. made wafers?
C. C. Wei
Why can we reply that query? Arizona fab’s for everyone. However majority is a U.S. buyer, you’re proper.
Jeff Su
Do you might have one other query? No? Okay. Nicely, if not, then this does conclude our Q&A session earlier than we conclude right now’s convention. Please be suggested that the replay of the convention will likely be accessible inside half-hour from now and the transcript will grow to be out there 24 hours from now, each of that are going to be out there by means of our web site, TSMC’s web site at www.tsmc.com.
So, thanks once more everybody for becoming a member of us right now. We hope everybody continues to remain properly and we hope you will note be a part of us once more subsequent quarter. Goodbye and have an important day. Thanks.