芯源系統 (MPWR) 2024 Q1 法說會逐字稿

Welcome everyone, to the MPS First Quarter 2024 Earnings Webinar. My name is Genevieve Cunningham, and I will be the moderator for this webinar.

歡迎各位參加 MPS 2024 年第一季財報網路研討會。我是 Genevieve Cunningham，將擔任本次研討會的主持人。

Joining me today are Michael Hsing, CEO and Founder of MPS; Bernie Blegen, EVP and CFO; and Tony Balow, VP of Finance.

今天與我一同出席的有：MPS 執行長暨創辦人 Michael Hsing；執行副總裁暨財務長 Bernie Blegen；以及財務副總裁 Tony Balow。

Earlier today, along with our earnings announcement, MPS released written commentary on the results of our operations. Both of these documents can be found on our website.

今天稍早，隨同我們的財報公告，MPS 也發布了關於營運結果的書面評論。這兩份文件都可在我們的網站上找到。

Before we begin, I would like to remind everyone that in the course of today's presentation, we may make forward-looking statements and projections that involve risk and uncertainty, risks, uncertainties and other factors that could cause actual results to differ from these forward-looking statements are identified in the safe harbor statements contained in the Q1 earnings release and in our latest SEC filings, including our Form 10-K, which can be found on our website. Our statements are made as of today, and we assume no obligation to update this information.

在開始之前，我想提醒各位，在今天的簡報過程中，我們可能會做出涉及風險與不確定性的前瞻性陳述與預測。可能導致實際結果與這些前瞻性陳述不同的風險、不確定性及其他因素，已在第一季財報新聞稿中的「安全港」聲明以及我們最新的 SEC 申報文件（包括可在我們網站上查閱的 Form 10-K）中加以說明。我們的陳述以今日為準，且我們不承擔更新此資訊的任何義務。

Now I'd like to turn the call over to Bernie.

現在我想把電話交給 Bernie。

Thanks, Gen. We're doing something a little different in today's earnings call. As a detailed recital performance metrics is included in the company's earnings commentary, accompanying the earnings release, I'll use this time to provide just a few comments on our Q1 2024 performance and our outlook before opening the call up to Q&A.

謝謝你，Gen。我們今天的財報電話會議會做一點不同的安排。由於公司隨同財報新聞稿發布的財報評論中，已包含對績效指標的詳細說明，我將利用這段時間就我們 2024 年第一季的表現與展望做幾點簡短評論，之後再開放問答。

Our financial performance improved in the first quarter of 2024 with revenue up both sequentially from Q4 '23 and year-over-year from Q1 '23, Ordering patterns consistently trended upwards through the quarter. Visibility into the second half of 2024, however, is limited and many of our customers remain cautious.

我們在 2024 年第一季的財務表現有所改善，營收不僅較 2023 年第四季呈現季增，也較 2023 年第一季呈現年增；訂單型態在整個季度持續向上。然則，對於 2024 年下半年的能見度仍然有限，且我們許多客戶依舊保持審慎。

Despite this uncertainty around the second half of 2024, customer engagement across all of our end markets remains high, and our design win pipeline continues to grow stronger. Additionally, we are continuing to expand our product portfolio and diversify our supply chain globally. We believe both actions position our company for further growth as the market improves.

儘管對 2024 年下半年存在不確定性，我們在所有終端市場的客戶互動仍維持高檔，而我們的設計導入（design win）管線也持續增強。此外，我們也持續擴充產品組合並在全球分散供應鏈。我們相信這兩項作為將使公司在市場回溫時具備進一步成長的有利位置。

In summary, we saw consistent improvement through the first quarter, but we continue to be cautious about the second half of 2024 business conditions. Overall, our proven long-term growth strategy remains intact, and we can swiftly adapt to market changes as they occur.

總結來說，我們在第一季期間看到持續改善，但對於 2024 年下半年的商業環境仍保持審慎。整體而言，我們經驗證的長期成長策略依然完好，並且能在市場變化發生時迅速調整。

I'll now open the webinar up for questions.

我現在開放研討會進入提問。

(Operator Instructions) Our first question is from Tore Svanberg of Stifel.

（接線員指示）我們的第一個問題來自 Stifel 的 Tore Svanberg。

Congratulations on this enviable consistency. I have two questions. My first question is on share gains. And historically, Michael, Monolithic Power tends to really accelerate share gains during downturns. And I know there's a lot of focus on Enterprise Data right now, which is 1/3 of your revenue, but -- can you maybe call out some other verticals or applications where you are seeing more share gains in the current downturn?

恭喜你們展現出令人稱羨的一致性。我有兩個問題。第一個問題是關於市占率提升。歷來 Michael，Monolithic Power 往往會在景氣下行期間加速取得市占。我知道目前市場很關注企業資料（Enterprise Data），它占你們營收的三分之一，但——你是否能點出在目前這波下行期間，除了企業資料之外，還有哪些垂直領域或應用看到更多市占提升？

Clearly, we just wait -- if we haven't out of this downturn. You see this, we compare the rest of the industries, we're like a 10% up. And -- so that's really a share gained. And so in terms of -- so the rest of vertical market as auto is clearly is the one. We won many sockets in there. And MPS is still small compared to all these established -- our competitors.

很明顯，我們只是等待——如果我們還沒有走出這波下行。你看，我們和其他產業相比，我們大概還成長了 10%。所以——那其實就是市占提升。至於——其他垂直市場方面，汽車顯然是其中之一。我們在那裡贏得了很多插槽（socket）。而相較於那些既有的——競爭對手，MPS 仍然算小。

And other ones, even other ones -- even the servers and notebooks areas and these are the ones, all gain shares. And the last one is the consumer, as we said it a year ago or so, we gave it up some of the shares because of the capacity constraints. Now, we have a lot more capacity now and you will see some growth in the near future.

另外一些領域，甚至其他領域——包括伺服器與筆電等，這些也都在提升市占。最後一個是消費性市場，正如我們大約一年前所說，因為產能受限，我們曾讓出了一些市占。現在我們的產能大幅增加，你們將在不久的將來看到一些成長。

And if I could add one additional comment there is that we've had a number of greenfield opportunities that have been waiting to launch in an improved macro environment. So in addition to the end markets that Michael just referenced, I think that you'll also see share gains in both communications and industrial.

如果我可以再補充一點，我們也有一些綠地（greenfield）機會，一直在等待更好的總體環境才會啟動。因此，除了 Michael 剛提到的終端市場之外，我認為你們也會在通訊與工業領域看到市占提升。

Great. And for my follow-up question and specifically on Enterprise Data, which is now 1/3 of our revenues. There's been some chatter lately about lower power management content in next-generation AI data centers due to liquid cooling and other techniques to lower overall power. Is this impacting MPS broadly? Or is this kind of more sort of a very specific use case in the server power management market?

很好。我的追問，特別是關於企業資料（Enterprise Data），它現在占我們營收的三分之一。最近有些討論提到，由於液冷以及其他降低整體功耗的技術，下一代 AI 資料中心的電源管理含量可能會降低。這是否會對 MPS 造成廣泛影響？還是這比較像是伺服器電源管理市場中非常特定的使用情境？

All these cooling systems, the new format or vertical powers, MPS involve all of them. And if they transition to those markets, those systems -- and MPS in the same gains, okay? We will gain -- and we will grow with it. Is there any other comment? Tony? Okay.

所有這些冷卻系統、新的規格或垂直電源，MPS 都有參與其中。如果它們轉向那些市場、那些系統——MPS 也會同樣受益，好嗎？我們會受益——並且隨之成長。還有其他補充嗎？Tony？好。

No, I think that the point that we've demonstrated, particularly with Enterprise Data is the ability to leverage up content as we go into higher-value technology. For example, the water cooled and vertical represent opportunities, not threats. And if you look down the line not to -- not far out, we'll also be going into rack power as well.

沒有，我想我們已經證明的一點，特別是在企業資料方面，是當我們進入更高價值的技術時，有能力提升單機含量（content）。例如，水冷與垂直方案代表的是機會，而不是威脅。而且如果你往後看——不需要看得太遠——我們也將進入機櫃電源（rack power）。

That's good. Yes, good answer.

很好。是的，回答得很好。

Our next question is from Quinn Bolton of Needham.

下一個問題來自 Needham 的 Quinn Bolton。

I'll add my congratulations on very steady performance in a challenging macro environment. I guess I wanted to follow up on Tore's question on enterprise data. Lots of your competitors making noise about perhaps gaining share at your largest enterprise data customer. I'm wondering if you could just sort of address your latest thoughts on competitive landscape in both lateral power and then perhaps looking forward to vertical power? And then I've got a follow-up.

我也要恭喜你們在充滿挑戰的總體環境下仍維持非常穩健的表現。我想延續 Tore 關於企業資料的問題。你們許多競爭對手最近在放話，說可能在你們最大的企業資料客戶那裡取得市占。我想請你們談談對競爭態勢的最新看法：在橫向電源（lateral power）方面，以及展望未來的垂直電源（vertical power）方面。然後我還有一個追問。

Yes. As you may remember, 2016, we have a power curve versus -- the power curve CPU computing capabilities and versus the power densities. And we projected it for 2018 or '19. That would be the crossover point that MPS product. And for those common footprint, driver [mass] couldn't fit into the peripheral powers. That's around 700 watts. If you remember that in 2016, we said that. And at the time, we projected with the CPU was wrong. It's actually turned out to be GPUs. But the power is the power, power densities.

是的。如你所記得，2016 年我們有一條功率曲線——功率曲線是 CPU 的運算能力相對於功率密度。我們把它預測到 2018 或 2019 年，那會是 MPS 產品的交叉點。對於那些通用封裝尺寸（common footprint），驅動器 [mass] 無法再塞進周邊電源（peripheral power）。那大約是在 700 瓦。如果你記得 2016 年我們就這麼說過。當時我們對 CPU 的預測是錯的，結果變成是 GPU。但功率就是功率、功率密度。

And -- but 2018 or '19 -- and that's pretty much the peripheral power that reach the limit. And now from that time -- and everybody goes to a vertical for even higher power AI computations or learning. So all over these power go to verticals.

而——在 2018 或 2019 年——周邊電源基本上就達到極限了。從那時起——為了更高功率的 AI 運算或訓練，大家都轉向垂直方案。所以所有這些電源都走向垂直化。

And peripheral powers reached the limit. And I can't comment on our competitors, okay, so far, MPS, well, as always, we want to bring the best technologies. And we are not a champion of volumes, okay? We don't do that. We want to every -- whatever we do, we do the best. And not the volumes that we ship.

周邊電源已達極限。至於競爭對手我不便評論，好嗎？到目前為止，MPS 一如既往，我們希望帶來最好的技術。我們不是追求出貨量的冠軍，好嗎？我們不做那個。我們希望——我們做的每一件事都做到最好，而不是看我們出貨多少。

And -- so when the market became normalized and became a normal -- not like the last couple -- last year or so. And there will be more -- more solution come on board. A lot of our competitors start to copy our products, so be it. That's fine. So we are -- we stick our MPS model, we will have a very diversified growth.

而——當市場回到常態、變成正常——不像過去幾個——去年左右那樣。那麼就會有更多——更多解決方案加入。很多競爭對手開始複製我們的產品，那就讓它發生吧。沒關係。所以我們——會堅持 MPS 的模式，我們將會有非常多元化的成長。

I think it's important to keep in mind when you look at our history, as Michael just referenced there, that we have always won opportunities due to our innovation. And as we look at the next generation of GPU or TPU or ASIC products that are in this high power end market. We're an enabling technology meaning that at the front of the design cycle, we're consulted, we're integrated, in fact, with the development of the next generation of products. So we believe that strategically that, yes, there will be competitive influences on the market but we want to continue to position ourselves as the leader.

我認為回顧我們的歷史時要記住一點，正如 Michael 剛才提到的，我們一直是靠創新贏得機會。當我們看向下一代 GPU、TPU 或 ASIC 產品——它們位於這個高功率終端市場——我們是一項賦能技術（enabling technology），也就是在設計週期的前端，我們就會被諮詢、被整合，事實上也會參與下一代產品的開發。因此我們相信在策略上，是的，市場會受到競爭因素影響，但我們希望持續把自己定位為領導者。

Got it. Maybe just a quick follow-up. Just any comments you guys have on when you think vertical power may go to volume production? Is that something that happens later in '24? Is that not in volume until sometime in 2025?

了解。也許再快速追問一下。你們對於你們認為垂直供電（vertical power）何時可能進入量產，有沒有任何看法？這會是在 2024 年稍晚發生嗎？還是要到 2025 年某個時點才會有量產規模？

And then the follow-up, Bernie, is you look at the consensus estimates, the Street's modeled up 9% sequential growth in your December quarter. Your typical seasonality, I think, is down 1% to 3%. Wondering if you have any comments as to that sort of a-seasonal pattern out in Q4? I know you're not guiding out that far, but wondering if you could make any comments about that sort of atypical growth.

然後再追問一下，Bernie，你看市場一致預估，華爾街模型推算你們 12 月季度將有 9% 的季增。你們典型的季節性，我認為是下滑 1% 到 3%。想請問你對於 Q4 這種反季節性的走勢有沒有任何評論？我知道你們不會指引到那麼遠，但想看看你是否能對這種非典型成長做些評論。

I'll answer the first part of it. And the vertical power is happening now. There's multiple of our customers, they are launching the vertical power now and we are shipping those products.

我先回答第一部分。垂直供電現在就正在發生。有多家客戶正在推出垂直供電，我們也正在出貨這些產品。

Yes. And I'll pick up the second part of your question there is that seasonality, particularly as you are exiting a downturn is hard to predict. And what we tried to indicate with our prepared comments is that there are signs of optimism from the standpoint of improved ordering patterns. But how that translates into the second half is hard to predict.

是的。我來接著回答你第二部分的問題：季節性，特別是在走出景氣下行時，很難預測。我們在事先準備的評論中想表達的是，從訂單模式改善的角度來看，確實有一些樂觀的跡象。但這會如何反映到下半年，仍然很難預測。

And so we have more of a profile of the guide that we've given for Q2. But that really as far as the difference between Q3 and Q4, we see them higher than Q2 perhaps. But between the two of them, flattish.

因此，我們對於 Q2 的指引輪廓會更清楚一些。但就 Q3 與 Q4 的差異而言，我們可能認為它們會比 Q2 高一些；但兩者之間大致持平。

Yes. We see AI probably still continue to grow and grow very fast rate.

是的。我們看到 AI 可能仍會持續成長，而且成長速度非常快。

Our next question is from Rick Schafer of Oppenheimer.

下一個問題來自 Oppenheimer 的 Rick Schafer。

My congratulations guys, and I just had a couple of questions. The first one just since we're talking about power, and I sort of have a point in one on server CPU power. Some of the new x86 platforms that are due out later this year are pushing 500 watts. I mean is there an expectation? Just like the call you made back in 2016, maybe that was a little premature. But is there expectation that x86 is eventually going to move to 48-volt still?

恭喜各位。我有幾個問題。第一個，既然我們在談電源，我想聚焦在伺服器 CPU 供電。有些預計今年稍晚推出的新 x86 平台功耗正推到 500 瓦。我的意思是，市場是否有預期——就像你們在 2016 年做出的那個判斷，或許當時有點太早——但是否預期 x86 最終仍會轉向 48 伏？

And if so, I guess, where are we in that transition? How far off do you think that is?

如果是的話，我想問我們在這個轉換進程中的位置在哪裡？你認為還要多久才會發生？

Yes. Okay. Now since you mentioned 2016, I can go back to earlier. 2014 we were not even -- we were outside the door, in 2016, we can join, we were invited as a guest. And we gave some tokens so you can play it. And so the transitions and from VR 13.5 to 14, then this type -- this time, when these new CPU power, we have a significant market share. That will keep -- and it's not reflected into revenue yet. And once those CPU release, we will gain a significant amount of shares.

是的。好。既然你提到 2016 年，我可以再往前說。2014 年我們甚至——我們還在門外；到 2016 年，我們可以加入了，我們是以受邀來賓的身分被邀請進來。我們也給了一些代幣讓你可以玩。於是轉換從 VR 13.5 到 14，然後這一次，當這些新的 CPU 功率上來時，我們已經有相當可觀的市占。這會持續——而且目前還沒有反映到營收裡。等到那些 CPU 發布後，我們會拿到相當多的市占。

And any comment, Michael, on the transition to 48-volt power for X86?

Michael，對於 x86 轉向 48 伏供電，你有什麼評論嗎？

That we don't -- that's probably -- the higher powers is over 700 watts. I think those will transition to vertical powers. And -- which were ready and other ones below that, we believe, are still using a traditional rack power that came in and use 12-volt supply.

我們不——那大概是——更高功率、超過 700 瓦的部分。我認為那些會轉向垂直供電。而我們已經準備好了；至於低於那個功率的，我們相信仍會使用傳統的機架式供電，進來後使用 12 伏供應。

Got it. And just as a follow-up, shifting gears a little bit to auto. I was just curious what your expectations are for your auto business this year? Obviously, that market is under a little bit of pressure near term. But I believe you guys have been pretty open about some of your material share gains.

了解。再追問一下，稍微換個話題到車用。我想了解你們對今年車用業務的預期是什麼？顯然短期內那個市場有些壓力。但我相信你們也相當公開地談到一些實質的市占提升。

For instance, with China, within China, ADAS also with some of your top auto -- you with your top auto customer. You've also got some pretty significant share gains, I believe, ramping later this year, but I was curious if you could provide any update there.

例如在中國、在中國境內的 ADAS，以及你們一些主要車用客戶——你們在你們最大的車用客戶那邊。我相信你們也有一些相當顯著的市占提升，並且會在今年稍晚放量，我想請問你能否提供一些最新進展。

We saw - yes, go ahead.

我們看到——是的，你先說。

I just was curious if you could add in there, what your expectations are from -- I know you had a couple of launches that some OEMs delayed in the second half last year, and I didn't know if you still expect it to benefit from those this year. I know that's kind of a lot in one question, but there's a lot going on in your auto business.

我也想順便問一下，你們的預期是——我知道去年下半年有幾個 OEM 延後了幾個新案的導入（launch），我不確定你們今年是否仍預期能從那些案子受益。我知道這一題問得有點多，但你們車用業務確實有很多事情在進行。

Yes, certainly. We actually care less with the share gain, what the revenue expectations. Of course, we have to prepare all the inventories, okay -- and that's the only thing we care. Whatever this or whatever it is, okay. We want to do -- if we're not -- if we're not the best, we will not win those market segments. And particularly, these are new applications and new features.

是的，當然。我們其實比較不在意市占提升，而是更在意營收預期。當然，我們必須準備好所有庫存，好——這是我們唯一在意的事。不管是這個或那個，好。我們想要做到——如果我們不是最好的，我們就不會贏得那些市場區隔。尤其這些都是新應用與新功能。

So far, we can tell you -- in this year's or end of the last year's, and Chinese EV makers, they produce a lot more. And with those features that we are in, they export it in -- not to U.S., but to the other possible world, they increased somewhere 5 million to 6 million units, okay, million cars. And that's where we see the upside so far.

到目前為止，我們可以告訴你——在今年或去年年底，中國的電動車廠商產量增加很多。而且搭載我們所切入的那些功能後，他們把車出口到——不是美國，而是其他可能的市場；他們增加了大約 500 萬到 600 萬台，好，約 500 萬到 600 萬輛車。這就是我們目前看到的上行空間。

Our next question is from Ross Seymore of Deutsche Bank.

下一個問題來自德意志銀行（Deutsche Bank）的 Ross Seymore。

Congratulations on the stability. Just a question on the visibility into the second half. I understand the caution, especially given everything going on in the broader market. But you also said that your bookings are improving, the order rates, the engagements, et cetera, et cetera. So is the visibility improving? It's just not as good as it used to be? I'm just trying to reconcile the bookings side improving, but the visibility not.

恭喜你們展現穩定性。我想問一下對下半年能見度的問題。我理解你們的謹慎，尤其在整體市場發生這麼多事情的情況下。但你們也提到接單在改善、下單速度、客戶互動等等。所以能見度是在改善嗎？只是沒有以前那麼好？我想釐清的是：接單面在改善，但能見度卻沒有改善，這兩者如何一致？

Sure, Ross. If we can kind of reflect that over the course of about the last 6 quarters, ordering patterns have been well below whatever we call normal. And when we're seeing improvement, doesn't mean that we're seeing that they have necessarily stabilized or that they're as predictive as when you have like 5 or 6 consecutive quarters of strong ordering patterns. So all we're trying to do right now is remain cautiously optimistic.

當然，Ross。如果我們回顧過去大約 6 個季度，下單模式一直遠低於我們所謂的正常水準。而我們現在看到的改善，並不代表它們已經必然穩定，或像你有連續 5 到 6 個季度強勁下單時那樣具有可預測性。所以我們現在所做的，就是保持審慎樂觀。

Ross, it's Tony. The only thing I'd probably add on that, right, is really focused on the design win engagement, making sure that pipeline is healthy because it's difficult to call when the market would come back. As Bernie said, it's still pretty choppy. But if we have that strong design win pipeline as well as having the supply chain diversification we were talking about, we're set to take advantage when the markets do come back.

Ross，我是 Tony。我唯一想補充的是，我們真的把重點放在設計導入（design win）的互動上，確保那個管線是健康的，因為很難判斷市場何時回來。正如 Bernie 說的，現在仍相當顛簸。但如果我們有強健的 design win 管線，再加上我們剛才談到的供應鏈多元化，我們就能在市場回溫時把握機會。

I guess as my one follow-up, a near-term question for you. Any outliers in the growth that you're guiding to in the second quarter by end market segments versus the 7% total?

我再追問一個偏近期的問題。你們對第二季指引的成長（總體 7%）中，按終端市場區隔來看，有沒有任何特別突出的異常值（outliers）？

No. Outliers means like AI stuff, Yes. As I said earlier, we still -- we still see a lot of growth, okay? I mean we are getting -- try to get all the inventory ready, okay, start to go even further, okay, and they even have a lot more upside.

沒有。你說的異常值是像 AI 那些嗎？是的。如我先前所說，我們仍然——我們仍看到很多成長，好嗎？我的意思是我們正在——嘗試把所有庫存準備好，好，開始再往前推一步，好，而且那邊甚至還有更大的上行空間。

Yes. If I look sequentially between Q2 and Q1, we do see the continuing demand profile for Enterprise Data. And I think we see also some contribution from Automotive, but the rest of our end markets are pretty flattish.

是的。如果我看 Q2 相對 Q1 的季比，我們確實看到企業資料（Enterprise Data）的需求輪廓持續存在。我也認為車用會有一些貢獻，但我們其他終端市場大致持平。

Our next question is from William Stein of Truist.

我們的下一個問題來自 Truist 的 William Stein。

Congrats on the good results, and thanks for the change in format. It's a breath of fresh air.

恭喜取得不錯的成果，也感謝你們改變呈現格式。這讓人耳目一新。

And -- you appreciate it. We try to make it life and make it easy for you.

而且——你們會喜歡的。我們努力讓它更有條理，也讓你更容易理解。

I'm hoping you can update us on the progress you pursued to try to diversify your manufacturing geographic footprint and both on the front end and back end? And then I have a follow-up, please.

我希望你們能更新一下，為了讓製造的地理布局更多元化，你們在前段與後段所推動的進展如何？然後我還有一個追問，謝謝。

Yes, I'll start, and then I'll let Michael and Bernie pick up. So I think we continue to make progress on diversification of our supply chain globally. And so I think that is proceeding as planned. And I think that we're set as customers would ask for that capability that we have it ready for them right now. So I think based on what we said in the past, so I would say no change in the expectation there. So Bernie or Michael, anything you want to add?

好的，我先開始，然後再請 Michael 和 Bernie 補充。我認為我們在全球供應鏈多元化方面持續取得進展，而且我認為一切都按計畫推進。我也認為，當客戶要求我們具備這種能力時，我們現在就已經準備好可以提供。因此，根據我們過去所說的，我會說對此的預期沒有改變。Bernie 或 Michael，你們有什麼要補充的嗎？

Yes. We built all these capacities our customer demands, okay. There's a few of them that really demand a lot outside of China. So we have more capacity. So that's the comment. Okay.

是的。我們是依照客戶需求建立這些產能，好嗎？其中有少數客戶確實強烈要求在中國以外的產能，所以我們也配置了更多產能。以上就是我的補充。

Okay. Let me try a different topic. Some product types that I know have been ramping maybe over a long period of time. One is converter -- data converters, which is a category, there's -- I think there's really one very dominant supplier, but you guys have started to get into that area. And then the other is the modules that you make when these are much more complex and they have many, many chips, I think, in each one. If you can talk to us about the recent growth in those newer categories, that would be helpful color.

好。我換個主題。有些產品類型我知道可能已經拉升了一段時間。其中一個是 converter——資料轉換器（data converters），這是一個品類，我想這裡面大概有一家非常主導的供應商，但你們也開始切入。另一個是你們做的模組，這些模組更複雜，每個裡面我想有非常多顆晶片。如果你們能談談這些較新類別近期的成長情況，會很有幫助。

Yes. Okay. Let me answer the first ones, data converter -- data converters are -- it's a technology that we're talking about, we developed in the last few years. And we want design win, we start to shipping products, and -- but however, there's only 1 or 2 products, okay? And now we're going to release a family of a product in less than a year, less than a year time. That will meet a more general market.

好的。我先回答第一個。資料轉換器——data converters——是一項技術，我們在過去幾年開發出來。我們拿到設計導入（design win），也開始出貨產品——不過目前只有 1 到 2 個產品，好嗎？接下來我們會在不到一年、不到一年的時間內，推出一個產品家族（family of products），以滿足更一般性的市場。

The other topic that's when you're talking about -- these modules. Yes, so called e-commerce modules, and we're actually doing well, okay, not exactly the e-commerce we talk about in 2018. And it's going over well over $100 million, okay. I mean -- but these are multi-chip modules is more than $100 million. Actually, all these AI products that we're shipping these all modules. And vertical powers and also the rack powers. We're shipping all these modules. So I -- in the past, I say -- I said and I'm sick and tired of selling silicon only. I want to leverage -- our leverage and monetize, our know-how will sell these power modules sales solutions.

另一個你提到的——這些模組。是的，所謂的 e-commerce 模組，我們其實做得不錯，好嗎？但不完全是我們在 2018 年談的那種 e-commerce。這部分規模已經遠超過 1 億美元，好嗎？我的意思是——這些多晶片模組（multi-chip modules）超過 1 億美元。事實上，我們正在出貨的所有 AI 產品，全部都是模組。垂直電源（vertical powers）以及機櫃電源（rack powers），我們出貨的都是這些模組。所以——我過去說過——我已經厭倦只賣矽（silicon）了。我想要運用我們的槓桿並將我們的 know-how 變現，去銷售這些電源模組的整體解決方案。

And I think that we've sort of hit on one of the strategic differentiators between MPS and a lot of our competitors is that we can offer the most flexible architecture of whether it is delivering a module or silicon die, so however the customer wants to build our silicon or modules into their end application, we have the flexibility to do just that.

我認為我們已經點出 MPS 與許多競爭對手之間的一個策略性差異：我們能提供最具彈性的架構，不論是交付模組或是矽裸晶（silicon die）。因此，不管客戶想把我們的矽或模組以何種方式整合到其終端應用中，我們都有彈性做到。

Our next question is from Matt Ramsay of Cowen.

我們的下一個問題來自 Cowen 的 Matt Ramsay。

I guess it's my first question. guys, I wanted you to maybe talk a little bit about the content per socket progression that you see for the company in the AI business, in particular, as your -- silicon providers that are driving these platforms make generational transitions there. I think there's a little bit of confusion that the big primary guy in AI is not just launching 1 product, but they're launching 3 different ones that I expect to have slightly different content. And you obviously have other customers there in AI, not just NVIDIA but others.

我想先問第一個問題。各位，我希望你們能談談在 AI 業務中，你們看到每個插槽（socket）的內容價值（content per socket）如何演進，特別是當推動這些平台的矽供應商進行世代轉換時。我認為目前有點混淆，因為 AI 的主要大客戶不只是推出 1 個產品，而是推出 3 個不同產品，我預期它們的內容配置會略有不同。而且你們在 AI 方面的客戶不只有 NVIDIA，還有其他客戶。

So if you could talk generally, Michael, about just what you're seeing from a content perspective, generation to generation there? And maybe Bernie, what are the variables on sort of blended content increases that you might see as we move forward?

所以 Michael，如果你能從內容配置的角度，概括談談你看到的世代到世代的變化？另外 Bernie，從你角度看，未來推進時，影響整體加權平均內容增加（blended content increases）的變數有哪些？

Yes. I can't talk about the -- well I don't -- we don't know how our customers -- and these are AI GPU providers, how they use it. I mean, they sell -- they have a different model, they have a different model -- they have a different variety of products. We -- as long as we know, okay, they use some -- they use IC chipping in general for peripheral power and verticals is all more than 700-watt powers, over 1,000-watt power. So these are from modules. And how many modules per CPUs, we have rough ideas, but we don't know exactly.

是的。我沒辦法談——嗯，我們也不知道客戶——這些 AI GPU 供應商——到底怎麼使用。我是說，他們銷售——他們有不同的模式，他們有不同的模式——他們有不同的產品組合。我們就我們所知，好嗎？他們一般會在周邊電源與垂直電源上使用一些 IC，功率都超過 700 瓦，甚至超過 1,000 瓦。這些都是模組。至於每顆 CPU 需要多少模組，我們有大概的概念，但不確定精確數字。

Yes. And to follow up on your second point, it's a very content-rich environment for us. It also includes memory, for example, and the CPU process -- processor in addition...

是的。接著回應你第二點，對我們而言這是一個內容非常豐富的環境。它也包含例如記憶體，以及 CPU 的製程——處理器，除此之外……

Or even optical.

甚至還有光學（optical）。

An optical as well. So I think it's really too early to put limits on content availability. In fact, we're finding new areas in order to deploy.

也包括光學。所以我認為現在要對內容可得性設上限還太早。事實上，我們正在找到新的領域來部署。

So far, we just want to handle the growth. All the demand from a memory side from opticals, from even the GPU, CPU powers. And this period is really the growth period.

到目前為止，我們只想先應付成長。來自記憶體端、光學端，甚至 GPU、CPU 電源端的所有需求。而這段期間確實就是成長期。

Got it. That's helpful. I know there's a lot of moving parts. I was going to kind of step back and ask a little bit about gross margin trajectory, it comes up a lot in my conversations. I know you're kind of in that 55% range. I guess, Bernie, what are the puts and takes here? I know there are big customers ramping in enterprise data and there are variables around that.

了解。這很有幫助。我知道變動因素很多。我想退一步問一下毛利率走勢，這在我與市場交流時常被提到。我知道你們大概在 55% 左右。我想問，Bernie，這裡面的利多與利空因素是什麼？我知道企業資料中心有大客戶正在放量，這方面也有一些變數。

There's also mix between your segments and where you're sourcing supply from, there's a lot of variables here. So what would -- I mean, are we kind of at a relative floor in the 55% or slightly above range? And what would be the variables that could drive the margin back higher?

另外，你們各事業部之間的組合，以及你們從哪裡取得供應來源，也都有很多變數。所以——我的意思是——我們是否大致已在 55% 或略高於 55% 的相對底部？而哪些變數可能推動毛利率回升？

Sure. This is a pretty simple question. So as you know, the range that we target is between 55% and 60%. We demonstrated that during the post pandemic stimulus that our margin was able to go to the higher end of that limit. And right now, for the last couple of quarters as well as our guidance here, we've maintained a non-GAAP rate of 55.7%. And really, what will enable us to go up is a change in the mix of business.

當然。這其實是個相當簡單的問題。如你所知，我們目標區間是 55% 到 60%。在疫情後刺激需求期間，我們證明了毛利率可以達到該區間的上緣。而目前，過去幾個季度以及我們這次的指引，我們都維持非 GAAP 的 55.7%。真正能讓毛利率往上走的，會是業務組合（mix of business）的改變。

Our next question is from Gary Mobley of Wells Fargo.

我們的下一個問題來自 Wells Fargo 的 Gary Mobley。

Congrats, and thank you for persistently seeing the expectation. Most of the interesting questions have been asked and answered, but I wanted to touch on pricing trends. I know you don't overlap a whole lot with the analog chip market leader. But clearly, they're being aggressive on price. So maybe if you can speak to the different product groups or business segments that might be affected by that? And as well, just generally speak about the pricing environments for your broad set of products.

恭喜，也謝謝你們持續如預期般地執行。大多數有趣的問題都已經被問過並回答了，但我想觸及一下定價趨勢。我知道你們與類比晶片市場的龍頭重疊不算多，但很明顯他們在價格上相當積極。所以也許你們可以談談，哪些不同的產品群或業務部門可能會受到影響？另外，也請更一般性地談談你們廣泛產品組合的定價環境。

Tony, you want to?

Tony，你要回答嗎？

I'll start and then let Bernie jump in, right, I think what you've heard from us in the past is that for us, because we lead on innovation and we're trying to work on the next-generation platforms and work on where we really have a differentiated advantage to our technology. We're probably less susceptible to pricing than some of the segments that are volume related or might be more mainstream. So I do know there's lots of reports out there about pricing right now. But I'd say right now, that's not impacting us in a way -- that's not impacting us in any way, and you can kind of see that in our outlook. And so Bernie will add some more on that.

我先開始，然後再讓 Bernie 補充。對，我想你們過去從我們這裡聽到的是：對我們而言，因為我們在創新上領先，並且致力於下一代平台，專注在我們技術真正具備差異化優勢的領域，所以相較於一些與出貨量相關、或更偏主流的區塊，我們可能較不容易受到價格影響。我確實知道現在市場上有很多關於價格的報導。但我會說，目前這並沒有以任何方式影響到我們，你也可以從我們的展望中看到這點。接著 Bernie 會再補充一些。

Yes. And fundamentally, part of the issue that's driving prices both additional capacity in China as well as additional capacity that's coming on with a large North American company. And on the cost side, we've always been very, very competitive regardless of the end market opportunity and...

是的。而從根本上來看，推動價格的部分原因在於中國新增產能，以及一家大型北美公司正在投放的新增產能。至於成本端，我們一向都非常、非常具有競爭力，不論終端市場機會如何，而且……

The cost regardless of what the margin is, we always have to -- we always drive the cost down. I shouldn't give you more than you asked, and it's a little bit from the futures. And so far, have nobody asked the question yet. We saw in the near-term market in the next couple of years, the AI truly trickle down to all applications. not only in the cars, and all -- in the sensors, all these -- in building for the phones and all kind of things trickle down to a different level. And that we see now all kind of activities is going on now.

不管毛利率是多少，我們在成本方面都必須——我們始終推動成本下降。我不該講超出你問的內容，這有點像是在談未來。而到目前為止，還沒有人問到這個問題。我們看到在短期市場、接下來幾年，AI 真的會滲透到所有應用，不僅是在汽車上，還有——在感測器、各種——用於手機等各種建置的東西，都會滲透到不同層級。而我們現在看到各種活動正在進行中。

And we -- there's a lot of design initial phase performing the product ideas. And these are requested from our customers. And we don't see that in the end of the last year, even in the beginning of this year, recently, we see a lot more. So that will drive -- those products definitely will drive the gross margin up.

而我們——有很多設計處於初期階段，正在落實產品構想。這些都是客戶提出的需求。我們在去年年底、甚至今年年初都還沒看到這樣的情況；但最近我們看到多了很多。所以這將會帶動——這些產品肯定會推升毛利率。

Thank you for that comprehensive answer. Last quarter, your distribution inventory, I think you characterized it -- I think you characterized the distribution inventory as being a little bit above the target level, what is it now? Have you been able to work that down to more of a normalized level?

謝謝你這個全面的回答。上一季，你們的通路庫存，我想你曾形容——我想你曾形容通路庫存略高於目標水準，現在如何？你們是否已經能把它降到較正常化的水準？

Sure. So there's sort of two ways to look at our channel inventory right now. If you keep it narrow to really the AI supply chain, we're trying to keep that elevated so that, that inventory is basically available on demand. But then with regard to all of the other end markets, we've seen it coming down nicely.

當然。所以目前看我們的通路庫存大致有兩種角度。如果把範圍縮小到真正的 AI 供應鏈，我們正試著維持在較高水位，讓那部分庫存基本上能隨需供應。但至於其他所有終端市場，我們看到庫存已經很不錯地下降。

Our next question is from Tore Svanberg of Stifel.

我們下一個問題來自 Stifel 的 Tore Svanberg。

I just had two quick follow-ups. I'm intrigued by the consumer segment, Michael, I know that's surprising when everyone is talking about AI. But I think this is the lowest percentage has ever been. But you did talk about getting some share there in the downturn. So any particular applications that we should keep an eye on as you start to grow back in the consumer area?

我有兩個簡短的追問。我對消費性部門很感興趣，Michael，我知道在大家都在談 AI 的時候這聽起來有點意外。但我想這可能是史上最低的占比。不過你確實提到在景氣下行時你們拿到了一些市占。所以在消費性領域開始回升時，有沒有特定應用是我們應該留意的？

Yes. Okay. That's good question, I mean, yes, that's -- I said earlier that fell into an unhealthy percentage. We want to have a diversified growth, we neglected our consumers. But on the other hand, is during this period, a lot of our competitors care less about cost. We have less effect, but we did lower the price in the notebook market segment. And that's probably you see it now, we grow the notebook revenues faster than notebook market growth. And these are low-margin products.

是的。好的。這是個好問題，我的意思是，是的，那——我先前說過，這掉到一個不健康的占比。我們希望成長更分散，我們忽略了消費性市場。但另一方面，在這段期間，很多競爭對手對成本不那麼在意。我們受到的影響較小，但我們確實在筆電市場區隔降價。這大概就是你現在看到的：我們的筆電營收成長快於筆電市場的成長。而這些是低毛利產品。

But when looking at -- looking forward, audio product. A lot of it related to a consumer product. Maybe, Tony, you can mention -- you can talk about this is okay.

但往前看——音訊產品，其中很多都與消費性產品相關。也許 Tony，你可以提一下——你可以談談這個，沒問題。

I'll go and Michael can certainly talk a lot more, but I'd point you back to what we announced last quarter on Axign acquisition. And we really looked at their technology and saw an opportunity to participate in the high-end audio market and MPS's scale and Axign's technology really offered an opportunity to go after customers, and we're seeing some traction in that area. So consumers got a wide base of potential applications, and I think we'll continue to pick our shots where we can sort of really add value with our innovation.

我先說，Michael 當然可以補充更多。不過我想請你回顧我們上季宣布的 Axign 收購案。我們檢視了他們的技術，看到一個切入高階音訊市場的機會；MPS 的規模加上 Axign 的技術，確實提供了爭取客戶的機會，而我們也在那個領域看到一些進展。所以消費性市場有很廣的潛在應用基礎，我想我們會持續挑選我們能以創新真正創造價值的切入點。

Yes. I can tell you, we will start to ship in Q3 but -- Axign and MPS together as a bundle.

是的。我可以告訴你，我們會在第三季開始出貨——Axign 與 MPS 會以套裝（bundle）的方式一起出貨。

Yes. That's exciting. And my last question is on communications. It was actually up sequentially this quarter. Everything that we're hearing, of course, is that, that market is still very weak. In your prepared remarks, you said that, that growth was driven by networks. So can you elaborate a little bit what's going on there? Are you actually starting to see that market turning? Or is this just pure share gains?

是的，這很令人振奮。我最後一個問題是關於通訊。本季其實是季增。當然，我們聽到的都是那個市場仍然非常疲弱。在你們的準備稿中，你提到成長是由網路帶動的。所以你能否多說一點那邊發生了什麼？你們是否真的開始看到市場轉向？還是這純粹是市占提升？

It remains a flattish market. I think that in quarter-by-quarter, we're going to see infrastructure go up or down. But the full year outlook is that it will probably be flat with '23.

它仍然是一個大致持平的市場。我認為按季度來看，基礎設施會上上下下。但全年展望應該會與 2023 年大致持平。

I'm betting on the communication market, it will trickle down from the speed, and demand of higher speed will trickle down all these segments. And we have a lot of design wins in 5Gs or in other high speed or Wi-Fi, and we're ready for that.

我押注通訊市場會因為速度而逐步滲透；對更高速的需求會滲透到所有這些區隔。我們在 5G 或其他高速、或 Wi‑Fi 上有很多設計導入（design wins），我們已經準備好了。

So Michael, what you mean is AI will trickle down. And as that happens, the comm sector will revive?

所以 Michael，你的意思是 AI 會逐步滲透。而隨著這發生，通訊產業就會復甦？

That's right. Yes, it has to. That's no brainer.

沒錯。是的，必然如此。這是顯而易見的。

Our last question is from Melissa Fairbanks of Raymond James.

我們最後一個問題來自 Raymond James 的 Melissa Fairbanks。

Michael, I love a no-brainer. I mean -- I love that.

Michael，我喜歡「顯而易見」。我的意思是——我很喜歡這個。

Yes, I don't need my brain to talk about it.

是的，談這個我都不需要動腦。

I'm adjusting my model accordingly. So I know you guys addressed auto earlier, and I know everyone is focused on enterprise data. But I would like to understand better, we're coming up on the model year builds for like model year 2025. I would like to know where you feel comfortable guiding the auto revenue from here?

我會相應調整我的模型。所以我知道你們先前談到汽車，我也知道大家都聚焦在企業資料。但我想更了解的是，我們即將進入像 2025 年式的車款年度（model year）投產。我想知道，從現在起，你們在指引汽車營收方面，覺得可以把握到什麼程度？

Yes. So let me take that one. A lot of the exposure we have with new content is with the EV companies. And so Michael touched on the opportunity, particularly as it relates to Chinese OEMs where we have a good installed base, particularly as it relates to ADAS or autonomous driving.

是的。那題我來回答。我們在新內容（new content）上的曝險很大一部分來自電動車公司。因此 Michael 提到的機會，特別是與中國 OEM 相關的部分——我們在那裡有不錯的既有基礎，尤其是在 ADAS 或自動駕駛方面。

And those tend not to be as seasonally driven as the internal combustion. So while we're seeing some sequential improvement quarter-by-quarter during the year, we don't necessarily -- we're not looking for that hockey stick that used to occur with automotive sales in Q3.

而這些通常不像內燃機那樣受季節性驅動。所以雖然我們看到全年按季度有一些逐季改善，但我們不一定——我們不期待像過去汽車銷售在第三季常出現的那種「曲棍球棒」式跳升。

I would now like to turn the webinar back over to Bernie.

我現在想把網路研討會交回給 Bernie。

So I'd like to thank you all for joining us on this conference call and look forward to talking to you again during the second quarter conference call, which will likely be in late July. Thank you. Have a nice day.

所以我想感謝各位參加這次電話會議，也期待在第二季電話會議時再與各位交流，時間很可能在七月下旬。謝謝。祝各位有美好的一天。