MACOM Technology Solutions Holdings, Inc. (MTSI) 2017 Q2 法說會逐字稿

Good afternoon, and welcome to MACOM's Fiscal Second Quarter Financial Results Conference Call. (Operator Instructions) As a reminder, this conference call is being recorded today, Tuesday, April 25, 2017.

下午好，歡迎參加 MACOM 會計年度第二季財務業績電話會議。(接線員指示) 提醒各位，本次電話會議將於今天（2017 年 4 月 25 日，星期二）錄音。

I will now turn the call over to Steve Ferranti, Vice President, Investor Relations, at MACOM. Steve, please go ahead.

現在我將把電話交給 MACOM 投資人關係副總裁 Steve Ferranti。Steve，請開始。

Thank you, Nicole. Good afternoon, everyone, and welcome to MACOM's Fiscal Second Quarter 2017 Earnings Conference Call. Joining me today are MACOM's President and Chief Executive Officer, John Croteau; and Senior Vice President and Chief Financial Officer, Bob McMullan.

謝謝你，Nicole。各位下午好，歡迎參加 MACOM 2017 會計年度第二季財報電話會議。今天與我一同出席的有 MACOM 總裁暨執行長 John Croteau，以及資深副總裁暨財務長 Bob McMullan。

If you have not yet received a copy of the earnings press release, you can obtain a copy at MACOM's website at www.macom.com under the Investor Relations section.

若您尚未收到財報新聞稿，可至 MACOM 網站 www.macom.com 的「投資人關係」專區取得。

Before I turn the call over to John, I would like to remind everyone that management's prepared remarks and answers to your questions contain forward-looking statements, which are subject to certain risks and uncertainties. Because actual results may differ materially from those discussed today, MACOM claims the protection of the safe harbor for forward-looking statements contained in the Private Securities Litigation Reform Act of 1995.

在我把電話交給 John 之前，我想提醒各位，管理層的事先準備發言以及對各位提問的回答中包含前瞻性陳述，且受某些風險與不確定性影響。由於實際結果可能與今日討論內容有重大差異，MACOM 主張適用《1995 年私人證券訴訟改革法》所載之前瞻性陳述「安全港」保護。

For more detailed discussion of the risks and uncertainties that could result in those differences, we refer you to MACOM's filings with the Securities and Exchange Commission, including its current report on Form 8-K filed today, quarterly report on Form 10-Q filed on February 1, 2017, and Annual Report on Form 10-K filed on November 17, 2016. Any forward-looking statements represent management's views only as of today, April 25, 2017, and MACOM assumes no obligation to update these statements in the future.

關於可能導致上述差異之風險與不確定性的更詳細討論，請參閱 MACOM 向美國證券交易委員會（SEC）提交的文件，包括今日提交的 Form 8-K 當前報告、2017 年 2 月 1 日提交的 Form 10-Q 季度報告，以及 2016 年 11 月 17 日提交的 Form 10-K 年度報告。任何前瞻性陳述僅代表管理層截至今日（2017 年 4 月 25 日）的觀點，MACOM 不承擔未來更新此等陳述之義務。

The company's press release and management's statements during this conference call will include discussions of certain adjusted non-GAAP measures and financial information, including all income statement amounts and percentages other than revenue referred to on today's call, unless otherwise noted.

公司新聞稿以及管理層在本次電話會議中的陳述，將包含對若干經調整之非 GAAP 指標與財務資訊的討論，包括除營收外於今日電話會議中提及之所有損益表金額與百分比（除非另有說明）。

These financial measures and a reconciliation of GAAP to adjusted non-GAAP results are provided in the company's press release and related Form 8-K, which was filed with the SEC today and can be found on the Investor Relations section of MACOM's website.

上述財務衡量指標以及 GAAP 與經調整非 GAAP 結果之調節表，已載於公司新聞稿及相關 Form 8-K（今日已向 SEC 提交），並可於 MACOM 網站的投資人關係專區查閱。

For those of you unable to listen to the entire call at this time, a recording will be available via webcast for at least 30 days in the Investor Relations section of MACOM's website.

若各位目前無法收聽完整電話會議，網路直播的錄音將於 MACOM 網站投資人關係專區提供，至少保留 30 天。

And with that, I'll turn over the call to John for his comments on the quarter.

接下來，我把電話交給 John，請他就本季發表評論。

Thank you, Steve. Welcome, everyone, and thanks for joining us today. I'll begin today's call with an overview of our fiscal second quarter results for 2017 and then turn the call over to Bob McMullan, our CFO, who will review our financial performance in further detail. I'll then conclude today's prepared comments by providing a summary of our execution and key highlights during the quarter, followed by an update on our 100-day integration plan for AppliedMicro and guidance for the fiscal third quarter of 2017.

謝謝你，Steve。歡迎各位，感謝今天加入我們。我將先概述我們 2017 會計年度第二季的業績，接著把電話交給我們的財務長 Bob McMullan，他將更詳細檢視我們的財務表現。之後我將以總結我們本季的執行情況與重點亮點作為事先準備發言的結尾，並提供 AppliedMicro 的 100 天整合計畫最新進度，以及 2017 會計年度第三季的財測指引。

Straight to the results. I'm pleased to report that for our fiscal second quarter of 2017, revenue came in at $186.1 million, adjusted gross margin was 58.5%, with adjusted earnings coming in at $0.63 per diluted share.

直接看結果。我很高興報告，我們 2017 會計年度第二季營收為 1.861 億美元，經調整毛利率為 58.5%，經調整每股稀釋盈餘為 0.63 美元。

Looking at our end markets. Networks was up sequentially with and without the contribution of AppliedMicro, growing 46.3% year-on-year, inclusive of the acquisition. Multi-market grew sequentially and was up 19.3% year-on-year, while Aerospace and Defense was also up sequentially, growing 23.3% year-on-year.

再看我們的終端市場。網路業務在納入與不納入 AppliedMicro 貢獻的情況下，皆較前一季成長；在包含本次收購的情況下，年增 46.3%。多元市場業務較前一季成長，年增 19.3%；航太與國防業務亦較前一季成長，年增 23.3%。

Now let me turn it over to Bob to review our fiscal second quarter financials in further detail.

接下來我把電話交給 Bob，請他更詳細檢視我們 2017 會計年度第二季的財務狀況。

Thank you, John, and good afternoon, everyone. I will review MACOM's fiscal second quarter results and financial position. But before, I would like to provide comments on our GAAP fiscal second quarter results and operating cash flow.

謝謝你，John，各位下午好。我將回顧 MACOM 2017 會計年度第二季的業績與財務狀況。不過在此之前，我想先就我們 GAAP 口徑的第二季結果與營運現金流提供一些說明。

MACOM's fiscal second quarter generated a GAAP net loss of $130.1 million and an EPS loss of $2.14 per share, respectively. The results include expenses related to the acquisition of AppliedMicro closed on January 26, 2017, including the noncash amortization of inventories stepped up to the fair market value, transaction expenses, certain change in control, payments to former AppliedMicro employees and a noncash valuation reserve of $88 million for MACOM's deferred tax asset recorded as the majority of the $89.8 million income tax expense recorded during the second quarter.

MACOM 2017 會計年度第二季 GAAP 淨損為 1.301 億美元，GAAP 每股虧損為 2.14 美元。上述結果包含與 2017 年 1 月 26 日完成之 AppliedMicro 收購相關的費用，包括將存貨按公允市價調升所產生之非現金攤銷、交易費用、若干控制權變更相關支出、支付給 AppliedMicro 前員工的款項，以及針對 MACOM 遞延所得稅資產提列的 8,800 萬美元非現金估值備抵；該備抵為第二季認列之 8,980 萬美元所得稅費用中的主要部分。

Such step-up amortization, transaction expenses and change in control payments are common in acquisitions.

此類調升攤銷、交易費用及控制權變更付款在併購案中相當常見。

Prior to the AppliedMicro acquisition, MACOM had an $88 million deferred tax asset balance, representing the estimated after-tax value of U.S. NOLs and other tax credit carryforwards primarily from MACOM's previous acquisitions. The current need for the deferred tax asset valuation reserve and the associated P&L expense during the fiscal second quarter was because of the AMCC acquisition and related expenses and the compute business will cause a MACOM U.S. tax loss in fiscal 2017, adding to other U.S. tax losses over the last few years.

在收購 AppliedMicro 之前，MACOM 的遞延所得稅資產餘額為 8,800 萬美元，代表美國淨營業損失（NOL）及其他稅額抵減結轉的稅後估計價值，主要源自 MACOM 先前的收購案。本會計年度第二季之所以需要提列遞延所得稅資產估值備抵及其相應的損益費用，是因為 AMCC 收購及相關費用，以及運算（compute）業務將使 MACOM 在 2017 會計年度於美國產生稅務虧損，並在過去數年其他美國稅務虧損的基礎上進一步增加。

AppliedMicro added an additional $70 million to the deferred tax asset balance. MACOM, under GAAP rules, had to evaluate the realizability of the total $158.6 million balance on an all-or-nothing basis with data that can change quarter-to-quarter and year-to-year.

AppliedMicro 使遞延所得稅資產餘額額外增加 7,000 萬美元。依 GAAP 規定，MACOM 必須以「全有或全無」的方式評估合計 1.586 億美元餘額的可實現性，而相關數據可能隨季度與年度而變動。

Today, it is influenced by MACOM's previous U.S. tax losses over the past few fiscal years, which led us to record the valuation reserve, reducing the deferred tax asset carrying value to 0. This is an appropriate and diligent way to account for the changing tax environment.

目前，該評估受到 MACOM 過去幾個會計年度在美國的稅務虧損影響，促使我們提列估值備抵，將遞延所得稅資產帳面價值降至 0。這是因應稅務環境變化而採取的適當且審慎的會計處理方式。

It is important to note that this valuation reserve does not prevent MACOM from utilizing its U.S. NOLs or other tax credit carryforwards over the next decade.

需要強調的是，這項估值備抵並不會阻止 MACOM 在未來十年內使用其美國 NOL 或其他稅額抵減結轉。

The cash payments from the transaction and change in control expenses were recorded through GAAP operating cash flow, not as part of the investing section of the cash flow statement. These onetime payments totaled $28.1 million. Adjusting cash flow from operations for the acquisition-related outflows increased the $400,000 GAAP reported number to $28.5 million.

交易費用與控制權變更費用所支付的現金，係透過 GAAP 營運現金流列示，而非列入現金流量表的投資活動部分。這些一次性付款合計為 2,810 萬美元。將營運現金流就收購相關現金流出進行調整後，可把 GAAP 報告的 40 萬美元提高至 2,850 萬美元。

Now to the operating results. Revenue in the fiscal second quarter was $186.1 million, growing 39.3% year-over-year and up from $151.8 million sequentially. Revenue, including AppliedMicro, Networks was $140.9 million and 75.7% of total revenues, up 46.3% year-over-year; Multi-market, $22.9 million and 12.3% of total revenues, up 19.3% year-over-year; and Aerospace and Defense, $22.3 million and 12% of total revenue, up 23.3% year-over-year.

接著看營運結果。2017 會計年度第二季營收為 1.861 億美元，年增 39.3%，並較前一季的 1.518 億美元成長。營收（含 AppliedMicro）方面：網路業務為 1.409 億美元，占總營收 75.7%，年增 46.3%；多元市場為 2,290 萬美元，占總營收 12.3%，年增 19.3%；航太與國防為 2,230 萬美元，占總營收 12%，年增 23.3%。

Of the total network revenue, Optical was $67.9 million or 36.5% of total revenue and up 9.4% year-over-year. Data center was $50.6 million or 27.2% and up 268% year-over-year. And other network revenue was $22.4 million or 12.1% and up 6.8% year-over-year.

在網路業務總營收中，光學為 6,790 萬美元，占總營收 36.5%，年增 9.4%。資料中心為 5,060 萬美元，占 27.2%，年增 268%。其他網路營收為 2,240 萬美元，占 12.1%，年增 6.8%。

Non-GAAP gross profit and gross margin in the fiscal second quarter was $108.9 million and 58.5% of revenues respectively, compared to $77.6 million and 58.1% of revenues respectively year-over-year, and $86.9 million and 57.2% respectively, on a sequential basis.

本會計年度第二季的非 GAAP 毛利與非 GAAP 毛利率分別為 1.089 億美元與營收的 58.5%，相較於去年同期分別為 7,760 萬美元與營收的 58.1%，以及按季比較分別為 8,690 萬美元與 57.2%。

Adjusted gross margins increased sequentially to improve product mix.

調整後毛利率因產品組合改善而按季上升。

In terms of operating expenses, total non-GAAP operating expenses was $60.3 million compared to $45.2 million year-over-year and $45.9 million sequentially. Adjusted operating expenses were up 33.4% year-over-year and up 31.4% sequentially due to the inclusion of AppliedMicro's expenses for part of the quarter and higher investment levels.

就營業費用而言，非 GAAP 營業費用總額為 6,030 萬美元，去年同期為 4,520 萬美元，按季為 4,590 萬美元。由於本季部分期間納入 AppliedMicro 的費用以及投資水準提高，調整後營業費用較去年同期增加 33.4%，較前一季增加 31.4%。

Adjusted R&D and SG&A expenses were $34.9 million and $25.4 million respectively, in the fiscal second quarter.

本會計年度第二季調整後研發費用與銷售、一般及行政（SG&A）費用分別為 3,490 萬美元與 2,540 萬美元。

Non-GAAP income from operations and operating margins were $48.6 million and 26.1% of revenues, up 50% in dollars and 7.4% on a percentage basis respectively year-over-year, and up 18.7% in dollars and down 3.2% or 90 basis points respectively, on a sequential basis.

非 GAAP 營業利益與營業利益率分別為 4,860 萬美元與營收的 26.1%；相較去年同期，金額增加 50%，比例增加 7.4%；按季比較，金額增加 18.7%，比例則下降 3.2%（或 90 個基點）。

The operating expenses, including the partial quarter of AppliedMicro versus the timing of its quarterly revenues, caused non-GAAP operating margin to decline which we expect to be a 1-quarter event.

由於營業費用（包含 AppliedMicro 併入的部分季度）與其季度營收認列時點不一致，導致非 GAAP 營業利益率下滑；我們預期這將是僅持續 1 個季度的事件。

Net interest expense was $6.7 million. Other income of $1.9 million represent income from the consulting contract that continues through fiscal 2017, associated with our Automotive business divestment.

淨利息費用為 670 萬美元。其他收入 190 萬美元，代表與我們汽車事業處分相關、將持續至 2017 會計年度的顧問合約收入。

Our normalized non-GAAP income tax rate for the fiscal second quarter was 10%, down from the previous 12%, due to the expanding mix of international revenues managed by our offshore entities and lower or no-tax jurisdictions.

本會計年度第二季的常態化非 GAAP 所得稅率為 10%，低於先前的 12%，主因為由我們海外實體管理的國際營收占比擴大，以及較低稅率或免稅司法管轄區的占比提高。

As to cash taxes, we had refunds and payments of $200,000, which essentially netted to 0.

就現金稅負而言，我們收到退稅並支付稅款合計 20 萬美元，基本上相抵後為 0。

Our fiscal second quarter non-GAAP net income and EPS were $39.4 million and $0.63 per fully diluted share respectively, growing from $25.7 million and $0.46 year-over-year and $31.8 million and $0.57 sequentially.

本會計年度第二季非 GAAP 淨利與每股盈餘（EPS）分別為 3,940 萬美元與每股稀釋後 0.63 美元，較去年同期的 2,570 萬美元與 0.46 美元成長，亦較前一季的 3,180 萬美元與 0.57 美元增加。

Non-GAAP net income grew 53.6% year-over-year and 23.9% sequentially. Non-GAAP EPS grew 36.9% year-over-year and 10.5% sequentially.

非 GAAP 淨利較去年同期成長 53.6%，較前一季成長 23.9%。非 GAAP 每股盈餘較去年同期成長 36.9%，較前一季成長 10.5%。

The share count used to calculate non-GAAP EPS was 62.8 million fully diluted shares.

用於計算非 GAAP 每股盈餘的股數為 6,280 萬股稀釋後流通股。

Adjusted EBITDA, or earnings before interest, taxes, depreciation and amortization was $56.6 million, up 45.2% from $39 million in our fiscal 2016 second quarter and up 17.1% from $48.4 million sequentially.

調整後 EBITDA（即息稅折舊攤銷前盈餘）為 5,660 萬美元，較 2016 會計年度第二季的 3,900 萬美元增加 45.2%，亦較前一季的 4,840 萬美元增加 17.1%。

GAAP cash flow from operations was $400,000, which included acquisition-related cash payments of $28.1 million. Cash flow from operations in the fiscal 2016 second quarter was $19.6 million and $20.4 million during the first fiscal quarter of 2017.

GAAP 營運活動現金流為 40 萬美元，其中包含與併購相關的現金支付 2,810 萬美元。2016 會計年度第二季的營運活動現金流為 1,960 萬美元，而 2017 會計年度第一季為 2,040 萬美元。

After deducting capital expenditures and excluding AMCC's acquisition-related payments, adjusted non-GAAP free cash flow was $17.1 million and 43.7% of non-GAAP net income in the fiscal second quarter compared to $8.6 million and 33% of non-GAAP net income in the fiscal second quarter of 2016, and $15.5 million and 49% of non-GAAP net income sequentially.

扣除資本支出並排除 AMCC 併購相關支付後，調整後非 GAAP 自由現金流為 1,710 萬美元，為本會計年度第二季非 GAAP 淨利的 43.7%；相較之下，2016 會計年度第二季為 860 萬美元、占非 GAAP 淨利的 33%；按季比較為 1,550 萬美元、占非 GAAP 淨利的 49%。

Turning to the balance sheet. At the fiscal second quarter end, our cash, cash equivalents and short-term investments were $144.2 million after using a net $256.2 million to acquire AppliedMicro.

接著看資產負債表。截至本會計年度第二季季末，在以淨額 2.562 億美元收購 AppliedMicro 之後，我們的現金、約當現金及短期投資為 1.442 億美元。

Accounts receivable were $127.7 million, up from $112.3 million sequentially. Days sales outstanding was 63 days compared to 67 days sequentially.

應收帳款為 1.277 億美元，較前一季的 1.123 億美元增加。應收帳款週轉天數（DSO）為 63 天，前一季為 67 天。

Inventories were up to $139.6 million compared to $115.2 million sequentially. Inventory turns were 2.2x compared to 2.3x sequentially.

存貨增加至 1.396 億美元，前一季為 1.152 億美元。存貨週轉率為 2.2 倍，前一季為 2.3 倍。

Long-term debt was $586.2 million, inclusive of capital leases. We also had $160 million of availability in an undrawn credit line.

長期負債為 5.862 億美元，包含資本租賃。我們另有 1.6 億美元的未動用循環信用額度可用額度。

Capital expenditures in the fiscal second quarter, which included $4.4 million of AMCC's related items, was $11.4 million or 6.1% of revenue compared to $4.9 million or 3.3% of revenue sequentially.

本會計年度第二季資本支出（其中包含 440 萬美元 AMCC 相關項目）為 1,140 萬美元，約占營收 6.1%；相較之下，前一季為 490 萬美元，約占營收 3.3%。

Depreciation expense for the fiscal second quarter was approximately $6.2 million as compared to $5.5 million sequentially. Our investment in capital expenditures exceeded our current levels of depreciation, reducing our free cash flow by $5.2 million in the fiscal second quarter.

本會計年度第二季折舊費用約為 620 萬美元，前一季為 550 萬美元。我們在資本支出上的投資高於目前折舊水準，使本會計年度第二季自由現金流減少 520 萬美元。

Back to you, John.

交給你了，John。

Thanks, Bob. Let's dive right into it. Last quarter was a watershed moment for MACOM. We saw a fascinating dichotomy between cyclical weakness in our carrier-based Optical businesses, which was more than offset by strong growth in cloud data centers. This underscored our previous realization that this is, in fact, a separate and distinct growth driver for MACOM in what is arguably the strongest area of secular growth in the electronics industry today, cloud data centers.

謝謝，Bob。我們直接切入重點。上一季對 MACOM 而言是個分水嶺。我們看到一個耐人尋味的對比：以電信營運商為主的光學業務出現週期性疲弱，但被雲端資料中心的強勁成長所大幅抵銷。這也再次印證我們先前的認知：雲端資料中心確實是 MACOM 一個獨立且明確的成長驅動力，而在今日電子產業中，雲端資料中心可說是長期結構性成長最強勁的領域之一。

Based on that realization, we've made the decision to focus our resources, with top priority on fueling that breakout in what we believe is shaping up to be a predictable high-value growth driver for the company for many years to come.

基於這項認知，我們已決定聚焦資源，並以最高優先順序推動這波突破；我們相信這正逐步成形為公司未來多年可預期、且高價值的成長驅動力。

With rapidly developing backlog, program awards and customer commitments, it's now clear that 100G and greater Optical connectivity in data centers is real and it's far bigger and far sooner than anyone in the industry has predicted, including us.

隨著在手訂單快速累積、專案得標與客戶承諾增加，現在已很清楚：資料中心 100G 及以上的光學連接需求是真實存在的，而且規模更大、來得更快，超出產業內任何人的預期，包括我們自己。

In aggregate, revenue from our network businesses grew sequentially as well as year-on-year, with and without the contribution of AppliedMicro. Global demand in metro and cloud data centers more than made up for very strong headwinds in Chinese carrier-based deployments in PON and provincial networks.

整體而言，我們網路業務的營收不論按季或年增皆成長，且無論是否納入 AppliedMicro 的貢獻皆然。都會區與雲端資料中心的全球需求，遠遠彌補了中國以電信營運商為主的 PON 與省級網路部署所面臨的強勁逆風。

Once again, this highlights the virtue of our diversification strategy across regions, end markets, channels and products, which has once again delivered growth despite choppiness in notoriously cyclical carrier deployments.

這再次凸顯我們在區域、終端市場、通路與產品上的多元化策略之價值；即便眾所周知電信營運商部署高度循環、波動不定，我們仍再次交出成長表現。

Last month at OFC, we observed a distinct transition from a carrier-centric to a cloud-centric focus for the show. That included MACOM. For those who attended OFC, it was readily apparent from our announcements and our product demonstrations that MACOM is well positioned with a portfolio spanning high-performance analog, lasers, silicon photonics, optical subassemblies and mixed-signal PHYs across all protocols and form factors. We're now being sponsored, actively pulled by top cloud data center customers, to provide various solutions through various customers and channels to fulfill what we expect will be insatiable demand for connectivity for many years to come.

上個月在 OFC，我們觀察到展會的重心明顯從以電信營運商為中心轉向以雲端為中心。MACOM 也包含在內。對於參加 OFC 的各位，從我們的公告與產品展示可以清楚看出，MACOM 具備良好布局：產品組合涵蓋高效能類比、雷射、矽光子、光學次組件，以及跨所有協定與外形規格的混合訊號 PHY。我們目前正獲得頂尖雲端資料中心客戶的支持並主動拉動，透過不同客戶與通路提供各式解決方案，以滿足我們預期在未來多年將近乎無止盡的連接需求。

Personally, I haven't seen such a breakout in our part of the industry since the advent of digital handsets and smartphones decades ago. It's shaping up, yes, to be that big.

就我個人而言，自從數十年前數位手機與智慧型手機問世以來，我在我們這個產業領域還沒看過如此程度的突破。它正在成形——沒錯——就是這麼大。

Having now sat at the architectural table with these major cloud customers, thanks to the AppliedMicro acquisition, and having shared our vision for delivering requisite technology and economics, we believe MACOM is well positioned to meet the requisite capacity, supply chain flexibility and pace of innovation that carrier-centric optical players have not delivered to date.

由於收購 AppliedMicro，我們如今已與這些主要雲端客戶同坐在架構設計的討論桌前，並分享我們在提供所需技術與經濟效益方面的願景；我們相信，MACOM 已具備良好條件，能夠滿足所需的產能、供應鏈彈性與創新速度，而這些正是以電信營運商為中心的光學業者迄今尚未能提供的。

It looks like we're facing an extremely supply-elastic market. Not price-elastic, supply-elastic. Unit demand in cloud data centers will outstrip carrier demand by orders of magnitude.

看起來我們正面臨一個供給極具彈性的市場。不是價格彈性，而是供給彈性。雲端資料中心的單位需求將以數個數量級超越電信營運商的需求。

Assuming supply at the right capacity and economics, whatever we can produce, and for that matter, what everyone else in the industry can produce at the right costs, will be deployed post-haste.

假設供給具備適當的產能與經濟效益，無論我們能生產多少，以及就此而言，產業中其他公司在正確成本下能生產多少，都將被迅速部署。

Long-term share will be determined by short-term operational execution. Better yet, the cloud data center is a North American phenomenon that's governed by Moore's law and fuels hyper growth in cloud-based economies. By comparison, cyclical carrier budgets invest at a more deliberate pace and are governed by public policy, which has proven time and again to be unpredictable. A great example today is China. For that reason, we view cloud data centers as a long-term sustainable, high-value market that we are well positioned to service with high-performance analog, mixed signal and photonic semiconductors.

長期市占將由短期的營運執行所決定。更重要的是，雲端資料中心是一種北美現象，受摩爾定律所支配，並推動雲端型經濟的超高速成長。相較之下，具週期性的電信營運商資本預算以更審慎的步調投資，且受公共政策所左右，而公共政策一再證明難以預測。當前一個很好的例子就是中國。因此，我們將雲端資料中心視為一個長期可持續的高價值市場，而我們也具備良好條件，以高效能類比、混合訊號與光子半導體為其提供服務。

Given that this market is currently and perhaps perpetually insatiable in demand for greater bandwidth density, we've made the decision to focus operational resources on maximizing MACOM's short-term share capture in this market above all else, thereby maximizing short-, medium- and long-term value to our shareholders.

鑑於此市場目前、且或許長期都對更高的頻寬密度有著難以滿足的需求，我們已決定將營運資源聚焦於在此市場最大化 MACOM 的短期市占取得，優先於其他一切，從而為股東最大化短期、中期與長期價值。

Operational execution and ramping MACOM production will be paramount over the coming months and quarters, so top-down focus and prioritization is key.

在未來數月與數季，營運執行與提升 MACOM 產能將是重中之重，因此自上而下的聚焦與優先排序至關重要。

Diving deeper into data centers, we had another breakout quarter with 70.3% year-on-year growth, excluding any contribution from AppliedMicro. Including AppliedMicro, data centers is now 27% of total revenue. It's now the single-largest end market we service, and it's poised to be our strongest growth driver as well.

進一步看資料中心，我們又迎來一個亮眼季度，在不含 AppliedMicro 貢獻的情況下，年增 70.3%。若包含 AppliedMicro，資料中心目前占總營收 27%。它已成為我們服務的最大單一終端市場，也有望成為我們最強的成長驅動力。

There were 3 major developments last quarter in the space. First, our high-performance analog products enabled more than 1.6 million ports in the first half of this year. That's up from 1 million for all of last year. This includes everything from laser and mixel drivers, to CDRs and TIAs, across all standards, from SR, PSM4, CWDM4, and LR4, servicing all major cloud service providers.

上季在該領域有三項重大進展。第一，我們的高效能類比產品在今年上半年支援了超過 160 萬個埠口。相較之下，去年全年為 100 萬個。這涵蓋從雷射與 mixel 驅動器，到 CDR 與 TIA，跨越所有標準，包括 SR、PSM4、CWDM4 與 LR4，服務所有主要雲端服務供應商。

Second, we saw a meaningful revenue contribution from 25GB lasers into 100G transceivers. Given the supply elasticity in this market, we've made ramping 25G laser production our top priority from a factory qualification and operational standpoint.

第二，我們看到 25GB 雷射導入 100G 收發器帶來顯著的營收貢獻。鑑於此市場的供給彈性，從工廠認證與營運角度來看，我們已將 25G 雷射的產能爬坡列為最高優先事項。

The long-term economic potential of our Etched Facet lasers and self-aligned Etched Facet L-PICs cannot be overestimated. Short-term execution is therefore imperative.

我們的蝕刻端面（Etched Facet）雷射與自對準蝕刻端面 L-PIC 的長期經濟潛力不可低估。因此，短期執行至關重要。

Third, at OFC, we showcased production units of our quad L-PICs for CWDM and 400GB PAM4 as well as our full complement of 400GB PAM4 products spanning PHYs lasers, L-PICs, drivers and TIAs. We're now working across all customers and channels and delivering these leadership solutions for the major cloud service providers.

第三，在 OFC 上，我們展示了用於 CWDM 與 400GB PAM4 的四通道 L-PIC 量產單元，以及涵蓋 PHY、雷射、L-PIC、驅動器與 TIA 的完整 400GB PAM4 產品組合。我們目前正與所有客戶與通路合作，並為主要雲端服務供應商交付這些領先解決方案。

At this juncture, visibility is good into short-term growth in data centers, with strong backlog and order intake, which we believe will continue to drive growth for the foreseeable future.

在此階段，資料中心短期成長的能見度良好，積壓訂單與接單強勁；我們相信這將在可預見的未來持續推動成長。

We will prioritize operational resources on cloud data centers, which we expect can more than offset any weakness or cyclicality on the carrier side of the business.

我們將把營運資源優先投入雲端資料中心，我們預期其足以抵消電信營運商端業務的任何疲弱或週期性波動。

So moving over to those Optical businesses. PON remained weak during the quarter, and market demand and inventory challenges persist as we enter Q3 with poor visibility. We remain cautious in our outlook for PON for next quarter.

接著談談那些光學業務。PON 在本季仍然疲弱，隨著我們以不佳的能見度進入第三季，市場需求與庫存挑戰仍在持續。我們對下季 PON 的展望仍保持審慎。

Our value proposition and market position with Etched Facet lasers remains intact. Our goal at this juncture is to sustain unit share and maximize gross margin dollars and percentage until 10G PON ramps materially, which we expect to see in 2018.

我們在蝕刻端面（Etched Facet）雷射上的價值主張與市場地位仍然穩固。目前我們的目標是在 10G PON 明顯放量之前（我們預期將在 2018 年看到），維持單位市占並最大化毛利金額與毛利率。

Metro/Long-haul, revenue was up sequentially and up 80% year-on-year.

都會／長途（Metro/Long-haul）方面，營收較前一季成長，且年增 80%。

Strength in metro and additional share gains in other regions more than offset the decline in China.

都會網的強勁表現以及其他地區額外的市占提升，足以抵消中國市場的下滑。

Looking back, there was clearly an inventory build in the December quarter in anticipation of Chinese provincial deployments that have now pushed out. We began to see a correction last quarter, which we expect will become more pronounced in the June quarter.

回顧來看，12 月季度顯然出現了庫存建立，係為了因應中國各省部署的預期，但這些部署如今已延後。我們在上季開始看到修正，我們預期在 6 月季度將更為明顯。

To close out my comments on our Optical and data center businesses, last quarter was a great example of the benefits of our diversification strategy and not falling into the trap of servicing the captive needs of one standard, one region, one part of the network or one transceiver supplier. We're taking share aggressively, and we believe we're uniquely positioned in the highest growth segments of our end markets.

總結我對光學與資料中心業務的評論，上季很好地展現了我們多元化策略的效益，也避免落入只服務單一標準、單一地區、網路中的單一環節或單一收發器供應商之封閉需求的陷阱。我們正積極搶占市占，並相信我們在終端市場中成長最快的區隔具備獨特的有利定位。

MACOM's ability to deliver the capacity, supply chain flexibility and pace of innovation that's necessary at the right cost structure is cementing our preeminent position as the supplier of choice across our target markets.

MACOM 以正確的成本結構提供所需的產能、供應鏈彈性與創新速度的能力，正鞏固我們在目標市場中作為首選供應商的領先地位。

Now touching briefly on A&D and MMICs. Last quarter, MMIC revenue across all of our end markets was up sequentially and grew 76% year-on-year. We continue to fill a void in the marketplace that's been left as competitors undergo consolidation. Our A&D business in total grew sequentially last quarter and was up 23.2% year-on-year. While we have healthy nonperishable backlog to realize our full growth aspirations, qualifications of certain high-rel space products are taking longer than anticipated. Again, demand in backlog remains very healthy in this business, so we expect to fully deliver EPS and shareholder value as anticipated, albeit over a slightly longer period.

接著簡要談談航太與國防（A&D）以及 MMIC。上季，我們在所有終端市場的 MMIC 營收較前一季成長，且年增 76%。隨著競爭對手進行整併，我們持續填補市場上留下的空缺。我們的 A&D 業務上季整體亦較前一季成長，且年增 23.2%。雖然我們有健康的非易腐性積壓訂單以實現完整的成長企圖，但某些高可靠度太空產品的認證所需時間比預期更長。再次強調，該業務的積壓訂單需求仍非常健康，因此我們預期仍將如期交付 EPS 與股東價值，只是時間可能會略為拉長。

With that, let's touch on GaN. We're right on schedule and squarely fixated on completing qualifications for 900 megahertz and 1.8 gigahertz customer programs in the June time frame. Again, these are mainstream programs in frequency bands servicing the heart of macrocell base station production today.

接著談談 GaN。我們完全按計畫進行，並專注於在 6 月時程內完成 900 兆赫與 1.8 吉赫客戶專案的認證。同樣地，這些是在頻段上屬於主流的專案，服務於當今宏蜂巢基地台量產的核心。

Last quarter, we broadened those engagements and today have line of sight to more than 20 additional programs, from 900 megahertz to 3.5 gigahertz, at just our first 2 lead customers, with production scheduled in the December to February time frame.

上季我們擴大了這些合作，而目前僅在我們前兩大領先客戶中，就已可望看到超過 20 個額外專案，頻率範圍從 900 兆赫到 3.5 吉赫，並排定於 12 月至 2 月期間量產。

At this juncture, our ability, and for that matter, our customers' ability to execute operationally will determine the magnitude and slope of the GaN revenue ramp at major base station customers over the coming quarters and years.

在此階段，我們的執行能力，以及同樣重要的，我們客戶的營運執行能力，將決定未來數季與數年在主要基地台客戶端 GaN 營收爬坡的規模與斜率。

Before I close, I'd like to provide a quick update on the AppliedMicro integration. We're right on track with our 100-day plan, which has gone exceptionally well on all fronts, including we've retained all key people, we've integrated all key business and operational systems and we've consummated MACOM engagements with all of their top-tier customers, notably 5 of the 7 major cloud customers.

在我結束之前，我想快速更新一下 AppliedMicro 的整合進度。我們的 100 天計畫完全按軌道推進，且各方面進展非常順利，包括我們保留了所有關鍵人才、整合了所有關鍵業務與營運系統，並完成了 MACOM 與其所有頂級客戶的合作接洽，尤其是 7 大雲端客戶中的 5 家。

We also continue to move the ball forward with our process to divest the compute business and are working to bring it to a timely conclusion.

我們也持續推進剝離運算業務的流程，並正努力使其及時完成。

I've now had the opportunity to meet personally with AppliedMicro's top customers, and I can tell you that the combined MACOM-APM opportunity is more attractive than even we had imagined. The opportunities for MACsec and PAM4 are not only real, but the transaction has elevated MACOM to a strategic status at major accounts for products and opportunities across our entire portfolio.

我現在已經有機會親自會見 AppliedMicro 的主要客戶，我可以告訴各位，MACOM 與 APM 結合後的機會，比我們原先想像的還要更具吸引力。MACsec 與 PAM4 的機會不僅確實存在，而且這筆交易也使 MACOM 在主要客戶中提升為策略性地位，涵蓋我們整體產品組合中的各項產品與機會。

To put this transaction in perspective though, AppliedMicro and PAM4 is the icing on the cake for MACOM in cloud data centers. The breakout I spoke about earlier has been in high-performance analog, lasers, and silicon photonics for standards well before and beyond single-lambda PAM4. AppliedMicro has accelerated and elevated MACOM's relevance and preeminence in everything we do, in what could be the most exciting secular growth driver in our space for many years to come.

不過，若要把這筆交易放在脈絡中來看，AppliedMicro 與 PAM4 對 MACOM 在雲端資料中心而言只是錦上添花。我先前提到的突破，發生在高效能類比、雷射與矽光子領域，這些標準早在單波長 PAM4 之前就已存在，且也延伸到其後。AppliedMicro 加速並提升了 MACOM 在我們所做一切事情上的相關性與領先地位，而這可能會成為我們產業在未來多年最令人振奮的長期結構性成長驅動力。

So wrapping it all up. Last quarter was yet another testament to our diversified business strategy across several markets and several growth drivers. We continue to ride the Optical super cycle with our Long-haul and Metro portfolio. We continue to lead the technology transition to GaN in base stations. Our MMIC products continue to fill the void created by industry consolidation. And last, but not least, we're well positioned to service insatiable demand in arguably the biggest secular growth driver of them all, connectivity within cloud data centers.

總結一下。上季再次證明了我們在多個市場與多項成長驅動因素上的多元化業務策略。我們持續受惠於光通訊超級循環，憑藉我們的長途與都會網產品組合。我們持續引領基地台向 GaN 的技術轉換。我們的 MMIC 產品持續填補產業整併所造成的空缺。最後但同樣重要的是，我們已做好準備，服務可說是所有長期結構性成長驅動因素中最大者——雲端資料中心內部連接的無止盡需求。

For these reasons, you'll understand why we remain bullish on MACOM's ability to continue to outperform as we have over the past 3 years, through good times and bad.

基於這些原因，各位會理解為何我們仍看多 MACOM 持續優於市場的能力，就像過去 3 年無論景氣好壞我們所展現的一樣。

With that, let's talk about next quarter guidance. For the fiscal third quarter ending June 30, 2017, we expect revenue to be in the range of $194 million to $198 million. Adjusted gross margin is expected to be between 58% and 61%. And adjusted earnings per share between $0.67 and $0.71 on an anticipated 66.3 million fully diluted shares outstanding.

接下來，我們談談下季財測指引。針對截至 2017 年 6 月 30 日的會計年度第三季，我們預期營收將落在 1.94 億至 1.98 億美元之間。調整後毛利率預期介於 58% 至 61%。在預估完全稀釋流通股數為 6,630 萬股的情況下，調整後每股盈餘預期介於 0.67 至 0.71 美元。

Operator, you can now open the call to questions.

接線員，現在可以開放提問。

(Operator Instructions) Our first question comes from the line of Harlan Sur of JPMorgan.

(接線員指示) 我們的第一個問題來自摩根大通（JPMorgan）的 Harlan Sur。

You delivered gross margins, they were good but they were nevertheless in the lower half of your guidance range. It seems like given the growth of the different segments, it was probably mix related but I wanted to get your views on the profile in March. And if you could just help us understand, what are the drivers that are going to take the gross margins 100 basis points higher in the June quarter?

你們交出的毛利率表現不錯，但仍落在指引區間的下半部。看起來，考量各個事業部的成長，可能與產品組合（mix）有關，但我想聽聽你們對 3 月季度的看法。另外，能否也協助我們理解，哪些因素會在 6 月季度把毛利率提高 100 個基點？

Harlan, thanks for that question. There is a mix influence to growth margin. We seized the opportunity, given the higher-margin data center business and the heavy demand in that marketplace, to really flush out more of the PowerPC-embedded revenue that was booked on a last-time buy when we acquired AMCC, and that margin brought it slightly down. So that has a backlog now that we're still working through towards the end of the year. But as that product continues to decline under the last-time buy situation, that's how we get the extra 100 basis points.

Harlan，謝謝你的問題。毛利率的確受到產品組合影響。鑑於資料中心業務毛利率較高且該市場需求強勁，我們把握機會，將我們在收購 AMCC 時取得、並在「最後一次採購」（last-time buy）時所入帳的 PowerPC 內嵌式（embedded）營收加速出貨消化，該部分毛利率使整體略為下滑。因此目前仍有積壓訂單（backlog），我們會持續處理到年底。但隨著該產品在最後一次採購情境下持續下滑，這就是我們額外取得 100 個基點的方式。

100-plus.

100 以上。

100-plus.

100 以上。

Okay. And then that's a good segue into my next question, which is as we think about the June quarter in your revenue guidance, at a high level, it would seem that the full quarter rev rec of AMCC's business is driving most of the incremental revenue growth because you still had 1/3 of the connectivity and embedded business to recognize, which would imply that core MACOM is essentially flat sequentially. Is that the way to think about it? If not, if could you just help us understand what's growing and what's not in the June quarter?

了解。接著延伸到我的下一個問題：當我們看你們 6 月季度的營收指引時，從高層次來看，似乎 AMCC 業務整季的營收認列（rev rec）推動了大部分的增量成長，因為你們仍有 1/3 的連接與內嵌式業務尚待認列，這意味著 MACOM 核心業務環比基本持平。可以這樣理解嗎？若不是，能否協助我們了解 6 月季度哪些在成長、哪些沒有成長？

Yes. So parts of the portfolio that have nothing to do with AMCC, like A&D and Multi-market, are actually growing low single digits. So that's not entirely accurate, including the Networks business, is low single digits. So I would say the one thing that we want to be very clear about, and I think I was emphatic, is there's no question that there is a cyclical weakness on the carrier base side of the business. PON remains weak, visibility is low, and you've got the China correction happening with the Metro/Long-haul provincial deployments, so there's no question that there is some headwinds in demand. At the same time, we're more than offsetting those with the cloud growth and great backlog, great order intake. And so it's really a mixed bag. So if we didn't have those headwinds in China, I think things would be materially stronger.

是的。所以，產品組合中與 AMCC 無關的部分，例如航太與國防（A&D）以及多元市場（Multi-market），其實是低個位數成長。因此那樣的說法不完全正確；包含網路（Networks）業務也都是低個位數成長。我想我們要非常清楚的一點——我也強調過——是營運商基地台端（carrier base）業務確實存在週期性疲弱，這點毫無疑問。PON 仍然疲弱、能見度低，加上中國在都會/長途的省級部署出現修正，因此需求確實有一些逆風。同時，我們以雲端成長、強勁的積壓訂單與良好的接單量，已經遠遠抵銷這些影響。因此整體是好壞參半。如果沒有中國的這些逆風，我認為表現會明顯更強。

Okay. Just my last question. As you talk about prioritization of the data center focus, maybe, John, if you could just help us understand, what segments or product areas or manufacturing capacity are you going to take resources from? Is it going to be your PON business, which is obviously kind of weak here, maybe over the near term, or your Metro/Long-haul business? Any color would be greatly appreciated.

好的。最後一個問題：你們談到把重心優先放在資料中心，John，能否協助我們理解，你們會從哪些事業部、產品領域或製造產能挪出資源？你們會從近期顯然較疲弱的 PON 業務挪出嗎，或是從都會/長途業務？任何補充說明都非常感謝。

So I don't think it has any impact on the Metro/Long-haul, but one of the bottlenecks we have short-term is availability of material, epitaxial wafers for laser production. And actually, our revenue per wafer on the data center product, 25GB, is higher than that for the 2.5GB. So it's logical in every sense of the word, strategically as well as tactically, to allocate scarce materials to data centers. So there is -- it's kind of fortuitous, frankly, that the PON business is soft. We're certainly ramping epitaxial supply. But frankly, that data center demand is way faster and way bigger than we ever anticipated. If you asked us, and you probably did ask us 90 days ago, we did not anticipate it turning on as hard as it is. Another interesting tidbit there is we actually are hiring back all the people that we had let go in Ithaca. We had downsized Ithaca to a development fab. And we find that we just can't ramp Lowell fast enough, not equipment capacity, but product qualifications. So we're actually turning Ithaca back on as a production fab to service that short-term demand. I think the other place that we bottleneck is we have an awful lot of qualifications going on right now. Part of the stuff we talked about pushing out was qualifications on high-rel space products. We've got the GaN qualifications. We've got the 25GB laser qualifications. We've got the last of the Metelics qualifications in Lowell and there's just finite resources to complete all of these new product quals. So those are some of the trade-offs. And again, we talked last quarter about stickhandling mix, and that was -- that's what this is all about.

我認為這對都會/長途沒有任何影響，但我們短期的一個瓶頸是材料供應——用於雷射生產的外延晶圓（epitaxial wafers）的可得性。而且，我們在資料中心產品（25GB）上的每片晶圓營收，高於 2.5GB 的產品。因此，無論從策略或戰術角度，把稀缺材料分配給資料中心在各方面都合情合理。坦白說，PON 業務走弱在某種程度上反而是「剛好」的。我們當然正在提升外延供應。但坦白說，資料中心需求的成長速度與規模都遠超過我們先前的預期。如果你在 90 天前問我們——你們可能也確實問過——我們並不預期它會如此強勁地啟動。另外一個有趣的小細節是，我們其實正在把先前在 Ithaca 裁撤的人員全部重新聘回。我們曾把 Ithaca 縮編成一座研發晶圓廠（development fab）。我們發現 Lowell 的擴產速度怎麼樣都不夠快，問題不在設備產能，而在產品認證（product qualifications）。因此我們正把 Ithaca 重新啟動為生產晶圓廠，以因應短期需求。我認為另一個瓶頸是目前有大量的認證在進行。我們先前提到要延後的一部分，是高可靠度太空產品的認證。我們有 GaN 的認證。我們有 25GB 雷射的認證。我們還有在 Lowell 進行的最後一批 Metelics 認證，而完成所有這些新產品認證的資源是有限的。所以這些就是一些取捨。再者，我們上季談到要靈活調度產品組合（stickhandling mix），而這——這正是其核心。

Our next question comes from the line of C.J. Muse of Evercore.

我們的下一個問題來自 Evercore 的 C.J. Muse。

I guess, first question, to revisit an earlier question, can you share with us what AMCC contributed revenue-wise? And here, I'd love to hear what was put in data center for the core business and where the PowerPC revenues resided, and then how we should think about the incremental growth in that business into June.

我想，第一個問題，回到先前的一個問題，您能和我們分享 AMCC 在營收方面的貢獻嗎？另外，我也很想了解核心業務中有多少是歸在資料中心，以及 PowerPC 的營收是歸在哪裡，然後我們應該如何看待該業務到 6 月的增量成長。

C.J., we don't break out acquisitions after they're embedded. But generally, to answer the balance of the question, the majority of the connectivity business is in our network business, some of it's in data center, some of it is in other non-Optical businesses. There is a small insignificant portion that's also classified under the embedded PC business as Multi-market.

C.J.，併購案在整合進來之後，我們不會再單獨拆分揭露。但一般來說，為了回答你問題的其餘部分，連接性（connectivity）業務的大多數是在我們的網路業務中，有一部分在資料中心，還有一部分在其他非光學（non-Optical）業務中。另外還有一小部分、影響不大，分類在嵌入式 PC 業務下的多元市場（Multi-market）。

Yes, that's product that sells into printers. Clearly, not a network application.

是的，那是銷售到印表機的產品。顯然不是網路應用。

Okay. Well, I guess, maybe to help us model and understand the gross margin driver, could you help us understand what came through in low margin in March and how we should think of that falling off in the coming Qs?

了解。那我想，或許為了幫助我們建模並理解毛利率的驅動因素，你能否協助我們了解 3 月有哪些低毛利的項目進來，以及我們應該如何看待這些在接下來幾個季度的消退？

So the revenue will probably be about the same for the next couple of quarters. It's the -- that's not the -- it's not the headwind that it is this quarter because the mix of business in the other product areas are stronger for the higher-margin products. So what we set up is to keep it consistent over the balance of the fiscal year for the embedded PC products, yet the higher gross margin data center business products mix is going to push the margin higher.

所以接下來幾個季度的營收可能大致相同。只是——這並不是——它不像本季那樣形成逆風，因為其他產品領域的業務組合更強，偏向較高毛利的產品。因此我們的規劃是，讓嵌入式 PC 產品在本財年剩餘期間維持一致，但資料中心業務中較高毛利產品的組合將會把整體毛利率往上推。

Yes, actually, if I can provide a little more color. So there is very healthy growth in the underlying connectivity business quarter-on-quarter, which is the stuff that's long-term preserved, the reason why we did the acquisition. The issue of the embedded PowerPC business is an issue of how much we want to bleed and how much dilution we can afford on the gross margin line. So as we said last quarter, and we did last quarter, in the next quarter or 2, we're going to be stickhandling that mix so that we don't dilute things too much but we get that backlog behind us.

是的，其實我可以再補充一些細節。底層的連接性業務季對季有非常健康的成長，這些是長期會保留下來的部分，也是我們進行這項併購的原因。嵌入式 PowerPC 業務的問題在於，我們願意承受多少流失，以及在毛利率這一行我們能承受多少稀釋。因此如同我們上季所說、也確實在上季所做的，在接下來 1 到 2 個季度，我們會調整（stickhandling）這個產品組合，讓稀釋不要太嚴重，同時把手上的積壓訂單（backlog）消化掉。

Got you. And I guess as my follow-up, when you talked earlier about a supply-elastic market as you service the cloud data center, I'm curious how do you think about the gross margin profile there today and over time? And as you concentrate your time, energy, investments on that business, does that change the operating profile of the model near-term?

明白。那作為追問，先前你提到在服務雲端資料中心時，市場具有供給彈性（supply-elastic）。我想了解你如何看待該業務目前以及長期的毛利率輪廓？另外，當你把時間、精力與投資更集中在該業務上時，短期內會不會改變這個模型的營運輪廓？

No. If it changed, it would actually be accretive. Data center business today, the high-performance analog stuff and the laser margins are at the upper end of our portfolio. And the reason is -- there's a subtle point here, but very important. We're talking about -- we have engineered solutions from the analog, especially the lasers and the L-PICs, which are really geared for scale, capacity and cost structure, and a lot of that is rooted in our Etched Facet Technology for the lasers. This is the same reason why we could take preeminent share in PON, arguably the most cost-sensitive market, competing with localized Chinese competitors and have gross margins that are very attractive and very accretive to our corporate average. So because we have those things engineered for cost and designed for manufacture, we can supply what is shaping up. I mean, the market for connectivity is $40 million, growing to $50 million per year, year-on-year growing -- projected to be $70 million a few years out. The real question is what's the mix of 10G versus 40 versus 100G. And my comment about supply elasticity is we're now hitting the cost per bit such that it seems like literally, every unit that we can manufacture will actually ship and drive that conversion from 10GB or 40GB up to 100GB. In fact, we are actively engaged with these customers for 400GB. So it's literally, right now, an insatiable demand. But I think one of the -- the strongest elements of MACOM is we engineer this stuff for cost. It's not repricing telecom products at low gross margins. It's a profound difference. A great question, though.

不會。如果有改變，反而會是增益（accretive）。目前資料中心業務中，高效能類比產品以及雷射的毛利率位於我們產品組合的高端。原因是——這裡有個細微但非常重要的點。我們談的是——我們從類比端、特別是雷射與 L-PIC，提供工程化解決方案，這些方案確實是為了規模、產能與成本結構而設計，而其中很大一部分根基在於我們的雷射蝕刻端面技術（Etched Facet Technology）。這也是為什麼我們能在 PON（可說是最重視成本的市場）取得領先市占，與在地中國競爭對手競爭，同時仍能維持非常吸引人、且對公司平均毛利率具有增益的毛利率。因此，因為我們把這些產品以成本為導向來工程化、並以可製造性為設計前提，我們就能供應目前正在成形的需求。我的意思是，連接性市場規模是 4,000 萬美元，年增到 5,000 萬美元，年對年成長——預估幾年後可達 7,000 萬美元。真正的問題是 10G、40G 與 100G 的組合比例。而我提到供給彈性的評論是：我們現在已經把每 bit 成本打到一個程度，似乎真的就是，我們能製造的每一個單位都會出貨，並推動從 10GB 或 40GB 升級到 100GB 的轉換。事實上，我們也正積極與這些客戶就 400GB 合作。所以目前真的是一種無止盡的需求。但我認為 MACOM 最強的一點之一，是我們以成本為導向來設計這些產品。這不是把電信產品重新定價到低毛利。這是本質上非常不同的事情。不過這確實是個很好的問題。

Our next question comes from the line of Blayne Curtis of Barclays.

我們的下一個問題來自 Barclays 的 Blayne Curtis。

Maybe just on the Optical business. John, you talked about PON being a little bit weaker. You had hoped that it would bottom. Just kind of curious, your updated thoughts as to is there still inventory, and have those other suppliers not backed out? And on the flip side, you saw strong growth in your data center business. I don't believe you're shipping 25G lasers yet, but you talked about having the fabs ramp back up. So can you just talk for the timing of bringing 25G lasers to revenue?

想請教一下光學（Optical）業務。John，你提到 PON 稍微弱一些。你原本希望它會觸底。我想了解你最新的看法：是否仍有庫存？以及其他供應商是否還沒有退出？另一方面，你們的資料中心業務成長強勁。我不認為你們已經開始出貨 25G 雷射，但你提到晶圓廠（fabs）正在重新拉升產能。所以你能談談 25G 雷射開始貢獻營收的時間點嗎？

Oh, yes, 25G is in volume. It's several million dollars of revenue, both in the form of lasers and embedded in our Optical subassemblies, in our TOSAs. So that is in production. The task before us is scaling that from low millions to many millions, so it's -- and it's analogous, Blayne, to what we did with PON when we acquired BinOptics. We had to scale that in capacity to be able to take that share. That's exactly what we're facing this year, very much an analogous task and eminently doable. I think I missed the front end of your question.

喔，有的，25G 已經在量產。營收有數百萬美元，形式包括單獨的雷射產品，以及內嵌在我們的光學次組件（Optical subassemblies）中、也就是我們的 TOSA。所以那已經在生產中。我們眼前的任務是把規模從數百萬美元提升到數千萬美元，所以——而且 Blayne，這和我們收購 BinOptics 後在 PON 所做的事情很類似。我們必須把產能擴上去，才能拿下那個市占。這正是我們今年要面對的，非常類似的任務，而且完全做得到。我想我漏聽了你問題的前半段。

Just the update on the PON market, why -- I think you had hoped that it would bottom. It sounds like it's getting worse.

只是想更新一下 PON 市場的狀況：為什麼——我記得你原本希望它會觸底。聽起來好像變得更糟了。

No, no, no. It's not just getting worse, just we haven't seen a clear recovery. And it behaves like a commodity market. So until the order intake is there and the demand is firm, I don't want to be in the business of declaring a recovery. But yes, there's no question, the inventories have been burning down. Where they burned down and how well the market -- the end market performs is where the ambiguity is. So the message I want to deliver is we're not -- we're not planning on some big recovery to be able to achieve our guidance.

不，不，不。不是變得更糟，只是我們還沒有看到明確的復甦。而且它的表現像是商品型市場。所以在接單量回來、需求變得確定之前，我不想做宣告復甦這種事。但沒錯，毫無疑問，庫存一直在去化。至於庫存去化到什麼程度，以及終端市場表現如何，這就是不確定之處。所以我想傳達的訊息是：我們不——我們不會把達成指引建立在某種大幅復甦的假設上。

Got you. And then I just wanted to ask on the A&D side, you had good 20% year-over-year growth. At one point, you were hoping for maybe even better. Just kind of maybe an update there, in terms of some of the trajectory in A&D.

了解。另外我想問一下航太與國防（A&D）方面，你們年對年有不錯的 20% 成長。之前某個時間點，你們甚至希望能更好。能否更新一下 A&D 的一些走勢與動能？

Yes. So we have backlog to be able to achieve what I had previously said, which is 40% year-on-year growth. Frankly, as we face some of these trade-offs of priorities, specifically qualifications for high-rel space and some of the Metelics stuff, we -- that's nonperishable, fungible backlog. We can push that out into subsequent quarter or quarters and focus the priorities instead on the 25GB lasers. Where the 25GB, it's the reason why I was emphasizing in the scripted remarks that short-term execution is imperative. It's going to be determining long-term shares, so pushing out some, what I would call, tactical A&D backlog that's nonperishable, is a very logical thing to do in the concept of stickhandling mix.

是的。因此，我們有積壓訂單（backlog）來達成我先前所說的目標，也就是年增 40%。坦白說，當我們面對一些優先順序取捨時，特別是高可靠度太空領域的認證，以及一些 Metelics 相關事項，我們——那是非易腐、可互換的積壓訂單。我們可以把它延後到後續一季或數季，並把優先重點改放在 25GB 雷射上。至於 25GB，這也是我在事先準備的發言中強調短期執行至關重要的原因。它將決定長期市占，因此把一些我所稱的、非易腐的戰術性 A&D 積壓訂單往後推，是在「調度產品組合（stickhandling mix）」的概念下非常合乎邏輯的作法。

Got you. And then sorry for a third one, but just in that stickhandling mix, I just want to understand Bob's comment. You expect this embedded PowerPC business will shift through the fiscal year? Or when do you envision the shipments to stop?

了解。然後抱歉第三個問題，但就這個調度組合而言，我想理解 Bob 的評論。你預期這個嵌入式 PowerPC 業務會在本財年內逐步轉移嗎？或者你預計出貨會在什麼時候停止？

Blayne, the way the order book has come in, and remember, we noted in the last conference call that there was a last-time buy issue here, so we have backlog that we can stickhandle, that it will proceed through the end of the year, and thereafter, we don't expect any more business.

Blayne，從訂單簿進來的情況來看，而且請記得我們在上次電話會議提到這裡有一次「最後一次採購（last-time buy）」的議題，所以我們有可以調度的積壓訂單，會一路延續到年底；之後我們不預期還會有更多業務。

Yes, I think the comment last quarter was we're planning to get all of it behind us in this fiscal year. So we can exit the year -- we can exit the year clean. And this was again, this was a surprise to us, with the business coming in, they executed a last-time buy before closing, and there's nothing we can do about how they manage the business between signing and closing. But the quicker we can get through that, to be honest, the earlier question about gross margin dilution is a bit frustrating for us because the rest of the business is performing quite well in that respect. And this just optically blends down our gross margin so it's a little frustrating.

是的，我想上季的評論是我們計畫在本財年把這些都處理完。所以我們可以在年底——我們可以乾淨地結束這一年。而這再次是出乎我們意料的，因為業務進來時，他們在交割前執行了一次最後一次採購，而在簽約與交割之間他們如何管理業務，我們也無能為力。但老實說，我們越快把這件事處理完越好；因為先前關於毛利率稀釋的問題對我們來說有點令人沮喪，因為其餘業務在那方面表現其實相當好。而這件事在表面上把我們的毛利率拉低，所以確實有點令人挫折。

Our next question comes from the line of Quinn Bolton of Needham.

下一個問題來自 Needham 的 Quinn Bolton。

John, I just wanted to come back to your comments about the China business. You talked about clearly seeing an inventory build in the December quarter and you're kind of working through that in March. And I think you said in your script that it might get worse in June. And I just wanted to clarify, was that comment specifically around China Long-haul and provincial network builds, or were you referring back to your PON business?

John，我想回到你對中國業務的評論。你提到你們清楚看到 12 月季度出現庫存堆積，而你們在 3 月季度正在消化。我想你在講稿中說 6 月可能會更糟。我想確認一下，那個評論是特別針對中國長途（Long-haul）與省網建設，還是你指的是你們的 PON 業務？

No, that was specifically the provincial deployments, not PON.

不，那是特別指省級部署，不是 PON。

The 100GB . . . sorry, go ahead, John.

100GB……抱歉，你先說，John。

Yes, yes. It's not PON. We're not talking about PON getting worse. PON is definitely getting better. It was, specifically, we started to see some correction at the end of last quarter in the 100GB Long-haul Metro stuff, specifically in China. Chinese. That will be more pronounced in June.

是，是的。不是 PON。我們不是在說 PON 變得更糟。PON 的確正在改善。具體來說，我們在上季末開始看到 100GB 長途/都會（Long-haul Metro）相關業務出現一些修正，特別是在中國。中國市場。這在 6 月會更為明顯。

Okay. And sort of a follow-up question on that. I think you gave the splits for Optical at $68 million and data center at $57 million. In the March quarter, it sounds like the Optical bucket, because of that headwind in China, the Optical bucket probably down, but more than offset by growth in data center. Is that the right way to be thinking about growth in the June quarter?

好的。關於這點的追問。我想你給了光通訊（Optical）為 6,800 萬美元、資料中心為 5,700 萬美元的拆分。在 3 月季度，聽起來因為中國的逆風，光通訊那一塊可能下滑，但被資料中心成長所抵銷甚至超過。那麼在思考 6 月季度的成長時，這樣理解是對的嗎？

Yes, I believe that's correct. I don't have those numbers in front of me, but that's -- the net is still positive, that's correct.

是的，我相信這是正確的。我手邊沒有那些數字，但——整體淨額仍是正向的，沒錯。

Great. Second question, you talked about the GaN on silicon going through quals now and I think you said you expect first orders in the December to February time frame. Is that the right timing then for us to be thinking about the first production revenue in that December to February time frame?

很好。第二個問題，你提到矽基 GaN 正在進行認證（quals），我想你說你預期在 12 月到 2 月的時間範圍內會有第一批訂單。那麼我們是否可以把首次量產營收也視為會在 12 月到 2 月那段時間出現？

No. Actually, sorry for the confusion in my scripted remarks. So we have the initial programs, which would generate revenue, assuming we complete the customer quals as planned, prior to that. But those are just the first 2 programs, one 900 megahertz program and 1.8 gigahertz program. What I referred to in December to February what the 20 additional programs that we have line of sight to. We have products aligned and we're delivering on customer requirements for those programs. My comment was we've got to get through the gate for these first program quals before we can secure business on the subsequent ones. So it would begin to ramp earlier than that, but when you talk about really building material revenue, you're going to get beyond the first 2 programs, and that's December to February.

不。其實，抱歉我在事先準備的發言中造成混淆。我們有最初的專案，只要我們如計畫在那之前完成客戶認證，就會帶來營收。但那只是前兩個專案，一個 900 兆赫（megahertz）專案與一個 1.8 吉赫（gigahertz）專案。我提到 12 月到 2 月的是我們可望看到的另外 20 個專案。我們已對齊產品，並依那些專案的客戶需求交付。我的意思是，在我們能夠拿下後續專案的業務之前，必須先通過這前兩個專案認證的關卡。所以爬坡會比那更早開始，但若談到真正形成具規模的實質營收，就會超過前兩個專案，而那會落在 12 月到 2 月。

Got it. And then lastly, just on the 25GB laser business. Obviously, a lot of the Datacom module vendors have internal laser production, but I think there are a number of players in the Optical module space, especially over in Asia, that don't have internal laser productions. So I'm just wondering if you can comment, are you seeing demand from both the captive modular houses that have internal laser production, or are your first shipments to some of the folks that don't have internal laser production?

了解。最後一個問題，關於 25GB 雷射業務。顯然，許多數據通訊（Datacom）模組供應商有內部雷射產能，但我想在光模組領域也有不少業者，尤其在亞洲那邊，並沒有內部雷射產能。所以我想請你評論一下：你們看到的需求是同時來自那些有內部雷射產能的自製模組廠，還是你們的首批出貨是給那些沒有內部雷射產能的業者？

So it's both. We're obviously very sensitive to not be disclosing people who also have internal supply. But the reality is, and I think over time, it will become more obvious, our Etched Facet Technology is singularly and uniquely geared for the capacity supply chain flexibility and cost structure for data centers. Taking nothing away from their laser capability, our cost structure is something that we can thrive in a market like PON, with a value proposition closer to the 100GB, 400GB level. So I think even though those people with their own captive laser supply recognize that, I would expect a transition, but it would be inappropriate for us to comment further about those customers and their decisions.

兩者都有。我們當然非常謹慎，不去揭露哪些客戶同時也有內部供應。但事實是，而且我想隨著時間推移會更明顯，我們的蝕刻端面技術（Etched Facet Technology）在資料中心的產能供應鏈彈性與成本結構方面，是獨一無二且特別契合的。並非貶低他們的雷射能力，我們的成本結構使我們能在像 PON 這樣的市場中茁壯，其價值主張更接近 100GB、400GB 的層級。所以我認為，即使那些擁有自有雷射供應的業者也認知到這點，我預期會有一個轉移過程，但我們不適合再就那些客戶及其決策做進一步評論。

Our next question comes from the line of Tore Svanberg of Stifel.

下一個問題來自 Stifel 的 Tore Svanberg。

First question is on data center. So if you just look at your portfolio now, John, with the IP portfolio, I think you mentioned in your investor presentation that you want to be #1 in 5 per data center. Could you just update us where you stand there? And is there anything else you need to invest in at this point to kind of become #1 in PHYs?

第一個問題是關於資料中心。John，如果只看你們目前的產品組合，包含 IP 產品組合，我想你在投資人簡報中提到你們希望在每個資料中心的 5 個項目中做到第 1 名。你能否更新一下你們目前的進展？以及在這個時間點，為了在 PHY 領域成為第 1 名，你們還需要再投資什麼嗎？

No. I think we're complete. To be clear, we're already #1 in the analog content, laser drivers, CDRs, TIAs and so on. I think based upon my comments I just made on the 25GB lasers, I think we're destined pretty quickly to be #1. You're talking about millions of units, not hundreds of thousands. To put in context, the 100GB Long-haul Metro data center market, even though it's growing 40% compound growth this year and next year, it's still only 250,000 growing to 500,000 units. We're talking about tens of millions of units, 2 orders of magnitude higher for these lasers. So we've got that economic advantage that's been parlayed into the L-PICs, which is the silicon photonics with the flip-chip laser so -- and the economics of that start attacking manufacturing costs, so we have that. For both CWDM4, so well before you need PAM4 or any PHY, and then you've got the icing on the cake, as I described it, which is the PAM4 single-lambda 100GB, but also the quad 400GB PAM4 and the 0 ambiguity, what the solution is for 400GB. So -- and by the way, the real subtlety is with the TOSA/ROSA capability that we picked up with FiBest, we're not just delivering spot components, we're providing components that are fully validated to people who can ramp those things in mass volume. So I would challenge people to find someone else with anything close to that portfolio and capability.

沒有。我認為我們已經完整了。講清楚一點，我們在類比內容、雷射驅動器、CDR、TIA 等方面已經是第 1 名。我想根據我剛才對 25GB 雷射的評論，我們很快就注定會成為第 1 名。你談的是數百萬個單位，而不是幾十萬。為了提供背景，100GB 長途/都會型資料中心市場，即使今年和明年以 40% 的複合成長率成長，它也只有 25 萬台，成長到 50 萬台。我們談的是數千萬台，這些雷射的量級高出 2 個數量級。因此我們具備那個經濟優勢，並已將其轉化到 L-PICs，也就是採用倒裝晶片雷射的矽光子技術——而且其經濟性開始衝擊製造成本，所以我們具備這點。對於 CWDM4 而言，在你需要 PAM4 或任何 PHY 之前很久就能做到；然後還有我所說的「錦上添花」，也就是 PAM4 單波長 100GB，但也包括四通道 400GB PAM4，以及對於 400GB 解決方案毫無歧義的答案。另外，真正微妙之處在於：透過我們從 FiBest 取得的 TOSA/ROSA 能力，我們不只是交付零散元件，而是提供已完整驗證的元件給那些能把這些產品大規模量產爬坡的客戶。所以我會挑戰大家去找另一家在產品組合與能力上有任何接近我們的公司。

That's very helpful. And moving on to China, would you be able to share with us what percentage of MACOM's business is in China for Optical? And do you have any visibility at all sort of beyond the June quarter? Because obviously, we heard a lot about the near-term weakness, but there's also some suggestions that maybe second half is going to come back. So do you have any visibility at all beyond the June quarter?

這非常有幫助。接著談中國，您能否和我們分享 MACOM 在光學業務中，有多少百分比是在中國？以及您對於 6 月季度之後是否有任何能見度？因為很明顯，我們聽到很多關於短期疲弱的說法，但也有一些暗示認為下半年可能會回來。所以您對 6 月季度之後是否有任何能見度？

Yes. So the 2 areas of weakness are PON, and there's that inventory correction. The overall market is not that ugly. It's modestly down for 2.5GB. Now 10GB is ramping, but it's just not material in 2017 compared to 2.5GB. And it's going to be a growth driver for 2018. We've got a great position there, but that's futures. But for PON, it's this inventory situation that was built up with these local Chinese competitors that are now out of business. So that will recover just as the inventory drains. The real salient issue are these provincial deployments, I mean looking back, the September and especially December quarters were like 80%, 90% year-on-year growth. In a market that's growing 40%, some would say 50%, but 40% port growth, just do the math. I mean, there was inventory build, and that's what started correcting. And I think June is the bottom of that. And if the provincial deployments move forward as expected, I think the back half of the year should be a strong recovery.

是的。所以疲弱的兩個領域是 PON，以及那裡的庫存修正。整體市場並沒有那麼糟。2.5GB 只是小幅下滑。現在 10GB 正在爬坡，但相較於 2.5GB，在 2017 年還不具規模。而它將成為 2018 年的成長驅動力。我們在那裡有很好的位置，但那是未來的事。但就 PON 而言，這是庫存狀況：先前因為一些中國本地競爭者囤了貨，而那些競爭者現在已經倒閉。因此只要庫存消化完就會復甦。真正關鍵的問題是這些省級部署；回頭看，9 月、尤其是 12 月季度的年增率大概是 80%、90%。在一個成長 40%（有人說 50%，但以 40% 的埠口成長） 的市場裡，你算一算就知道。也就是說，當時有庫存堆積，而這就是開始修正的原因。我認為 6 月是那個修正的谷底。如果省級部署如預期推進，我認為下半年應該會強勁復甦。

Very good. And just one last question on GaN. So you mentioned the 2 initial deployments and then 20 other engagements and the December to February time frame contribution potential from those. Were you referring to revenue contribution for those 20, or would that more kind of be the time frame when the quals will be completed with those 20?

很好。最後一個關於 GaN 的問題。您提到兩個初始部署，然後還有另外 20 個合作案，並且在 12 月到 2 月的時間範圍內可能會有貢獻。您指的是那 20 個案子的營收貢獻，還是比較像是那 20 個案子完成認證（qual）的時間點？

So the qual issue is much more the first customer qualifications of programs, where they qualify not just the GaN technology but the products in the Doherty designs and the whole thing. Once we get past that gate, being able to proliferate geometrically into these additional programs, you're through the gate and you're off. And we've got line of sight again. We have teams that have been servicing these base station guys for decades, literally. So we've got access, we've got all the -- customers have opened up all their requirements. We're delivering on those requirements. It's just a giant gate to get through these first quals. The second designs, the second wave there, that 20 that I referred to, becomes infinitely easier once you get past that initial gate.

所以認證（qual）這件事，更多是第一個客戶對專案的認證：他們不只是認證 GaN 技術，還包括 Doherty 架構設計中的產品以及整個方案。一旦我們跨過那道門檻，就能以幾何級數在這些額外專案中擴散；你過了門檻，就能放手去做。而且我們再次具備清楚的可視性。我們有團隊服務這些基地台客戶已經幾十年了，真的是幾十年。所以我們有接觸管道，客戶也把所有需求都攤開了。我們正在交付並滿足那些需求。只是要通過最初這些認證，是一道很大的關卡。第二套設計、第二波，也就是我提到的那 20 個，一旦跨過最初那道關卡，就會變得容易得多。

(Operator Instructions) Our next question comes from the line of Mark Delaney of Goldman Sachs.

(接線員指示) 我們的下一個問題來自高盛（Goldman Sachs）的 Mark Delaney。

My first question is going to be, I was hoping you could help us better understand some of the new sub segment classifications within Networks. So I think you said $50.6 million for data center. Can you help us understand how much of that is tied to some of these hyper-scale cloud customers, or is any of that on-premise type of data centers?

我的第一個問題是，我希望您能幫助我們更好地理解網路（Networks）內一些新的子區隔分類。所以我想您說資料中心是 5,060 萬美元。您能否幫我們理解其中有多少是與某些超大規模（hyper-scale）的雲端客戶相關，或者其中是否有任何是企業自建（on-premise）型資料中心？

It is -- that number, wait, does that include the...

那是——那個數字，等等，那是否包含……

That's the revenue...

那是營收……

That includes the revenue on the OTN framers on -- and AppliedMicro as well, right?

那包含了 OTN 成框器（framer）在——以及 AppliedMicro 的營收，對吧？

Yes, which is traditional standing data centers versus cloud.

是的，也就是傳統的既有資料中心，相對於雲端。

Yes. So I have to apologize. I don't have it off the top of my head the, what I would call, the legacy OTN framer mapper revenue coming in from AppliedMicro, but a big chunk of that right now is the analog content going into the cloud guys. And a very healthy contribution from the 25GB lasers, those in the form of standalone and TOSAs. So it's really a combination of the 2, and I apologize, I just don't have the numbers in front of me to split the 2.

是的。所以我得道歉。我一時想不起來我們所謂的、從 AppliedMicro 帶進來的傳統 OTN 成框/映射器（framer mapper）營收，但目前其中很大一部分是供應給雲端客戶的類比內容。另外 25GB 雷射也有非常健康的貢獻，形式包括獨立元件以及 TOSA。所以實際上是兩者的組合；很抱歉，我手邊沒有數字可以把兩者拆分。

I mean, I'm just trying to understand apples-to-apples the Optical piece, which I think was $67.9 million this quarter, in the December quarter, was $82.5 million. That decline from $82.5 million to $67.9 million, was any of that because of this new creation of the data center category? So apples-to-apples, the December quarter number would have been lower in Optical, or is that the -- on an apples-to-apples decline within Optical?

我的意思是，我只是想用同口徑來理解光學（Optical）這一塊；我想這季是 6,790 萬美元，而在 12 月季度是 8,250 萬美元。從 8,250 萬美元降到 6,790 萬美元的這個下滑，有沒有任何部分是因為新設立了資料中心這個類別？也就是說，用同口徑比較，12 月季度的光學數字本來就會更低，還是說——這就是光學內部同口徑的下滑？

It was not apples-to-apples. The previous number we quoted included data centers, and what we said this -- just starting this quarter, we'd separate the 2 to provide greater transparency. Next, pulling out data centers, apples-to-apples, we're actually up.

那不是同口徑。我們先前引用的數字包含資料中心，而我們說從本季開始，會把兩者分開以提供更高的透明度。接著，把資料中心拆出來後，用同口徑比較，我們其實是成長的。

Okay. And then I'm just trying to understand a little bit more on the provincial decline that the company has seen in China. I think last quarter, the commentary was the only issue of the company we're seeing was in PON, as I recall, and it spread. And I know forecasting is very difficult. We understand that in our side of the business. But I mean, could you just help us understand a bit more about what was surprising to you guys, what had you been expecting or what had your customers been telling you and why did that not materialize as you had been anticipating 90 days ago?

好的。接著我想更了解公司在中國看到的省級下滑。我記得上季的評論是，公司看到的唯一問題是在 PON，然後它擴散了。我也知道預測非常困難。我們在我們這邊的業務也理解。但我的意思是，您能否幫我們多理解一些：有哪些事情讓你們感到意外？你們原本預期什麼、或你們的客戶當時告訴你們什麼？以及為什麼這些在 90 天前你們所預期的情況沒有如期發生？

Specifically, in PON, you mean or...

具體來說，在 PON 方面，你的意思是或是……

The provincial metric, which is now weaker than, I think, the company had expected 90 days ago.

省級指標，目前比我想公司在 90 天前所預期的還要疲弱。

Yes, I think while we knew during the exuberance of the September and December quarters, I think if you go back, we were trying to provide some cautionary stuff about diversification because there was clearly an inventory build. The market was not growing at 80%, 90% year-on-year. So that is not a surprise. I think some of the timing of the provincial deployments pushing out and some of the behavior of the customers very recently, in recent weeks, that's a little bit of a surprise. And that's the -- that was -- and that's what's behind my comment about things being a little more pronounced in the June quarter. But until that provincial deployments start moving forward, we're going to be dealing with customers who are kind of locked up.

是的，我想在 9 月與 12 月季度的亢奮期間，我們其實就知道；如果你回頭看，我們一直試著就多元化提出一些審慎提醒，因為很明顯有庫存堆積。市場並沒有以年增 80%、90% 的速度成長。所以這並不令人意外。我認為，省級部署時程往後推，以及客戶在非常近期、最近幾週的一些行為，這就有點出乎意料。而那就是——那就是——也就是我評論說在 6 月季度情況會更為明顯的原因。但在省級部署開始往前推進之前，我們將會面對某種程度上被「卡住」的客戶。

Yes, Mark, the Long-haul Metro on an apples-to-apple basis did grow this quarter. It's -- so -- but we've had such substantial growth in prior quarters that it just didn't live up to the past growth expectations.

是的，Mark，以同口徑（apples-to-apple）來看，長途/都會（Long-haul Metro）這一季確實成長了。只是——所以——但我們在前幾季的成長幅度非常可觀，因此這一季就沒有達到過去的成長預期。

Hey, Mark, you know what, I was just looking through the spreadsheet I have in front of me here. It looks like the legacy part of the AppliedMicro portfolio is about 20% of that 27% total. So about 5% of our total would be legacy enterprise.

嘿，Mark，你知道嗎，我剛剛在看我面前這份試算表。看起來 AppliedMicro 產品組合中的傳統（legacy）部分約占那個 27% 總額的 20%。所以我們總體中大約 5% 會是傳統企業（legacy enterprise）。

Our next question comes from the line of Harsh Kumar of Stephens.

我們的下一個問題來自 Stephens 的 Harsh Kumar。

I was wondering if you could, John or Bob, give us a sense of how big PON and maybe Metro was in this quarter for you? You gave us good data on the data center piece because I was wondering, get the other part?

我想請問 John 或 Bob，能否讓我們了解一下本季你們的 PON 以及可能的 Metro 規模有多大？你們已經就資料中心部分提供了很好的數據，我想知道其他那一塊呢？

Yes. So PON is probably at its low level, is under 10%, down from a few points under 10% of where it was in Q1. And the second part of the question?

是的。所以 PON 可能處在低檔水準，低於 10%，相較於第一季時的水準（當時也在 10% 以下、低幾個百分點）有所下滑。那問題的第二部分是？

Was Metro the same thing?

Metro 也是一樣嗎？

Well, we don't break out Metro.

嗯，我們不會把 Metro 單獨拆分揭露。

Actually, Metro/Long-haul was up.

其實，Metro/Long-haul 是上升的。

It was up.

是上升的。

Like up. So there was the beginning in the correction in China, but our share gains outside of China actually more than offset what's going on in China.

就是上升。所以中國的修正才剛開始，但我們在中國以外的市占提升，實際上已經足以抵消中國正在發生的狀況。

Okay. But I was wondering if I'd get a percentage so we can get a good idea of like what -- how the Optical piece is made up of? Is that possible?

好的。但我想知道能不能給個百分比，這樣我們才能更清楚光學（Optical）這一塊的組成——大概是怎麼構成的？有可能嗎？

It's not a PAM, though.

但那不是 PAM（可服務市場）拆分。

Okay. I can get it later, Bob.

好的。Bob，我之後再跟你拿。

Yes, yes.

是的，是的。

Would you say that -- maybe I think, in some of the earlier calls, you had said that maybe you were expecting some initial revenues on GaN on silicon in this particular quarter, at least, the first part of the ramp. I know you said that, that's coming later, and then there's an additional piece coming in, in December and February. Is there a reason why it's getting pushed out, or is it just kind of the standard stuff that happens with this early sort of qualification stuff?

你會不會說——我想在之前幾次電話會議中，你們曾提到可能預期在本季就會看到一些矽基 GaN（GaN on silicon）的初期營收，至少是爬坡（ramp）的第一部分。我知道你說那會在之後才來，然後 12 月與 2 月還會有額外的一部分進來。請問延後的原因是什麼？還是只是這種早期認證（qualification）流程中常見會發生的標準狀況？

Well, certainly, in the last 90 days, I apologize if I miscommunicated 90 days ago, but been very consistent that the qualifications for these initial customer programs are early June. And we're still on track with those dates. So if I created a different perception, I apologize. But I mean, they're certainly delayed from what we had hoped a year or 2 ago. And it's just a long arduous process of getting locked down on these programs and completing all the intensive qualification issues. And to put a little color on that, you've got to realize for base stations, you're talking about product that spends decades in the field so reliability and qualifications are a much bigger deal than, for instance, a consumer product in a handset. But those schedules have been locked for over 90 days now.

嗯，當然，在過去 90 天內——如果我在 90 天前表達不清，我先道歉——但我們一直很一致地說，這些初期客戶專案的認證時間點是在 6 月初。而我們仍然按這些日期在推進。所以如果我造成了不同的認知，我很抱歉。但我的意思是，這些確實比我們在一、兩年前所希望的時程更晚。而這只是把這些專案敲定、並完成所有密集認證議題的一個漫長而艱鉅的流程。再補充一點背景，你必須了解，基地台產品是在現場運行數十年的，所以可靠度與認證的重要性遠高於例如手機等消費性產品。但這些時程在 90 多天前就已經鎖定了。

I got you. And for a last one, I know there was a compute piece that was up for sale and I know you are working on your 100-day plan. Do you, John and Bob, feel reasonably comfortable about the potential for maybe possibly selling that business?

了解。最後一個問題，我知道有一個運算（compute）業務正在出售中，也知道你們在做 100 天計畫。John 和 Bob，你們是否相當有把握這個業務有可能被出售？

Oh, jeez, yes. I mean, the real more salient issue is to whom, with what structure of transaction, right? We're really stuck that we really can't make public comments at this time because we've got a process going and telegraphing the type of interest and who and at what valuation would be so inappropriate. But we are -- we've had our heads down working that process from the day it closed. We're right on track from the schedules that were locked down there. We just reviewed options with our Board, and now we're moving to the next phase, the next round. And those who were still in the process are being informed.

噢，天啊，是的。我的意思是，更關鍵的問題其實是賣給誰、交易架構是什麼，對吧？我們真的受限於目前不能公開評論，因為我們正在進行一個流程；去透露利害關係的類型、對象是誰、以及在什麼估值水準，都是非常不恰當的。但我們——自交易完成那天起就埋頭在推進這個流程。我們完全按照當時鎖定的時程在走。我們剛和董事會檢視了各種選項，現在正進入下一階段、下一輪。仍在流程中的各方也正在被告知。

Harsh, I have your answer. So Long-haul Optical-data center interconnect this quarter was 26% compared to 28% last quarter.

Harsh，我有答案了。所以本季長途光學—資料中心互連（Long-haul Optical-data center interconnect）占比為 26%，相較上一季為 28%。

Our next question comes from the line of Richard Shannon of Craig-Hallum.

我們的下一個問題來自 Craig-Hallum 的 Richard Shannon。

I guess a few for me, and I'll start in the data center space. For the June quarter, John, are you expecting material growth in L-PIC, or is that coming after June?

我這邊有幾個問題，我先從資料中心領域開始。John，對於 6 月季度，你預期 L-PIC 會有顯著成長嗎？還是要等到 6 月之後？

So it will be after June. So we have production units and we're completing the designs. I think in the December quarter would be where we would start seeing the first volume of production, but we have the production devices completing qualification. Qualification is not a big issue as it is with GaN. But the customer designs and ramping into production is just the phase that we're in right now.

所以會在 6 月之後。我們已有量產單元（production units），並且正在完成設計。我認為在 12 月季度我們會開始看到第一批量產出貨量，但我們的量產器件正在完成認證。認證不像 GaN 那樣是個大問題。但客戶端設計以及導入量產的爬坡，正是我們目前所處的階段。

Okay. And when you talk about imperative in the short term to execute on the data center side, is that specific to L-PIC or more broadly across data center customers? And I'm curious why that's the case. Is there some specific window you're trying to hit here, or were you just shifting your strategic focus to that area?

好的。另外，當你談到短期內在資料中心端執行落地是當務之急時，這是特別針對 L-PIC，還是更廣泛地針對資料中心客戶？我也好奇為什麼會這樣。你們是想要趕上一個特定的時間窗口，還是只是把策略重心轉向那個領域？

Yes, we are absolutely shifting strategic focus. I mean, let me give you the context and you can do the math. So in the Long-haul Metro market, which is fantastic, it's growing 40% compound growth from around 250,000, 255,000 units in '16, up to around 350,000 to on the path towards close to 500,000 over the 2-year period. It's great and wonderful. To put in context, the unit port market, the number of ports in data centers, is 40 million going to 50 million, going to 70 million. That's the TAM. Currently, it's 10G, and we're converting 10G and 40G up to 100G. And so you do the math on the content, and it is a multibillion-dollar opportunity for us. It's analogous to the magnitude of handsets and smartphones in the compound semi space. It is a big opportunity. We're very well-positioned. And the first thing that comes to mind, it's execution on everything, including by the way, the PAM4 platform that AppliedMicro comes in with. But the first thing that is paramount is ramping 25GB laser production. It is in production. There was very material multiple millions of dollars of revenue in 25GB lasers, enabled revenue in TOSAs. But ramping to the tens of millions of units as we did for PON, I mean, we're shipping, at peak, 5 million to 8 million units per month of 2.5G PON. We need to do the same thing with 25GB. And we're right on track to do that, and it's right within our wheelhouse. It's just there's a lot of execution to be able to ramp that and get the yields up to where they need to be, to be able to deliver -- when we do that, we uniquely have the capacity, supply chain flexibility, and we are -- play a very strategic focus then -- role then in the industry for those cloud data center guys. The optical industry is not geared to supply that magnitude of units, that magnitude of ports. We have the HPA business, the previous Mindspeed business in Newport Beach that's been -- that has done that. They are in a lead position in the industry as we roll out the lasers and then layer the lasers literally on top of the L-PICs. That's when we stand alone. And then the PAM4 stuff layers in, there's just a lot of operational execution to be able to ramp from what we've done today with analog, which is 1.6 million units in the first half of the year, to tens of billions of units across these products. There's a lot of execution. But if you do the math, you'll see the reason why it is -- unquestionably, it dwarfs all else that we do.

是的，我們絕對正在轉移策略重心。我的意思是，讓我先提供背景，您可以自己算一算。因此，在長距離都會（Long-haul Metro）市場，這個市場非常棒，正以 40% 的複合成長率成長，從 2016 年約 250,000、255,000 台，成長到約 350,000 台，並在 2 年期間朝接近 500,000 台的路徑前進。這很棒、也很令人振奮。放在脈絡中看，單位埠（unit port）市場，也就是資料中心的埠數，從 4,000 萬增加到 5,000 萬，再到 7,000 萬。那就是 TAM（總可服務市場）。目前是 10G，而我們正把 10G 與 40G 轉換升級到 100G。所以您把內容價值算一算，對我們而言這是一個數十億美元規模的機會。其量級類似於複合半導體領域中手機與智慧型手機的規模。這是一個巨大的機會。我們的布局非常到位。首先想到的是在所有事情上的執行力——順帶一提，也包括 AppliedMicro 帶來的 PAM4 平台。但最關鍵的第一件事，是把 25GB 雷射的產能爬坡。它已在量產中。25GB 雷射已帶來非常可觀、數百萬美元等級的營收，並在 TOSA 中帶動營收。但要像我們在 PON 所做的那樣，把出貨爬坡到數千萬顆的量級——我的意思是，我們在 2.5G PON 的高峰期每月出貨 500 萬到 800 萬顆。我們需要在 25GB 上做到同樣的事。而我們正按計畫推進，這也完全在我們的專長範圍內。只是要把產能拉上來並把良率提升到所需水準，需要大量的執行工作；一旦做到，我們將具備獨特的產能、供應鏈彈性，並且——屆時我們將在產業中扮演非常具策略重心的角色——為那些雲端資料中心客戶服務。光學產業並未為供應如此規模的出貨量、如此規模的埠數而設計。我們有 HPA 業務，也就是先前位於 Newport Beach 的 Mindspeed 業務——他們一直在做這件事。當我們推出雷射，並把雷射真正地疊加在 L-PICs 之上時，他們在產業中處於領先地位。那時我們就會是獨一無二的。接著再把 PAM4 的部分疊加進來；要從我們今天在類比產品上所做到的規模——上半年 160 萬顆——爬坡到這些產品線合計數百億顆的規模，需要大量的營運執行。需要大量的執行。但如果您算一算，就會明白為什麼——毫無疑問，它的規模遠遠超過我們所做的其他一切。

Okay. I appreciate that perspective, John. That makes a lot of sense. Last quick question for me. You talked about, within the Optical carrier business, some share gains in China offsetting a little bit of weakness there. Can you characterize where those share gains are coming from?

好的。John，我很感謝這個觀點。這非常合理。我最後一個簡短問題。你提到在光學電信商（Optical carrier）業務中，中國的一些市占提升抵消了那邊一點疲弱。你能否說明這些市占提升主要來自哪裡？

Actually, the share gain is outside of China. Yes. And it's North America and Japan. From day 1, we've had a very strong position with certain Japanese customers. And there's an insatiable demand that they have servicing Metro and DCI. This is DCI-data center interconnect, not within data centers but still coherent. And I mean, we're supply-constrained in that space as bizarre as it sounds because some of the product is different than that for China.

其實，市占提升是在中國以外。是的。主要在北美和日本。從第一天起，我們在某些日本客戶那裡就有非常強的地位。他們在都會網與 DCI 方面有著近乎無止境的需求。這裡的 DCI 是資料中心互連（data center interconnect），不是資料中心內部，但仍然是相干（coherent）。而且，聽起來很奇怪，但我們在那個領域反而受限於供應，因為部分產品與供應中國的產品不同。

That is all the time we have for questions today. I'd like to hand the call back over to Mr. Croteau for any closing remarks.

今天的提問時間就到這裡。我想把電話交回給 Croteau 先生，請他做結語。

Very good. So before closing out today's call, I want to mention that we're going to be hosting a number of investor events over the coming months, including analyst forums in Boston and New York, where we will be focused on our cloud data center opportunity in more detail. That will be during the week of May 15. Stay tuned for additional details. We'll also be attending a number of investor conferences during the quarter, including the JPMorgan Conference in Boston, May 23; the Craig-Hallum Conference in Minneapolis, May 31; and then 2 events in San Francisco, the Stifel Tech Internet and Media Conference on June 5; and the Bank of America Merrill Lynch Global Tech Conference on June 6; followed by 2 events in New York, the Stephens Spring Conference on June 7; and the Citi Conference on June 8. If you'd like to request a meeting at one of these upcoming events, please email us at ir@macom.com or contact your sales representative directly. We look forward to providing an update and reporting on our continued progress on next quarter's call. That concludes today's remarks. Operator, you may now disconnect the call.

很好。在結束今天的電話會議之前，我想提到，未來幾個月我們將舉辦多場投資人活動，包括在波士頓與紐約的分析師論壇，我們將更詳細聚焦於雲端資料中心機會。活動將在 5 月 15 日當週進行。更多細節敬請期待。本季我們也將參加多場投資人會議，包括：5 月 23 日在波士頓的摩根大通（JPMorgan）會議；5 月 31 日在明尼阿波利斯的 Craig-Hallum 會議；接著在舊金山有 2 場活動，6 月 5 日的 Stifel 科技、網際網路與媒體會議，以及 6 月 6 日的美國銀行美林（Bank of America Merrill Lynch）全球科技會議；之後在紐約還有 2 場活動，6 月 7 日的 Stephens 春季會議，以及 6 月 8 日的 Citi 會議。若您希望在上述即將到來的活動中安排會面，請寄信至 ir@macom.com，或直接聯繫您的業務代表。我們期待在下季電話會議中提供最新進展並報告我們持續的推進成果。以上為今天的結語。接線員，您現在可以中斷通話。

Ladies and gentlemen, thank you for participating in today's conference. That does conclude today's program. You may all disconnect. Everyone, have a great day.

各位女士、先生，感謝您參與今天的會議。今天的議程到此結束。各位現在可以掛線。祝各位有美好的一天。