MACOM Technology Solutions Holdings, Inc. (MTSI) 2019 Q2 法說會逐字稿

Good afternoon, and welcome to MACOM's Second Fiscal Quarter 2019 Financial Results Conference Call. (Operator Instructions) As a reminder, this conference call is being recorded today, Tuesday, May 7, 2019. I will now turn the call to Steve Ferranti, Vice President of Investor Relations at MACOM. Steve, please go ahead.

下午好，歡迎參加 MACOM 2019 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（2019 年 5 月 7 日，星期二）錄音。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, Shannon. Good afternoon, everyone, and welcome to MACOM's Second Fiscal Quarter 2019 Earnings Conference Call. Joining me today are MACOM's President and Chief Executive Officer, John Croteau; and Principal Financial Officer, Conor Hegarty.

謝謝你，Shannon。各位下午好，歡迎參加 MACOM 2019 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，以及首席財務主管 Conor Hegarty。

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

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

Before turning the call over to John, I'd 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，而此類陳述受若干風險與不確定性影響。由於實際結果可能與今日討論內容存在重大差異，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條款之保護。

For more detailed discussion of the risks and uncertainties that could result in those differences, we refer you to MACOM's filings with the SEC, including its current report on Form 8-K filed today, its annual report on Form 10-K for fiscal year 2018 filed on November 16, 2018, and its quarterly report on Form 10-Q for the fiscal quarter ended December 28, 2018, filed on February 6, 2019.

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、於 2018 年 11 月 16 日提交之 2018 會計年度 Form 10-K 年度報告，以及於 2019 年 2 月 6 日提交之截至 2018 年 12 月 28 日止會計季度的 Form 10-Q 季度報告。

Any forward-looking statements represent management's views only as of today, May 7, 2019, and MACOM assumes no obligation to update these statements in the future. 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 referred to on today's call unless otherwise noted.

任何前瞻性陳述僅代表管理層截至今日（2019 年 5 月 7 日）的觀點，MACOM 不承擔未來更新此等陳述之任何義務。公司新聞稿以及管理層在本次電話會議中的發言，將包含對若干經調整之非 GAAP 指標與財務資訊的討論；除非另有說明，今日電話會議中提及之所有損益表金額與百分比均為此類經調整之非 GAAP 數據。

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 at 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「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，我將把電話交給 John，請他就本季發表評論。

Thanks, Steve. Welcome, everyone, and thanks for joining us today. I'll begin today's call with an overview of our second quarter results for fiscal 2019. I'll then turn the call over to Conor Hegarty who will review our financial performance in further detail. I'll conclude today's prepared remarks by providing an update on our secular growth drivers and cloud data centers, 5G Telecom and global and homeland security including a more in depth discussion of our pending joint venture with Goertek to service the China 5G opportunity. And finally, I'll close the call with guidance for the third fiscal quarter of 2019.

謝謝，Steve。歡迎各位，感謝今天加入我們。我將先概述我們 2019 會計年度第二季的業績表現。接著我會把電話交給 Conor Hegarty，由他更詳細地回顧我們的財務表現。我將在事先準備發言的最後，更新我們的長期結構性成長動能，以及雲端資料中心、5G 電信與全球及國土安全等領域的進展，並更深入討論我們與歌爾股份（Goertek）擬議中的合資企業，以服務中國 5G 機會。最後，我將提供 2019 會計年度第三季的財測指引。

Jumping right into the numbers. Consistent with our preliminary Q2 results that we provided on April 24, adjusted revenue for the fiscal second quarter was $121.5 million, down 19% sequentially driven largely by broad-based weakness among our data center businesses. Adjusted gross margin was 48.7% including an inventory charge of roughly $8 million that accounted for 600 basis points of negative gross margin impacts. Adjusted earnings per share was a loss of $0.18. That said, free cash flow was positive $13 million, and our cash balance was up $6 million sequentially. This was obviously a very challenging quarter. On top of the forecasted softness in China and seasonality in our I&D business, we had a number of unexpected headwinds, which were rooted in the Data Center inventory correction that we discussed in our last earnings call.

直接進入數字。與我們在 4 月 24 日提供的第二季初步結果一致，2019 會計年度第二季經調整營收為 1.215 億美元，較前一季下滑 19%，主要原因是我們資料中心相關業務普遍疲弱。經調整毛利率為 48.7%，其中包含約 800 萬美元的存貨費用，對毛利率造成 600 個基點的負面影響。經調整每股盈餘為虧損 0.18 美元。儘管如此，自由現金流為正 1,300 萬美元，我們的現金餘額較前一季增加 600 萬美元。這顯然是一個非常具挑戰性的季度。除了我們原先預期的中國市場疲弱與 I&D 業務的季節性因素之外，我們還遭遇多項未預期的逆風，其根源在於我們在上次財報電話會議中討論的資料中心存貨修正。

For much of 2018, the industry had faced a supply shortage as module suppliers struggled to keep up with demand for major cloud deployments. Service providers had put sustained pressure on their suppliers, our customers, to expedite shipments during this period.

在 2018 年的大部分時間裡，產業面臨供應短缺，因為模組供應商難以跟上主要雲端部署的需求。在此期間，服務供應商持續對其供應商（也就是我們的客戶）施加壓力，要求加速出貨。

Earlier this year, those cloud operators clamped down on new orders in order to digest inventory. This has been well reported by our customers and peers across the industry. This hard stop in orders following a prolonged period of hyper growth created a whipsaw effect that has reverberated throughout the supply chain over the past few months. To start, inventories of our data center component stacked up at module vendors to levels well beyond what we anticipated 90 days ago. This resulted in a shortfall in product revenue as compared to our original Q2 guidance.

今年稍早，這些雲端營運商為了消化庫存而緊縮新訂單。這點已由我們的客戶以及同業在整個產業中廣泛報導。在長時間的高速成長之後，訂單突然急停，造成「鞭甩效應」，並在過去幾個月於供應鏈中持續擴散。首先，我們的資料中心元件庫存在模組供應商端堆積至遠高於我們 90 天前所預期的水準。因此，產品營收相較於我們原先的第二季指引出現缺口。

Secondly, cloud operators deprioritized the qualification and ramp of new module suppliers creating uncertain prospects for our white box solutions customers. New entrants usually take the brunt of market downturns as incumbents leverage existing supply agreements to wind down inventory levels more smoothly. We do the exact same thing in product areas where we hold an incumbent position. In light of these market conditions and soft backlog, we believe it prudent to record inventory reserves on certain data center products, which we're in the process of ramping. This resulted in a charge to cost of goods sold of more than $8 million, which accounted for a $0.12 hit to EPS and 600 basis points of negative gross margin impacts. Excluding this unanticipated accounting adjustment, gross margin would have been close to the low end of our guidance range.

其次，雲端營運商降低了對新模組供應商的認證與放量優先順序，使我們白牌解決方案客戶的前景變得不確定。在市場下行時，新進者通常承受最大衝擊，因為既有供應商可利用既有供應協議，更平順地降低庫存水位。在我們具備既有供應商地位的產品領域，我們也會採取完全相同的做法。鑑於這些市場狀況與疲弱的訂單能見度，我們認為對部分正在放量的資料中心產品提列存貨備抵是審慎之舉。這導致銷貨成本增加超過 800 萬美元的費用，對每股盈餘造成 0.12 美元的衝擊，並對毛利率造成 600 個基點的負面影響。若排除此一未預期的會計調整，毛利率將接近我們指引區間的低端。

It's worth noting that we did bring on a new solutions customers during the quarter. We collected the first tranche of cash associated with the initial milestones but did not recognize revenue as originally anticipated due to current market uncertainties. We may or may not be able to recognize revenue at some point in the future so we're not factoring that into our Q3 guidance. I'll get into this further at the end of my prepared remarks. The magnitude of this correction has been deeper and far more disruptive than we originally anticipated. However, we believe we're at the bottom of the cycle, and our customers remain optimistic for a strong rebound in the second half of the calendar year.

值得一提的是，我們在本季確實新增了一位解決方案客戶。我們已收取與初始里程碑相關的第一筆現金款項，但由於目前市場不確定性，未如原先預期認列營收。我們未來可能會、也可能不會在某個時間點認列該營收，因此我們未將其納入第三季指引。我會在事先準備發言的最後進一步說明。此次修正的幅度比我們原先預期更深，且干擾程度也大得多。然而，我們相信我們已處於景氣循環的底部，而我們的客戶仍對下半年出現強勁反彈抱持樂觀。

All of the underlying demand drivers for hyperscale connectivity remain in place. Customer and market analyst forecasts call for a pickup in port shipments in the second half of 2019 and again strong growth in 2020. I'll talk more about this later in the call along with our secular growth drivers.

超大規模（hyperscale）連接性需求的所有基本驅動因素仍然存在。客戶與市場分析師的預測指出，2019 年下半年連接埠（port）出貨量將回升，且 2020 年將再次呈現強勁成長。我稍後也會在電話會議中，連同我們的長期結構性成長動能，一併進一步說明。

Before I turn it over to Conor for a more in-depth review of our fiscal second quarter financials, I'd like to take a moment to thank Bob McMullan for his service at MACOM. For the past 5 years, Bob was instrumental in rebuilding and reshaping our portfolio through numerous acquisitions, divestitures and corporate financing transactions. We all wish Bob continued success in his future endeavors.

在我把時間交給 Conor，請他更深入檢視我們會計年度第二季的財務表現之前，我想先花點時間感謝 Bob McMullan 在 MACOM 的服務。在過去 5 年裡，Bob 透過多項併購、資產剝離以及企業融資交易，在重建並重塑我們的產品組合方面扮演了關鍵角色。我們都祝福 Bob 在未來的發展一切順利、持續成功。

With that, I'll let Conor go through the financials. Take it away, Conor.

接下來，我就請 Conor 帶大家看財務數字。交給你了，Conor。

Thank you, John. Good afternoon, everyone.

謝謝你，John。各位下午好。

Adjusted revenue in fiscal Q2 was $121.5 million, down 19% year-over-year from $150.4 million. I'd like to highlight that our fiscal Q2 GAAP revenue includes $7 million associated with our data center solutions business. This amount was previously included in our non-GAAP revenue in the third fiscal quarter of 2018. Accordingly, this $7 million amount has been excluded from our fiscal Q2 2019 non-GAAP revenue. The revenue amounts and any associated percentages presented and discussed in our call today relate to non-GAAP revenue also referred to as adjusted revenue.

會計年度第二季的調整後營收為 1.215 億美元，較去年同期的 1.504 億美元下滑 19%。我想特別指出，我們會計年度第二季的 GAAP 營收包含與資料中心解決方案業務相關的 700 萬美元。這筆金額先前在 2018 會計年度第三季被納入我們的非 GAAP 營收。因此，這 700 萬美元已自我們 2019 會計年度第二季的非 GAAP 營收中排除。我們今天電話會議中所呈現與討論的營收金額及任何相關百分比，皆以非 GAAP 營收為準，亦稱為調整後營收。

Revenue by end markets. Industrial & Defense was $50 million and 42% of total revenue, up 15% year-over-year and in line with our forecast. Telecom was $47 million and 39% of total revenue, down 26% year-over-year, but in line with our forecast. As a reminder, fiscal Q2 last year included LR4 revenue of $12 million.

按終端市場劃分的營收。工業與國防為 5,000 萬美元，占總營收 42%，年增 15%，並符合我們的預測。電信為 4,700 萬美元，占總營收 39%，年減 26%，但符合我們的預測。提醒一下，去年會計年度第二季包含 1,200 萬美元的 LR4 營收。

Data center was $24 million and 19% of total revenue, down 44% year-over-year. This highlights the magnitude of the data center inventory correction.

資料中心為 2,400 萬美元，占總營收 19%，年減 44%。這凸顯了資料中心庫存修正的幅度。

Adjusted gross profit and adjusted gross margin in fiscal Q2 were $59.1 million and 48.7% of revenue, respectively, which includes $8 million in inventory reserves primarily associated with data center materials or roughly 600 basis points of gross margin impact. Total adjusted operating expenses were $63.3 million composed of R&D expenses of $40.4 million and SG&A expenses of $22.9 million. Total adjusted operating expenses were up approximately $500,000 or 1% sequentially.

會計年度第二季的調整後毛利與調整後毛利率分別為 5,910 萬美元與營收的 48.7%，其中包含 800 萬美元的存貨備抵，主要與資料中心材料相關，約造成毛利率影響 600 個基點。調整後營業費用合計為 6,330 萬美元，其中研發費用 4,040 萬美元，銷售、一般及管理費用（SG&A）2,290 萬美元。調整後營業費用合計較前一季約增加 50 萬美元，或 1%。

Adjusted loss from operations, $4.2 million, translating into negative 3% operating margin. Adjusted net interest expense was $8.4 million, up $600,000 sequentially, primarily due to approximately $400,000 of charges associated with leased equipment, which is not expected to continue in future periods. Our normalized adjusted income tax rate in fiscal Q2 continued at 8%.

調整後營業損失為 420 萬美元，換算為負 3% 的營業利益率。調整後淨利息費用為 840 萬美元，較前一季增加 60 萬美元，主要因約 40 萬美元與租賃設備相關的費用所致，預期未來期間不會持續發生。我們會計年度第二季的常態化調整後所得稅率維持在 8%。

Fiscal Q2 adjusted net loss was $11.6 million, translating into negative $0.18 per fully diluted share. The share count used to calculate fiscal Q2 adjusted EPS was 66 million fully diluted shares.

會計年度第二季調整後淨損為 1,160 萬美元，換算為每股完全稀釋後虧損 0.18 美元。用於計算會計年度第二季調整後每股盈餘（EPS）的股數為 6,600 萬股完全稀釋後股數。

Adjusted EBITDA, or earnings before interest, taxes, depreciation and amortization, was $3 million.

調整後 EBITDA（息稅折舊攤銷前盈餘）為 300 萬美元。

Moving on to cash flow. GAAP cash flow from operations was $24 million in fiscal Q2 aided by reductions in net working capital of $26 million. Capital expenditures in fiscal Q2 were $11 million or 9% of revenue, and depreciation expense was $8 million. Free cash flow was $13 million, up from negative $9 million sequentially due primarily to the improvements in working capital amounts.

接著談現金流。會計年度第二季 GAAP 營運活動現金流為 2,400 萬美元，主要受惠於淨營運資金減少 2,600 萬美元。會計年度第二季資本支出為 1,100 萬美元，約占營收 9%，折舊費用為 800 萬美元。自由現金流為 1,300 萬美元，較前一季的負 900 萬美元增加，主要因營運資金項目改善所致。

Cash, cash equivalents and short-term investments were $192 million, up from $186 million sequentially.

現金、約當現金及短期投資為 1.92 億美元，較前一季的 1.86 億美元增加。

Accounts receivable were $87 million. Days sales outstanding were 66 days.

應收帳款為 8,700 萬美元。應收帳款週轉天數（DSO）為 66 天。

Inventories were $120 million. Inventory turns were 2.1x, down from 2.2x sequentially. Long-term debt was $686 million inclusive of capital leases.

存貨為 1.20 億美元。存貨週轉率為 2.1 倍，較前一季的 2.2 倍下降。長期負債為 6.86 億美元，包含資本租賃。

Long-term debt of $657 million is termed with minimal annual principal repayment until maturity in May 2024 and covenant light. MACOM has an undrawn $160 million credit line available through November 20, 2021. Back to you, John.

其中 6.57 億美元的長期負債為定期貸款，直到 2024 年 5 月到期前，每年本金償還額很少，且契約限制較少（covenant light）。MACOM 有一筆 1.60 億美元的循環信用額度尚未動用，可用至 2021 年 11 月 20 日。交回給你，John。

Thanks, Conor.

謝謝，Conor。

For the remainder of my prepared remarks, I'll provide an update on our secular growth drivers and why, despite all of the near term ugliness, I'm as bullish as ever on the opportunity for MACOM that lies ahead. Across our opportunity set, we're better positioned today than we were just 90 days ago.

在我接下來的講稿中，我將更新我們的長期結構性成長動能，並說明為何儘管短期情況相當難看，我對 MACOM 未來的機會仍一如既往地樂觀。在我們的整體機會版圖中，我們今天的定位比 90 天前更好。

Starting with cloud data centers. The current inventory correction does not alter the fundamental need for high-speed connectivity in data centers. Cloud operators' revenue streams and future growth initiatives are dependent on ultra high-speed data links between server racks and across their data centers.

先從雲端資料中心談起。目前的庫存修正並不會改變資料中心對高速連結的根本需求。雲端營運商的營收來源與未來成長計畫，仰賴於伺服器機櫃之間以及跨資料中心的超高速資料連結。

As a prime example of this, last quarter, Facebook unveiled its next-generation hyperscale topology called F-16 to cope with massive growth of data traffic inside of its data centers. F-16 is designed to provide 1.6 terabits of bandwidth at the top of the rack switch based on 100 gig CWDM building blocks. This architectural shift is expected to drive a 4x increase in the number of 100 gig CWDM ports at Facebook while also providing a migration path to 200 and 400 gig in the future. We expect other cloud customers to follow a similar path. This is an excellent example of why we continue to believe that 100 gig has a long life cycle ahead. Customer forecasts bear this out, predicting strong growth for both 100 gig CWDM4 and PAM4 in the remainder of 2019 and 2020 while 200 and 400 gig ramp on top of that demand starting next year.

一個最具代表性的例子是，上季 Facebook 發表了其下一代超大規模（hyperscale）拓撲架構 F-16，以因應其資料中心內部資料流量的爆炸性成長。F-16 的設計目標是在機櫃頂端交換器（top-of-rack switch）提供 1.6 太位元的頻寬，並以 100G CWDM 的基本模組為基礎。預期這項架構轉變將使 Facebook 的 100G CWDM 連接埠數量增加 4 倍，同時也提供未來升級至 200G 與 400G 的遷移路徑。我們預期其他雲端客戶也會走類似的路徑。這是我們持續相信 100G 仍有很長產品生命週期的絕佳例證。客戶預測也印證了這一點：預期 2019 年剩餘期間與 2020 年，100G CWDM4 與 PAM4 都將強勁成長，而 200G 與 400G 將在明年起在此需求之上開始放量。

MACOM remains extremely well positioned to capitalize on all of these trends. This year's Optical Fiber Conference in San Diego was a watershed event in terms of highlighting the leadership and production readiness of our broad portfolio of analog, DSP, laser and L-PIC solutions for the cloud.

MACOM 仍然具備極佳的有利位置，可把握所有這些趨勢。今年在聖地牙哥舉辦的光纖通訊大會（Optical Fiber Conference, OFC）是一個分水嶺事件，突顯了我們面向雲端的類比、DSP、雷射與 L-PIC 解決方案廣泛產品組合的領先地位與量產就緒度。

Our lead PAM4 DSP customer finally unveiled their new 100 gig PAM4 transceiver at OFC, center stage in their booth, which was running live data traffic over a 10-kilometer link using our DSP chip.

我們的主要 PAM4 DSP 客戶終於在 OFC 上揭露其新的 100G PAM4 收發器，並在其攤位的核心位置展示，使用我們的 DSP 晶片在 10 公里鏈路上進行即時資料傳輸。

Those who attended the conference and had the chance to see that demo can now appreciate the significance that customer ascribes to that product for their overall cloud data center strategy. This is a really big deal for them, and it's a really big deal for the rest of the industry too. This is the first of several customers that are poised to enter production using our DSP.

參加會議並有機會看到該展示的人，現在可以理解該客戶對該產品在其整體雲端資料中心策略中的重要性評價。這對他們而言是非常重大的里程碑，對整個產業其他業者而言也同樣意義重大。這是多家客戶之中的第一家，這些客戶已蓄勢待發，準備採用我們的 DSP 進入量產。

Broadly speaking, DSP-based PAM4s gain traction within those cloud operators with fiber infrastructure that's rooted in PSM4. Initial 400 gig traction has been in the form of DR4, which much like PSM, breaks out 4 single wavelength streams over 4 fibers to 4 100 gig PAM4 transceivers. That means 4 100 gig transceivers will ship for each 400 gig transceiver, another reason we believe 100 gig has a long life cycle ahead.

概括而言，基於 DSP 的 PAM4 在那些光纖基礎設施以 PSM4 為根基的雲端營運商之間正逐步獲得採用。初期的 400G 動能主要以 DR4 的形式呈現；其與 PSM 類似，透過 4 條光纖分拆出 4 路單一波長的資料流，連接至 4 個 100G PAM4 收發器。這意味著每出貨 1 個 400G 收發器，就會出貨 4 個 100G 收發器；這也是我們認為 100G 在未來仍將擁有長生命週期的另一個原因。

Given our deep roots and architectural partnerships with key players in the data center industry, we believe our PAM4 DSP is poised to achieve a preeminent share in those 100 gig DR1 transceivers. That's why we're laser focused ramping production of our flagship PRISM DSP.

憑藉我們在資料中心產業的深厚根基，以及與關鍵業者的架構層級合作夥伴關係，我們相信我們的 PAM4 DSP 有望在那些 100G DR1 收發器中取得領先的市占率。這也是為什麼我們全力聚焦於加速量產我們的旗艦 PRISM DSP。

While many people equate PAM4 with DSPs, certain cloud operators are sponsoring a whole new dimension to their 200 and 400 gig strategies, Analog, CDR-based PAM4. This approach provides a smooth upgrade path to 200 and 400 gig for those with CWDM-based fiber infrastructure. CWDM multiplexes 4 wavelengths over a single fiber. This initiative plays directly to our strengths and incumbency in high-performance analog components.

雖然許多人將 PAM4 與 DSP 劃上等號，但某些雲端營運商正在為其 200G 與 400G 策略推動一個全新的面向：基於類比、採用 CDR 的 PAM4。對於採用 CWDM 架構光纖基礎設施的業者而言，這種方法可提供平順的升級路徑至 200G 與 400G。CWDM 會在單一光纖上多工 4 個波長。這項倡議正好直接契合我們在高效能類比元件方面的優勢與既有領先地位。

Earlier today, we announced the formation of the Open Eye MSA Group, an industry consortium founded by MACOM along with Semtech with participating members including Huawei's HiSilicon, Juniper, Innolight, AOI and Mellanox among nearly a dozen others. A multisource agreement aims to come up with interoperability and accelerated adoption of 200 and 400 gig PAM4 interconnects using an analog approach. Between MACOM, Semtech and other participants, MSA members supply nearly 100% of the CDRs to Cloud customers, demonstrating MSA interoperability across a broad group of electronics, laser and component suppliers thus achieves de facto industry standardization.

今天稍早，我們宣布成立 Open Eye MSA Group，這是一個由 MACOM 與 Semtech 共同創立的產業聯盟，參與成員包括華為海思、Juniper、Innolight、AOI 與 Mellanox 等，另有近十家其他公司。多來源協議（MSA）的目標是透過類比方法，建立互通性並加速採用 200G 與 400G PAM4 互連。在 MACOM、Semtech 與其他參與者之間，MSA 成員幾乎供應雲端客戶所需的 100% CDR，並在廣泛的電子、雷射與元件供應商群體中展現 MSA 的互通性，從而達成事實上的產業標準化。

Today, we simultaneously announced our portfolio of MSA-targeted CDRs, drivers and TIAs, which add low-cost, low-power features for the PAM4 function. Instead of using a 16- or 7-nanometer DSP, which contains power-hungry A/D and D/A blocks, our PAM4 CDRs add modulation, detection and analog equalization in a much more cost-effective process node. This is inherently lower cost and much lower power consumption staying entirely in the analog domain.

今天我們也同步宣布一系列針對 MSA 的 CDR、驅動器與 TIA 產品組合，為 PAM4 功能加入低成本、低功耗特性。相較於採用 16 或 7 奈米 DSP（其中包含耗電的 A/D 與 D/A 模組），我們的 PAM4 CDR 以更具成本效益的製程節點，加入調變、偵測與類比等化功能。這在本質上成本更低，且由於完全維持在類比領域，功耗也大幅降低。

The end result is 25% less power per gigabit and similarly, lower cost per gigabit than CWDM4 and DSP-based PAM4 solutions. Cloud customers expect to double their link rate with only minor incremental power and cost using the analog CDR-based approach. But I want to be clear here. We're not promoting the Open Eye MSA as a replacement for IEEE single-lambda, and we're not promoting CDRs instead of DSPs. We supply both with complete portfolios of analog and photonic components for both PAM4 approaches equally.

最終結果是每 Gbps 的功耗降低 25%，同時每 Gbps 的成本也同樣低於 CWDM4 與基於 DSP 的 PAM4 解決方案。雲端客戶預期透過基於類比 CDR 的方法，在僅有小幅增加功耗與成本的情況下，將鏈路速率提升一倍。但我想在此說清楚。我們並非將 Open Eye MSA 推廣為 IEEE 單波長（single-lambda）的替代方案，也不是在推廣以 CDR 取代 DSP。我們同時供應兩者，並針對兩種 PAM4 方法，均提供完整的類比與光子元件產品組合。

As I explained earlier, the 2 approaches coexist serving different parts of the industry with different fiber infrastructure and in their data centers.

如我先前所述，這兩種方法會共存，分別服務於產業中不同的區塊、不同的光纖基礎設施，以及其資料中心環境。

At OFC we also announced with GlobalFoundries 12-inch silicon Photonics wafer production to support our L-PIC ramp at hyperscale data center volumes. Our L-PIC portfolio spans PAM4 at 100, 200 and 400 gig covering both analog and DSP approaches.

在 OFC 上，我們也與 GlobalFoundries 共同宣布 12 吋矽光子晶圓量產，以支援我們的 L-PIC 在超大規模資料中心出貨量下的爬坡。我們的 L-PIC 產品組合涵蓋 100G、200G 與 400G 的 PAM4，並同時覆蓋類比與 DSP 兩種方法。

So we're covered whichever way the industry goes.

因此無論產業走向何方，我們都已布局到位。

Silicon Photonics, in general, including our L-PICs is poised to play a key enabling role for PAM4 in both DSP and analog forms. It obviates the need for otherwise costly and power-hungry EML lasers which have been a major impediment to PAM4 adoption. As part of today's product announcement, our FR4 L-PIC is poised to enable 200 gig transceivers later this year at or below the current costs of 100 gig modules.

整體而言，矽光子（包括我們的 L-PIC）有望在 DSP 與類比兩種形式的 PAM4 中扮演關鍵的使能角色。它免除了原本昂貴且耗電的 EML 雷射需求，而 EML 雷射一直是 PAM4 採用的一大阻礙。作為今日產品發布的一部分，我們的 FR4 L-PIC 有望在今年稍晚支援 200G 收發器，其成本可達到或低於目前 100G 模組的成本水準。

By this point, I hope it's evident we believe our secular growth opportunity in data centers is fully intact. Our portfolio is expanding geometrically to include DSPs, lasers and L-PICs and our share's growing in a market that is expected to recover if not snap back in the second half of this calendar year. Separating the short-term cyclical from the long-term secular, our future has never been brighter in cloud data centers.

到此為止，我希望已很明顯：我們相信資料中心的長期結構性成長機會依然完全存在。我們的產品組合正以幾何級數擴張，涵蓋 DSP、雷射與 L-PIC；而在預期於本曆年下半年復甦、甚至可能快速反彈的市場中，我們的市占率也在提升。將短期循環因素與長期結構性因素區分開來看，我們在雲端資料中心的未來從未如此光明。

Next, on to 5G Telecom.

接下來談 5G 電信。

Anyone who attended Mobile World Congress knows that 5G is real, and it's starting in a big way later this year on a global basis in applications ranging from sub-6 gigahertz microcell basestations to millimeter wave fixed wireless networks. We had a great roster of customer meeting in Barcelona with major network operators as well as leading basestation OEMs worldwide.

任何參加過世界行動通訊大會（MWC）的人都知道，5G 是真實存在的，而且將在今年稍晚以全球規模大舉啟動，應用範圍從 6GHz 以下的微型蜂巢基地台到毫米波固定無線網路。我們在巴塞隆納與主要網路營運商以及全球領先的基地台 OEM 進行了大量客戶會議。

One clear conclusion from field trials is that GaN is now a requirement for 5G microcell basestations. One major OEM actually tried to use LDMOS because it was all they could source in volume. That failed miserably.

從實地測試得出的一個明確結論是：GaN 現在已成為 5G 微型蜂巢基地台的必要條件。某家主要 OEM 實際上曾嘗試使用 LDMOS，因為那是他們唯一能大量取得的元件。結果徹底失敗。

The Chief Technology Officer of one of the largest wireless operators told us that their antennas and field trials were twice the target power consumption and 3x too expensive. Some antennas even had to be liquid cooled. Major players now understand that GaN on silicon uniquely provides it all: target efficiency, costs and a high-volume supply chain. We're now squarely fixated on ramping production with ST in support of the 5G buildout globally.

某家最大型無線營運商之一的技術長告訴我們，他們的天線與實地測試結果顯示功耗是目標的兩倍，成本則高出 3 倍。有些天線甚至必須採用液冷。主要業者如今已理解：矽基氮化鎵（GaN on silicon）能以獨特方式同時滿足所有需求——目標效率、成本，以及高產量供應鏈。我們現在正全力聚焦於與 ST 一起提升產能，以支援全球 5G 建設。

Marco Monti, President of SD Micro's Automotive and Discrete group, joined us at our investor event in Barcelona. He outlined how their GaN capacity expansion, copy exact in Italy and Singapore will be able to service over 80% of global demand for 5G basestations, exclusively through MACOM and our pending joint venture in China.

SD Micro 汽車與分離元件事業群總裁 Marco Monti 也在巴塞隆納的投資人活動中與我們同台。他說明了他們在義大利與新加坡的 GaN 產能擴張（完全複製的產線）將可服務全球超過 80% 的 5G 基地台需求，並將透過 MACOM 以及我們在中國待核准的合資企業獨家供應。

That brings me to the really big news this quarter, that JV in China. This is the capstone to MACOM's strategy to become a scale player in the multibillion-dollar 5G opportunity worldwide. We've taken a page out of the semiconductor industry playbook and formed a JV that can credibly and locally service a multibillion-dollar opportunity in China's domestic market. Over the last quarter, we ran a comprehensive process to find a Chinese investment partner. I'm very proud and excited to say that the process concluded with Goertek, a proven success story in China. Goertek is a high-volume supplier of semiconductor components, modules and assembled products to the most demanding, smartphone OEMs and service providers worldwide.

這就帶到本季真正的重大消息：我們在中國的合資企業（JV）。這是 MACOM 策略的關鍵一環，旨在成為全球數十億美元規模 5G 機會中的具規模參與者。我們借鑑半導體產業的策略手冊，成立一個能在中國國內市場以在地方式、可信賴地服務數十億美元機會的合資企業。在上一季，我們進行了全面性的流程以尋找中國投資夥伴。我非常自豪且興奮地宣布，該流程最終選定了歌爾股份（Goertek），這是一個在中國已被驗證的成功案例。Goertek 是半導體元件、模組與組裝產品的高產量供應商，服務全球要求最嚴苛的智慧型手機 OEM 與服務供應商。

Over the last 10 years, they've grown revenue from around $100 million, peaking at over $4 billion growing at over a 40% compound rate during that period. Our 5G GaN on silicon opportunity fits Goertek's operating model perfectly.

在過去 10 年間，他們的營收從約 1 億美元成長，最高曾超過 40 億美元，期間以超過 40% 的複合成長率成長。我們的 5G 矽基 GaN 機會與 Goertek 的營運模式完美契合。

They source proprietary semiconductor products and technology from industry leaders like TI, Broadcom and Infineon. In this case, they'll source MACOM's GaN-on-Silicon wafers directly from STMicro and supply 5G power amplifiers to China-based OEMs.

他們向 TI、Broadcom 與 Infineon 等產業領導者採購專有的半導體產品與技術。在此案例中，他們將直接向 STMicro 採購 MACOM 的矽基 GaN 晶圓，並向中國本土 OEM 供應 5G 功率放大器。

Okay. Here's the key facts about the transaction. We signed and announced a JV on April 24. We expect the transaction to close in 2 to 4 months. Given Goertek's size, the transaction is subject to antitrust approval from China State Administration for Market Regulation. We've taken great care that the only commodities and technologies that MACOM will share are freely exportable to China under U.S. law. The license products are civil, telecommunications grade or EAR-99. We might as well be selling fountain pens. No technology or existing U.S. business will transfer so we don't expect any issues with U.S. Export Control or CFIUS.

好的。以下是這筆交易的關鍵事實。我們於 4 月 24 日簽署並公告成立合資公司（JV）。我們預期交易將在 2 到 4 個月內完成交割。鑑於 Goertek 的規模，本交易須取得中國國家市場監督管理總局的反壟斷核准。我們已非常謹慎，確保 MACOM 將分享的唯一商品與技術，皆可依美國法律自由出口至中國。授權產品為民用、電信等級或 EAR-99。我們幾乎就像是在賣鋼筆一樣。不會移轉任何技術或既有的美國業務，因此我們不預期會在美國出口管制或 CFIUS 方面遇到任何問題。

Goertek will provide total consideration to MACOM of up to $134.6 million consisting of an upfront payment of $30 million and revenue and profit-based milestone payments. In exchange for a 51% equity stake in the venture. Separately, Goertek and MACOM will each initially contribute $25 million in working capital to fund the venture. So the net cash benefit to MACOM is up to $110 million. It's worth noting that Goertek's $134.6 million will more than return on our total investment to date in GaN-on-Silicon.

Goertek 將向 MACOM 提供最高達 1.346 億美元的總對價，包括 3,000 萬美元的預付款，以及以營收與獲利為基礎的里程碑付款。以換取該合資企業 51% 的股權。另外，Goertek 與 MACOM 將各自先行投入 2,500 萬美元的營運資金以資助該合資企業。因此，MACOM 的淨現金利益最高可達 1.10 億美元。值得注意的是，Goertek 的 1.346 億美元將足以使我們迄今在矽基 GaN 的總投資獲得超額回收。

In terms of ongoing economics, MACOM will be entitled to dividend preferences and royalties from the joint venture. Once the transaction closes, MACOM will receive cash payments from the JV of $8 million per quarter and IP license maintenance fees for a period of 2 years post-signing. Those fees began accruing as of April 24, the date of signing, and the accrued cash will be paid when the transaction closes.

就持續性的經濟利益而言，MACOM 將有權取得合資公司的股利優先權與權利金。交易完成後，MACOM 將在簽署後 2 年期間內，每季自合資公司收到 800 萬美元的現金付款，以及 IP 授權維護費。這些費用自 4 月 24 日簽署日起開始計提，累計的現金將於交易完成交割時支付。

MACOM also retains rights to GaN-on-Silicon sales outside of Greater China, which has traditionally comprised half the global market for RF power. So in total, on a combined basis, MACOM retains more than 75% of the global economic opportunity for GaN-on-Silicon in basestations.

MACOM 亦保留在大中華區以外的矽基 GaN 銷售權利；傳統上，大中華區以外市場約占全球 RF 功率市場的一半。因此，合併計算下，MACOM 在基地台矽基 GaN 的全球經濟機會中仍保留超過 75%。

Success in China in turn will provide MACOM a fantastic cost structure and supply chain and the ability to invest further in GaN innovation in the U.S., with which we can replicate that earnings power on a global basis.

反過來說，在中國的成功將為 MACOM 帶來極佳的成本結構與供應鏈，並使我們能在美國進一步投資 GaN 創新，藉此在全球範圍複製那樣的獲利能力。

So in summary, we believe that our progress last quarter -- expanding and adding manufacturing and investment partnerships with companies of the scale and strength of STMicro and Goertek -- forms a powerful combination to unlock the China 5G opportunity and provide tremendous earnings power for MACOM and our shareholders. I'll remind everyone that GaN-on-Silicon power amplifiers is but a subset of our overall 5G opportunity for MACOM. We have similar opportunities with other RF technologies like AlGaAs as well as our high-performance analog, DSP and photonic products in the 5G optical buildout. In the interest of time, I'll defer those discussions to our next earnings call.

總結而言，我們相信上季的進展——與 STMicro 與 Goertek 這類具規模與實力的公司擴大並新增製造與投資合作夥伴關係——形成強而有力的組合，可解鎖中國 5G 機會，並為 MACOM 與股東帶來可觀的獲利能力。我也提醒各位，矽基 GaN 功率放大器只是 MACOM 整體 5G 機會的一個子集合。我們在其他 RF 技術（如 AlGaAs）以及 5G 光纖建設中的高效能類比、DSP 與光子產品方面，也有類似機會。為節省時間，我將把這些討論留待下一次法說會。

Suffice to say that our telecom portfolio is poised for long-term secular growth on the back of 5G infrastructure investments beginning later this year. I look forward to providing more detail and delivering on the vision in the quarters ahead.

簡而言之，隨著今年稍晚開始的 5G 基礎建設投資，我們的電信產品組合已蓄勢待發，將迎來長期的結構性成長。我期待在未來幾季提供更多細節，並落實這一願景。

Finally, before closing today's scripted remarks, I'd like to touch briefly on noteworthy developments in global and homeland defense. Since our last earnings call, 2 of our major customers, Raytheon and Lockheed Martin, announced a teaming agreement to pursue the SENSR contracts. We believe this greatly simplifies the bidding landscape since combined Raytheon and Lockheed represent 80% of overall ground-based radar deployments globally. While many of the moving pieces are still playing out, we believe that this may ultimately result in a more timely deployment of what is shaping up to be the largest buildout of ground-based radar in the history of the United States.

最後，在結束今天的預先準備講稿之前，我想簡要談談全球與國土防衛領域值得注意的進展。自上次財報電話會議以來，我們兩家主要客戶 Raytheon 與 Lockheed Martin 宣布簽署合作協議，以共同爭取 SENSR 合約。我們認為這將大幅簡化投標態勢，因為 Raytheon 與 Lockheed 合計代表全球地面雷達部署總量的 80%。雖然許多變數仍在發展中，我們相信這最終可能使部署更為及時，而這正逐步成形為美國史上最大規模的地面雷達建置。

While SENSR deployment is still a couple of years out, final bids are expected to report full-scale preproduction builds, which we're ready to support in engineering and in manufacturing. Outside of SENSR, the threat from unauthorized drones continues to be a major challenge for open air venues, high-rise apartments and public facilities. Just last month, another unauthorized drone incursion took place over another public venue. This time it was over Fenway Park here in Boston. Now it's getting personal. There's a newfound urgency among government agencies including the FAA, FBI and DHS to address these vulnerabilities. This has driven a surge in interest in our radar tile technology from a wide variety of stakeholders. We've been very active recently with government agencies not only for SENSR but also for opportunities to protect these civil open air venues. After years of investment and development work, we're in the right place at the right time with a novel approach that offers proven field performance. The world's leading radar customers and government agencies understand the value that we're bringing to bear with Scalable Planar Arrays better than anyone and are teaming with MACOM for precisely that reason. We view global and homeland defense to be a strong, domestic secular growth engine stretching for arguably decades to come.

雖然 SENSR 的部署仍需再等幾年，但預期最終投標將涵蓋全規模的量產前（preproduction）建置，我們已準備好在工程與製造方面提供支援。除 SENSR 之外，未經授權的無人機威脅仍是露天場館、高樓住宅與公共設施面臨的一大挑戰。就在上個月，又有一起未經授權的無人機侵入事件發生在另一個公共場所上空。這次是在波士頓的芬威球場（Fenway Park）上空。現在這變得很切身。包括 FAA、FBI 與 DHS 在內的政府機關，對於解決這些弱點出現了新的急迫感。這也帶動各類利害關係人對我們雷達拼板（radar tile）技術的興趣大幅升溫。近期我們與政府機關互動非常積極，不僅針對 SENSR，也針對保護這些民用露天場域的機會。經過多年投資與研發工作，我們以一種具備已驗證實地表現的新方法，在對的時間站在對的位置。全球領先的雷達客戶與政府機關，比任何人都更了解我們以可擴展平面陣列（Scalable Planar Arrays）所帶來的價值，因此正是基於這個原因與 MACOM 合作。我們認為全球與國土防衛將是強勁的、以國內為主的結構性成長引擎，且可望延續數十年。

To conclude, while all of our secular growth opportunities and cloud data centers, 5G Telecom and global and homeland defense remain fully intact, we believe that the current inventory correction in data centers will continue through the June quarter. With that in mind, we're guiding fiscal Q3 to be essentially flat to Q2. More specifically, for the third fiscal quarter ending June 28, 2019, revenue is expected to be in the range of $120 million to $124 million. Gross margin is expected to be between 53% and 55% and adjusted earnings per share between a loss of $0.08 and a loss of $0.04 on an anticipated 66.5 million fully diluted shares outstanding.

最後總結，雖然我們所有的結構性成長機會——雲端資料中心、5G 電信，以及全球與國土防衛——仍完全不變，但我們認為資料中心目前的庫存修正將持續到 6 月季度。基於此，我們預期 2025 會計年度第 3 季（Q3）指引將與 Q2 大致持平。更具體而言，對於截至 2019 年 6 月 28 日的會計年度第三季，預期營收介於 1.20 億至 1.24 億美元。預期毛利率介於 53% 至 55%，調整後每股盈餘（EPS）介於虧損 0.08 美元至虧損 0.04 美元（以預估 6,650 萬股完全稀釋流通股計）。

Again, our fiscal Q3 guidance anticipates no material recovery in demand before the end of the quarter. Also, not factored into our guidance is potential recognition of licensing revenue from funds that we received last quarter from a new solutions customer, and any benefit of license maintenance fees if our JV with Goertek closes before the end of the quarter. Each of these could represent upside to our current Q3 guidance. We believe the worst is behind us. If not this quarter, we look forward to getting back on track with improving financial performance as the market recovers later this year. Some of our customers indicate they'll be exiting the quarter at more normalized inventory levels, which is a positive leading indicator of recovery.

再次強調，我們對會計年度 Q3 的指引預期在本季結束前需求不會出現實質性復甦。此外，我們的指引未納入：上季自一位新解決方案客戶收到資金所可能認列的授權收入，以及若我們與 Goertek 的合資案在本季結束前完成交割，可能帶來的授權維護費收益。上述各項都可能成為我們目前 Q3 指引的上行因素。我們相信最糟的情況已經過去。即使不是在本季，我們也期待隨著今年稍晚市場復甦，財務表現回到改善軌道。部分客戶表示，他們將在本季末回到較正常的庫存水位，這是復甦的正向領先指標。

Despite this temporary, albeit precipitous air pocket in data centers, the fundamental demand drivers for each of our growth opportunities remain intact. In each case, we're positioned with a portfolio of disruptive technologies, protected by strong and proven intellectual property. With the progress we made just in the last quarter, we've never been more optimistic about MACOM's future prospects for growth and profitability.

儘管資料中心出現這段短暫但相當急遽的數據「空窗期」，我們各項成長機會的根本需求驅動因素仍然完好無損。在每一個領域，我們都以一套具顛覆性的技術組合佈局，並由強大且經過驗證的智慧財產權加以保護。憑藉我們僅在上一季所取得的進展，我們對 MACOM 未來的成長與獲利前景從未如此樂觀。

This air pocket in data centers highlights an interesting challenge for the company that we started dealing with head-on last quarter. Several of our growth initiatives are coming to fruition simultaneously. Our challenge is to deploy sufficient resources to meet our customers' demands within the windows of opportunity. This raises a valid question of affordability. Our JV with Goertek is a great example of a proactive solution to address that challenge. Putting the business in an even better position for success while defraying operating expenses and at the same time retaining substantial earnings power for our shareholders. We believe there are similar opportunities for other parts of our portfolio. Operator, you can now open the call to questions.

資料中心的這段空窗期凸顯了公司一項有趣的挑戰，而我們在上一季已開始正面著手處理。我們多項成長計畫正同時開花結果。我們的挑戰是在機會窗口內部署足夠資源，以滿足客戶需求。這也引出一個合理的問題：可負擔性。我們與歌爾（Goertek）的合資企業（JV）就是一個主動解決此挑戰的絕佳例子。在分攤營運費用的同時，讓業務處於更有利的成功位置，並且仍為股東保留可觀的獲利能力。我們相信在投資組合的其他部分也存在類似機會。接線員，現在可以開放提問。

(Operator Instructions) Our first question comes from Quinn Bolton with Needham.

(接線員指示) 我們的第一個問題來自 Needham 的 Quinn Bolton。

This is Michelle on for Quinn. For my first question, I guess on the GaN on Silicon side of things, can you guys give us an update on your confidence level for volume shipments taking place in calendar 4Q of this year? Does that remain on track? And is there any update on the $23 million investment with STMicro and where that stands, too?

我是代替 Quinn 的 Michelle。我的第一個問題，我想是關於矽基氮化鎵（GaN on Silicon）這一塊，你們能否更新一下對於今年曆年第四季（4Q）開始進行量產出貨的信心程度？這是否仍按計畫進行？另外，與意法半導體（STMicro）相關的 2,300 萬美元投資目前進展如何，也能否更新一下？

Yes. So we have very clear line of sight to the first wave of deployments over in China specifically, and we're completing the product deliveries and qualifications, validations with our lead customer in line with collecting orders and ramping production in support of those 4Q shipments. I'd say, in general, with ST, those -- that capital equipment investment is in the process. There's equipment on order and going to be entering the first ST factory through the course of this year for completing the qualifications. At Mobile World Congress, we actually shared with Marco Monti the production ramp timeframe. It was basically big step-ups as you step through 2020 from both the Italy and the Singapore-based factories in ST.

可以。我們對於第一波部署（特別是在中國）有非常清晰的能見度；我們正依照取得訂單並提升產能、以支援第四季出貨的節奏，完成對主要客戶的產品交付與認證、驗證。整體而言，與 ST 相關的資本設備投資正在進行中。設備已下單，並將在今年內陸續進入 ST 的第一座工廠以完成資格認證。在世界行動通訊大會（Mobile World Congress）上，我們也與 Marco Monti 分享了量產爬坡的時間表。基本上，隨著 2020 年推進，ST 位於義大利與新加坡的工廠都會出現明顯的產能階梯式提升。

Okay. Great. To the PAM4 analog CDR-based chipset you announced today, it sounds like these are solutions are sampling today. Can you give us some more color on how you see those solutions ramping over time and what types of distances you're targeting?

好的。很棒。關於你們今天宣布的 PAM4 類比 CDR 架構晶片組，聽起來這些解決方案目前已在送樣。能否再多分享一些你們如何看待這些方案隨時間推進的放量節奏，以及你們鎖定的傳輸距離類型？

Yes. Actually, it's more than sampling. We've actually been working in the labs with our cloud customers demonstrating interoperability with specifically Semtech products so that's the goal of this multisource agreement, to demonstrate multiple sources for each of the components and transceivers, and that's very far along. We're now in process with the MSA group of documenting specification so that we can actually publish the specs later this year. But yes, those products are real and available, and cloud customers and transceiver guys are actually designing them in. So it's off to the races. It's not concepts. It's actually hard work in the lab demonstrating performance.

可以。其實不只是送樣。我們已經在實驗室與雲端客戶合作，展示與特定 Semtech 產品的互通性；這正是這項多來源協議（MSA）的目標：針對各個元件與收發器展示多家供應來源，而這項工作已進展到相當後期。我們目前正與 MSA 團隊一起撰寫規格文件，以便在今年稍晚正式發布規格。但沒錯，這些產品是真實存在且可供取得的，雲端客戶與收發器廠商也正在把它們設計導入。所以已經全面啟動。這不是概念。而是在實驗室裡透過艱苦的工作來驗證效能。

Our next question comes from Mark Lipacis with Jefferies.

我們的下一個問題來自 Jefferies 的 Mark Lipacis。

First one on the free cash flow, nice to see that up in the quarter. Can you give us a framework for thinking about how we should be modeling free cash flow going forward, either on a -- either on like a projection on what you think it's going to do for the rest of the fiscal year or as how we should think about it in terms of net income conversion? That's the first question. I have a follow-up.

第一個問題關於自由現金流，本季看到它上升很不錯。你們能否提供一個思考框架，讓我們在往後建模自由現金流時有所依據——不論是你們對本會計年度剩餘期間的預估走勢，或是從淨利轉換率的角度來看，我們應該如何理解？這是第一個問題。我還有一個追問。

In terms of the free cash flow, Mark, for Q2, obviously, a very strong performance with the improved working capital position we posted. So our free cash flow was much better than we expected for Q2 and actually is positive for the year-to-date. In terms of Q3, Mark, there was lots of puts and takes for Q3 and lots of variability around the free cash flow number. We will see a slightly higher CapEx in Q3 and hope to offset that with some further improvements in working capital. In terms of rest of the year, we don't give guidance into the rest of the year.

Mark，就第二季（Q2）的自由現金流而言，很明顯我們因營運資金狀況改善而繳出非常強勁的表現。因此，第二季自由現金流遠優於我們原先預期，且年初至今（YTD）也轉為正值。至於第三季（Q3），影響自由現金流的正負因素很多，數字的變動性也很大。我們會在第三季看到資本支出（CapEx）略為提高，並希望透過進一步改善營運資金來抵銷。至於年度剩餘期間，我們不提供全年其餘時間的指引。

Fair enough. That's helpful. John, can you describe what is your appetite for M&A today versus 2 years ago, let's say?

了解。這很有幫助。John，你能否描述一下，相較於例如兩年前，你們現在對併購（M&A）的意願／胃納如何？

Yes, that's a fantastic question. We're actually going the other way right now. What I had referred to at the end of the scripted remarks is this JV is a fantastic construct that we're working not to bring more technology into the company but actually get more leverage. Our big problem that we're dealing with right now, if you go through the scripted remarks, all the stuff in data centers, whether it's the L-PICs, the DSPs, as well as obviously the GaN on Silicon stuff, a lot of our growth initiatives actually come to fruition right now, and they become all-consuming in terms of management focus, operating expense, capital expense and so on. And this JV construct that we came upon in running that process over in China, I think, is a great model. I mean we put that business in China in much better hands to be able to service that domestic market locally. We defray operating expense $8 million per quarter begins accruing April 24 so that's directly offsetting OpEx. And we retain very substantial economic value for our shareholders. I mean between preferred dividends, royalties and the maintenance fees, it's very large share of the economic value in China and 100% outside of China we retain, so it's like well more than 75%, well more than 80% for that matter. So it's a great model and I think we're going to be exercising lots of things that we can do to be able to relieve pressure on the balance sheet and relieve pressure on the P&L.

是的，這是個非常好的問題。我們目前其實是往相反方向走。我在事先準備的結語中提到，這個合資企業（JV）是一個很棒的架構；我們推動它並不是為了把更多技術帶進公司，而是為了取得更高的槓桿效益。我們現在面臨的最大問題是——如果你回顧事先準備的內容——資料中心相關的所有事情，不論是 L-PICs、DSPs，當然也包括矽基氮化鎵（GaN on Silicon），我們許多成長計畫都在此刻同時成熟，因而在管理焦點、營運費用、資本支出等方面變得幾乎佔據全部資源。而我們在中國推動該流程時所形成的這個 JV 架構，我認為是一個很好的模型。我們把中國的這項業務交由更適合的在地團隊來經營，以便在當地服務國內市場。我們分攤營運費用——每季 800 萬美元——自 4 月 24 日起開始認列，這會直接抵銷 OpEx。同時，我們也為股東保留了非常可觀的經濟價值。我的意思是，透過特別股股利、權利金以及維護費，我們在中國保留了很大比例的經濟價值，而在中國以外我們保留 100%，因此整體來看遠超過 75%，甚至遠超過 80%。所以這是一個很好的模型；我認為我們會採取許多措施，以減輕資產負債表的壓力並減輕損益表（P&L）的壓力。

Our next question comes from Blayne Curtis with Barclays.

我們的下一個問題來自 Barclays 的 Blayne Curtis。

This is Tom O'Malley on for Blayne Curtis. Just looking into June here, can you give us the puts and takes across the businesses? You guys are guiding flattish revenue, but just in terms of segments, do you have any color on which ones are growing or which ones are standing back a bit?

我是代替 Blayne Curtis 的 Tom O'Malley。就 6 月這一季來看，你們能否說明各業務之間的正負因素？你們給出的營收指引大致持平，但就各個事業部／區塊而言，能否提供一些顏色，哪些在成長、哪些稍微放緩？

So they're all within a few percentage points of the same kind of thing, and the visibility is low. In telecom right now, it's China visibility is the issue. In data center it's -- there's zero visibility in terms of recovery. In I&D, we're still dealing with seasonality we were down sequentially but up 15% year-on-year, so that's, really an issue of lumpiness in defense programs, to be honest. But I would say that the precision of our forecasting is greater than the numbers that we have on the wall here.

所以它們都在同一類型的事情上相差幾個百分點，而且能見度很低。在電信方面，目前的問題是中國的能見度。在資料中心方面，就復甦而言完全沒有能見度。在 I&D 方面，我們仍在面對季節性，雖然較前一季下滑，但年增 15%，所以老實說，這其實是國防專案出貨不均所造成的波動問題。但我會說，我們的預測精準度其實比我們現在牆上看到的這些數字還要高。

Great. And then just talking about the broader opportunities and this is more longer term, you've kind of described the GaN opportunity being more Q4 to start and then into next year. The ground-based radar arrays is a little further out as well under the telco market you said, maybe improving slightly, but what's driving the growth in the back half of this calendar year? How do you guys keep growing revenue before some of these bigger opportunities hit?

很好。接著談談更廣泛的機會，這比較偏長期；你們先前描述 GaN 的機會大概是從第四季開始，然後延伸到明年。地面式雷達陣列也更往後一些，你也提到電信市場可能略有改善，但今年下半年這個曆年下半段的成長動能是什麼在驅動？在這些更大的機會真正到來之前，你們要如何持續讓營收成長？

Well, in the September quarter, you're going to see data centers begin to recover. It's very clear. When we have multiple customers talking about getting back to normalize inventories by the end of this quarter, that's indicative of a lift. It's far too early to call that a strong lift, but there are some out there that's talking about a snapback. So data center will look like growth. I would say in telecom, we're waiting for 5G. And the issue with the GaN on Silicon stuff now is we're working through the accounting treatment with Deloitte. So assuming we don't need to consolidate, which is my expectation, you're going to start getting P&L benefit, certainly $8 million of cash benefit per quarter. And again, that started accruing April 24. So when that transaction closes, we get that cash accrued through this quarter. If we close this quarter, you could get a bluebird on Q3 guidance, but if it goes into next quarter, that snowplows into the next quarter.

嗯，在 9 月季度，你會看到資料中心開始復甦。這非常明顯。當我們有多位客戶都在談到要在本季末把庫存恢復到正常水位，這就代表需求會有抬升。現在說這會是強勁的拉升還太早，但確實有人在談會出現快速反彈。所以資料中心看起來會呈現成長。我會說在電信方面，我們在等 5G。而目前矽基 GaN（GaN on Silicon）的問題是，我們正在和 Deloitte 一起釐清會計處理方式。所以假設我們不需要合併報表——這也是我的預期——你們將開始看到損益表（P&L）的效益，當然每季會有 800 萬美元的現金效益。而且這從 4 月 24 日就開始累積。所以當這筆交易完成時，我們會拿到截至本季為止累積的現金。如果我們在本季完成，你可能會在 Q3 指引上拿到一個意外之喜；但如果延到下季，那就會把效益順延到下一季。

Our next question comes from Mark Delaney with Goldman Sachs.

下一題來自高盛（Goldman Sachs）的 Mark Delaney。

First question, I was hoping to better understand the gross margin guidance for next quarter. I think there wasn't the inventory write-down. Gross margin was about 55 for this quarter, pretty flattish revenues and I think the midpoint of gross margins down a little bit for next quarter. So maybe just help us understand what is being factored in for gross margin for next quarter?

第一個問題，我希望能更了解下季的毛利率指引。我想這一季沒有庫存減記。本季毛利率大約 55%，營收相當持平，而我想下季毛利率的中位數會小幅下滑。所以能否協助我們理解，下季毛利率的假設因素是什麼？

Well we assume similar mix to this quarter excluding the inventory adjustment, the inventory reserves, to be honest, Mark. I mean across the board, we just -- the visibility is so poor that it's difficult to get much more precision than that. I would say the tough part of this downturn is as odd as it sounds our high-performance analog business and our connectivity business are among the highest gross margin in the company, and so it hurts. This depression in data center revenues is actually a pretty material dilution on our gross margin. When it comes back, you'll see a lift. So I'm very hopeful that in the September quarter, especially December quarter, we'll be snapping back and back on our March towards 60.

嗯，我們假設產品組合與本季相近，但不包含庫存調整、也就是庫存備抵，老實說，Mark。我的意思是，整體來看——能見度實在太差了，所以很難比這更精準。我會說這次景氣下行最棘手的部分是，聽起來很奇怪，但我們的高效能類比業務和連接（connectivity）業務其實是公司裡毛利率最高的幾個業務之一，所以衝擊很大。資料中心營收的低迷，確實對我們毛利率造成相當明顯的稀釋。等它回來，你就會看到毛利率上升。所以我非常希望在 9 月季度，尤其是 12 月季度，我們會快速反彈，並重新朝 60% 邁進。

Got it. That's helpful. And then for a follow-up on the telecom segment, if I recall properly last quarter guidance was for the Metro/Long-haul to be up very strongly in March driven by telecom applications. It sounded like that came in line with your expectations. Could you just talk a little bit about to what extent that did in fact come through in the regional -- in the regional drivers for China or other regions and what your view is going forward for that business?

了解。這很有幫助。接著追問電信部門，如果我沒記錯，上季的指引是 Metro/Long-haul 在 3 月季度會大幅成長，主要由電信應用帶動。聽起來實際結果符合你們的預期。你能否談談這在中國或其他地區等區域性驅動因素上，實際上反映的程度如何？以及你們對這項業務未來的看法？

Yes. Good question, Mark. Metro/Long-haul is the shining bright star in the portfolio right now. The Metro and client products were up something like 36% quarter-on-quarter, and I would say that ends up being lost in the telecom numbers with the other parts of the telecom portfolio like backhaul/fronthaul and PON. Because Metro/Long-haul business it's nice. It's just not a large driver that can make up for the big downturn in data center. But outside of the data center, things are looking -- I mean I think we're fully booked through next quarter. Not even this quarter, next quarter. So that's a very strong part of the business right now.

是的。好問題，Mark。Metro/Long-haul 目前是我們產品組合中最亮眼的明星。Metro 和 client 產品季增大約 36%，但我會說，這在電信整體數字裡會被其他電信產品組合的部分（例如 backhaul/fronthaul 和 PON）所掩蓋。因為 Metro/Long-haul 的業務表現很好。只是它的規模還不足以彌補資料中心大幅下滑所造成的缺口。但撇開資料中心不談，其他部分看起來——我的意思是，我們到下個季度都已經滿單了。不是這一季，是下一季。所以這目前是業務中非常強勁的一塊。

Our next question comes from Tore Svanberg with Stifel.

下一題來自 Stifel 的 Tore Svanberg。

John, could you elaborate a little bit more on the Goertek JV especially as it relates to applications you'll be selling in to. Will Goertek be selling PAs for basestations only? Or could they also be selling into smartphone or other applications?

John，你能否再多談一些關於歌爾（Goertek）合資公司（JV）的情況，特別是你們將銷售切入的應用領域？歌爾只會銷售基地台用的功率放大器（PA）嗎？還是也可能銷售到智慧型手機或其他應用？

The current license agreement is entirely basestations -- 100% basestations. They have ambitions to move beyond that. As we can quantify demand and frankly get consideration for that, then we'll consider expanding it. I'm very bullish -- the Goertek team, Long Jiang, the CEO, University of Maryland grad, Jason Lieu a Stanford grad, he's the CEO of the JV. I think it's going to be a great partnership to have to be able to put a lot of products into China. I think my lesson having worked at NXP, they were very successful setting up very similar JVs. It's one thing to supply boutique 50 -- $50 Million businesses like we traditionally do into China. It's an entirely different thing to ship hundreds of millions of dollars of, if not $1 billion of GaN PAs. So I think we have a long, healthy partnership with those guys before us.

目前的授權協議完全是基地台——100% 基地台。他們有企圖心要往那之外擴展。當我們能量化需求，而且坦白說也能因此取得相應對價時，我們就會考慮擴大範圍。我非常看好歌爾團隊——CEO 龍江是馬里蘭大學畢業，合資公司 CEO Jason Lieu 是史丹佛畢業。我認為這會是一個很棒的合作夥伴關係，能讓我們把大量產品導入中國。我想我在 NXP 工作時學到的一課是，他們透過建立非常類似的合資公司取得很大成功。把我們傳統上在中國供應的那種精品型、5,000 萬美元規模的業務做起來是一回事。但要出貨數億美元、甚至可能達到 10 億美元的 GaN PA，則完全是另一回事。所以我認為我們與他們之間會有一段長期且健康的合作關係。

As my follow-up on the PAM4 analog side, so I understand it's going to coexist with DSP. So what's the size of the market either by dollar terms or maybe millions of ports?

我追問一下 PAM4 類比這一塊；我理解它會與 DSP 共存。那這個市場的規模有多大？可以用美元金額，或是以埠（ports）的百萬數量來衡量？

It ends up that the division of the guys who come at PAM4 from a PSM4 heritage and those who come at it from a CWDM4 heritage. It's about a balance of 50-50. It isn't 80-20. Maybe it's 60-40 and can't quantify it. But if you look into the guys who have been PSM4 consumers, big names and so the IEEE single-lambda absolutely has a lot of traction and a bright future. Likewise, the guys with CWDM4 heritage, very strong names on that side so I mean the common denominator for us to be honest is a chockful of high-performance analog, laser and L-PIC content. We've been quantifying things even in the DSP instantiations 70% of the bill of materials is analog and photonic. So it's -- we've got very interesting position with L-PICs that go 100, 200, 400, 28 gigabaud for the analog, 53 gigabaud for the DSP-based stuff. PAM4 is going to be a very nice growth driver for us for the foreseeable future, frankly, once the industry recovers.

結果是，從 PSM4 傳承切入 PAM4 的那群人，與從 CWDM4 傳承切入的那群人之間，大致上是 50-50 的分布。大概各半。不是 80-20。也許是 60-40，但我無法量化得那麼精確。但如果你看那些一直是 PSM4 使用者的公司，都是大牌，所以 IEEE 的單波長（single-lambda）確實有很強的動能與光明的前景。同樣地，具有 CWDM4 傳承的那群人，在那一側也有非常強的品牌；所以老實說，對我們而言共同分母是滿滿的高效能類比、雷射與 L-PIC 內容。我們甚至在 DSP 的實作中量化過，物料清單（BOM）有 70% 是類比與光子元件。所以——我們在 L-PIC 上有非常有趣的定位，可支援 100、200、400；類比端是 28 gigabaud，而基於 DSP 的方案是 53 gigabaud。一旦產業復甦，PAM4 在可預見的未來將會是我們非常好的成長驅動力。

Our next question comes from Mark Kelleher with D. A. Davidson.

我們的下一個問題來自 D. A. Davidson 的 Mark Kelleher。

I just wanted to ask about a little clarification. You mentioned your white box module vendors were having some difficulty getting qualification. Can you just kind of revisit that? How important is that that to your data center ramp? And I think you connected that to some -- to the inventory charge if I heard that right. Is that right?

我只是想請教一點釐清。您提到你們的白盒模組供應商在取得認證（qualification）方面遇到一些困難。您能否再稍微回顧一下那部分？這對你們資料中心的爬坡（ramp）有多重要？而且我想您把這件事跟某些——跟庫存費用（inventory charge）連在一起，如果我沒聽錯的話。是這樣嗎？

No. I don't think there was a direct connection to the inventory charge, but the solution stuff, I would put it in context. We run a very, very healthy business of selling components to people who buy components and design and manufacture their own transceiver solutions. That's a very successful business model for us. We do have some customers that have taken higher level deliverables from us and planned ramps, and I'd just put it this way. The whole industry has just gone into a pause. The inventories have built up throughout the supply chain. What tends to happen is the end customers and the transceiver customers, they end up going back on their incumbents because, frankly, they prioritize keeping their incumbent suppliers healthy, and the new entrants have to wait. So I don't necessarily consider it long-term bad news for the new entrants. It's just not good news in the short term while we have this glut with inventory. So it's a pause is the best way to put it.

不是。我不認為那與庫存費用有直接關聯，但關於解決方案（solution）的部分，我會把它放在脈絡裡說明。我們有一個非常、非常健康的業務：向那些購買元件並自行設計與製造其收發器（transceiver）解決方案的客戶銷售元件。那對我們來說是一個非常成功的商業模式。我們確實也有一些客戶向我們採購更高層級的交付成果並規劃爬坡，我會這樣說：整個產業目前都進入了暫停狀態。供應鏈各環節的庫存都累積起來了。通常會發生的是，終端客戶與收發器客戶最後會回到他們原本的既有供應商（incumbents），因為坦白說，他們會優先確保既有供應商的健康，而新進者就得等待。所以我不一定把這視為對新進者的長期壞消息。只是短期在我們面臨這種庫存過剩（glut）時，這不是好消息。所以用「暫停」來形容是最貼切的。

Okay. And then just as a follow-up. Could you just clarify what the inventory charge was? I might have missed that?

了解。那再追問一下。您能否釐清一下那筆庫存費用是什麼？我可能漏聽了？

Yes. The biggest part of it was we were ramping our 25-gig DML lasers very aggressively in anticipation of supply shortages, which is what our customers were indicating that they had anticipated this year. And just as we were bringing them to fruition in the first calendar quarter, the whole supply chain turned over on us. I went to OFC and one of our biggest target customers told me that they were sitting on 6 months inventory of lasers, which I almost instantly fainted. But so these inventory reserves, every quarter, we do it the same way. It's a calculation. You look at your backlog, you look at your sales in context of your inventory and it's a straight forward calculation and in this particular instance, we had the ramping whip, right? It's a lot of whip. It's not necessarily finished goods although there is some. It's just we don't have any backdrop because there's no orders to be had. So it's not necessarily scrap. It's just writing down the valuation. It's the reason why we have good cash flow despite the operating loss.

好的。其中最大的一部分是：我們原本非常積極地拉升（ramp）我們的 25G DML 雷射，因為我們預期會出現供應短缺，而客戶也表示他們預期今年會如此。但就在我們於第一個曆年季度把它們推向量產成果之際，整個供應鏈情勢卻反過來了。我去 OFC 時，我們其中一個最大的目標客戶告訴我，他們手上有 6 個月的雷射庫存，我幾乎當場昏倒。不過，這些庫存準備（inventory reserves），我們每一季都用同樣方式處理。這是一個計算。你看你的訂單積壓（backlog），再看你的銷售，並結合你的庫存情況，這是一個直接的計算；而在這個特定情況下，我們有那個爬坡的「鞭子效應」（ramping whip），對吧？鞭子效應很大。它不一定是成品，雖然也有一些。只是我們沒有任何支撐，因為沒有訂單可拿。所以它不一定是報廢（scrap）。只是把價值做減記（writing down the valuation）。這也是為什麼即便有營業虧損，我們仍有良好現金流的原因。

Our next question comes from Tim Savageaux with Northland Capital.

我們的下一個問題來自 Northland Capital 的 Tim Savageaux。

First question on the joint venture in China. The $8 million a quarter that you mentioned, is that part of the milestone cash consideration? Is that -- to the extent there's 135 minus the 30 upfront? Or is that something different or incremental?

第一個問題關於在中國的合資企業（JV）。您提到每季 800 萬美元，那是里程碑現金對價（milestone cash consideration）的一部分嗎？也就是在 1.35 億減去 3,000 萬的預付款（upfront）之後那部分？還是那是不同的或額外的？

That's different. So there's license maintenance fees, which are $8 million, they started accruing April 24. And then separate from that, there's the $30 million upfront and the milestones, which are entirely revenue and profit gross profit-based earnouts. And we'll, as we ramp, we'll be burning through those and ticking off those -- that consideration and the last thing is we actually collect royalties on sales as well. And there's actually preferred dividends. So it's economically speaking, the ongoing economics are very attractive. But the $8 million a quarter (corrected by Company after the call) happens regardless.

那是不同的。所以有授權維護費（license maintenance fees），是 800 萬美元，從 4 月 24 日開始計提（accruing）。另外，還有 3,000 萬美元的預付款以及里程碑款，這些完全是以營收與毛利（gross profit）為基礎的分期獲利（earnouts）。而隨著我們爬坡，我們會逐步消化那些並逐一達成——那部分對價；最後一點是，我們也會對銷售收取權利金（royalties）。而且其實還有優先股股利（preferred dividends）。所以從經濟性來看，持續性的經濟利益非常有吸引力。但每季 800 萬美元（公司在電話會議後更正）是不論如何都會發生的。

Understood. Did you say that was 2 years for the agreement for that?

了解。您是說那個協議的期間是 2 年嗎？

Yes, the initial agreement is 2 years. I would fully expect -- there's no reason it won't continue on beyond 2 years. It's just for 2 years there was clearly a dependency on us supplying the product, licensing the product and not just the technology so it made sense to maintain it for 2 years, but there's -- again, I don't want to speculate 2 years out, but I would see no reason why that wouldn't continue.

是的，初始協議是 2 年。我完全預期——沒有理由它不會在 2 年之後繼續。只是這 2 年期間很明顯地依賴我們供應產品、授權產品，而不只是技術，所以維持 2 年是合理的；但——同樣地，我不想對 2 年後做推測，但我看不出有任何理由它不會延續。

Got it. I don't know if we think of that in a similar fashion to the solutions revenue over on the datacom side. Obviously, it's not recurring but for your kind of customers that you license, I wonder if you could maybe drill down a little bit more on the dynamics of being able to recognize revenue from your most recent solutions customer. Is that part of the whole qualifications slowdown process? And wasn't that a (inaudible) ?

明白。我不確定我們是否可以用類似資料通訊（datacom）那邊解決方案營收（solutions revenue）的方式來看待它。顯然它不是經常性（recurring）的，但對於你們這類被授權的客戶，我想請您能否再更深入談談，關於你們最新的解決方案客戶在認列營收方面的動態。那是否也是整個認證放緩流程的一部分？而且那不是一個（聽不清）嗎？

You got a double question there, let me make sure I cover both. So the $8 million per quarter we're going through the accounting treatment right now with Deloitte and first question is consolidate/not consolidate, right, assuming as it doesn't consolidate? Because we're relinquishing control not just equity ownership, but I mean the economics of dealing with the customers in the supply chain are now in the hands of the JV partner. But with that, the $8 million comes into how would it be treated? Is it contra-OpEx, which is the spirit of the maintenance fees? Or is it treated as consulting fee? If you remember when we sold the automotive business with Autoliv, you were covering back then, that showed up as a consulting fee, but so the second thing is do the math. The substantial impact on EPS, again accruing effective April 24 so it's got a material contribution this quarter if it closes. And if it doesn't close this quarter, that snowplows with a full quarter next quarter so that's -- that clock is already ticking. Separately, you asked the question about the license revenue that we did not declare revenue on. We just cleared -- no one's asked the question, but we just cleared the $7 million of non-GAAP revenue and the GAAP revenue recognition from a year ago. And you get into the new accounting standards and, basically, if you have initial milestones and a multi-milestone program, the new accounting standards, at least as we're interpreting them, are -- don't declare revenue, at least GAAP revenue until the completion of the entire program. So that's the reason why it took so long for the $7 million and the initial customer. In this particular instance, we actually got customer acceptance and the initial tranche of milestones, but we said, okay we're not doing non-GAAP revenue again. You can see it sits in liabilities and other liabilities on the balance sheet and we'll declare revenue when we complete that contract. So that could happen this quarter. I mean there's customer completion stuff this quarter. But until it's there, it's not there. So we're not going to declare it as revenue. But that's upside to the Q3 guidance as well.

你問了兩個問題，我確認我兩個都回答到。所以每季 800 萬美元，我們目前正與 Deloitte 針對會計處理進行討論；第一個問題是要不要合併（consolidate / not consolidate），對吧，假設它不需要合併？因為我們放棄的是控制權，不只是股權持有；我的意思是，與客戶及供應鏈打交道的經濟權益現在都在合資夥伴手上。在此情況下，這 800 萬美元要如何處理？它是作為營業費用的抵減（contra-OpEx），這符合維護費的精神？還是把它當作顧問費（consulting fee）？如果你記得我們與 Autoliv 出售汽車業務時——你當時也有追蹤——那筆就是以顧問費呈現；第二點是你算一下。對 EPS 的影響相當顯著，而且是自 4 月 24 日起生效開始計提，所以如果本季完成交割（close），本季就會有實質貢獻。如果本季沒有完成交割，那就會順延（snowplows）到下季，並在下季反映完整一季，所以——那個時鐘已經在走了。另外，你問到我們尚未宣告認列的授權營收。沒有人問這題，但我們剛剛把一年前那筆 700 萬美元的非 GAAP 營收以及 GAAP 營收認列處理完畢。而這牽涉到新的會計準則；基本上，如果你有初始里程碑以及多里程碑的計畫，新的會計準則——至少依我們的解讀——是：在整個計畫完成之前，不要宣告營收，至少不要宣告 GAAP 營收。所以這就是為什麼那 700 萬美元以及第一位客戶花了那麼久才認列的原因。在這個案例中，我們其實已取得客戶驗收（customer acceptance）以及第一批（tranche）的里程碑，但我們說，好，我們不再做非 GAAP 營收認列。你可以看到它在資產負債表上列在負債與其他負債（liabilities and other liabilities）裡，等我們完成那份合約時才會宣告認列營收。所以這可能在本季發生。我的意思是，本季確實有一些客戶完成（completion）的事項。但在它真正完成之前，就還不算完成。所以我們不會把它宣告為營收。但這也會是對第三季指引（Q3 guidance）的上行空間（upside）。

Got it. And last question for me, I think at the close of your initial comments, you kind of talked to the model on the GaN on Silicon JV and then maybe the solutions business, too. That's something that might be kind of extensible to the rest of your business. I wonder off the top of your head you'd be willing to share kind of any thoughts on what parts of your business might be -- might also be relevant to the type of opportunity?

了解。我這邊最後一個問題，我想在您一開始的評論結尾，您有提到 GaN on Silicon 的合資企業（JV）模式，然後可能也提到解決方案業務。那似乎是可以延伸到你們其他業務的一種作法。我想請教您，您是否願意就您腦中第一時間的想法，分享一下你們業務中哪些部分——也可能與這類機會相關？

Yes. So it's a very good question. So it's obviously too early to provide any specifics but here's the kind of the recipe for success, and the problem statement is this, and I realize and appreciate investors have been waiting a long time whether it's GaN on Silicon or the L-PICs or the DSPs or various growth initiatives that we've been putting money into that have taken longer. They are now coming to fruition. Those were in my scripted remarks in here exactly what the status is of all these things come to fruition. And the problem is they're demanding -- the customers are demanding more deliverables, more effort, more resources, both operating expense and capital expense, and this downturn has just put a big spotlight on the fact that we're burning too rich. It's a valid question of affordability. So I love the JV constructs. If we can do more of that, great. And the recipe is you put the business in the hands in a better position for success. The Goertek JV is a good example. We'd be kidding ourselves, frankly, if we thought we're going to have a $1 billion business per year in China servicing a domestic market. Even suppliers of magnitude of my partner before, NXP, they use the JV construct to address the domestic market. So one thing is you want to put it in better hands for success. Secondly, you absolutely want to the defray operating expense. If you put it into a JV, you can probably transfer the resources out. We couldn't do that with Goertek because there's export control issues so we have to do the maintenance fee structure, and the last thing is retaining substantial earnings power for our investors. And I'll tell you when you start running the numbers and I would suggest people do it. You look at the magnitude of the opportunity over China for instance and the probability of what share and you run the economics and you see what earnings power this unlocks as a good example, it's very eye-opening when we put that way. And I'll tell you, the other things I spoke about in the scripted remarks are coming to fruition as well. There is real GaN on Silicon, there's direct analogous progression that we're making in other parts of the business. So we're taking our situation very seriously. We took the OpEx issue head on. We delivered $8 million per quarter of OpEx defraying expenses. Now until it closes, we can't declare that, but it's accruable as of April 24. If we can pull off more of those deals and be able to get unlocking earnings power for our shareholders, we're going to do so. Because there's certainly limits to what we can fund with our current base of revenue even after recovery. Great question.

是的。這是一個非常好的問題。顯然現在要提供任何具體細節還太早，但我先分享一下成功的「配方」，以及我們要解決的問題：我理解也感謝投資人等了很久，不論是 GaN on Silicon、L-PICs、DSPs，或是我們投入資金的各項成長計畫，確實都花了更長時間。現在它們正在開花結果。我在事先準備的講稿裡也有提到，這些事情目前的狀態正是逐步落地、走向成果。但問題在於，客戶正在要求——要求更多交付、更多投入、更多資源，不論是營運費用（OpEx）或資本支出（CapEx）；而這次景氣下行也讓「我們燒錢燒得太兇」這件事被放大檢視。這確實是關於可負擔性的合理問題。所以我很喜歡 JV 這種架構。如果我們能做更多這樣的安排，那很好。而其成功配方是：第一，把業務放在更有利於成功的位置與手上。Goertek 的 JV 就是一個好例子。坦白說，如果我們以為自己能在中國做出每年 10 億美元、服務國內市場的業務，那是在自欺欺人。即便是我以前的合作夥伴 NXP 這種量級的供應商，也會用 JV 架構來切入國內市場。所以第一點是，你要把它放到更有利於成功的手上。第二，你一定要分攤（defray）營運費用。如果放進 JV，你很可能可以把資源轉出去。我們在 Goertek 這案子上做不到，因為有出口管制問題，所以我們必須採用維護費（maintenance fee）的結構；最後一點是，要為投資人保留可觀的獲利能力。我跟你說，當你開始跑數字——我也建議大家自己算算看。你看例如中國的機會規模、可能拿到的市占機率，然後把經濟模型跑一遍，看看這能解鎖多少獲利能力；用這種方式呈現時，會非常令人震撼。而且我也要說，我在講稿裡提到的其他事情也正在開花結果。確實有真正的 GaN on Silicon，我們在業務的其他部分也有非常直接、類似的進展路徑。所以我們非常嚴肅地看待我們的處境。我們正面處理 OpEx 的問題。我們已經實現每季 800 萬美元的 OpEx 分攤／抵減費用。當然在交易完成之前我們不能正式宣告，但自 4 月 24 日起已可計提（accruable）。如果我們能再做成更多這類交易，並為股東解鎖獲利能力，我們就會去做。因為即使在復甦之後，以我們目前的營收基礎，能夠支撐的投入也確實有其上限。好問題。

Our next question comes from Harlan Sur with JPMorgan.

我們下一個問題來自摩根大通（JPMorgan）的 Harlan Sur。

On the JV formation with Goertek, obviously, it's a great channel into the China infrastructure market. Where are the China equipment suppliers in the process of evaluating and qualifying new PA vendors? I would have assumed that they're pretty far along in the process much like you with your existing customers. So what is the timeline for you guys getting qualified, potentially getting qualified with your GaN PAs with some of these potential China equipment guys?

關於與 Goertek 成立 JV，顯然這是一個切入中國基礎建設市場的很棒通路。中國的設備供應商目前在評估與認證新的 PA 供應商這件事上，進展到哪個階段了？我原本以為他們應該已經走到相當後面的階段，就像你們與既有客戶的流程一樣。所以你們的時間表是什麼？你們的 GaN PA 產品要在這些潛在的中國設備廠商那邊完成認證、甚至可能取得認證，大概需要多久？

So that's a great question. Let me emphasize something that's perhaps not obvious. There is a crisis over in China with the local OEMs and the European subsidiaries of Nokia and Ericsson to be able to get access to enough GaN material. I had an opportunity one on one with the top sourcing guy and one of the top basestations OEMs over there, he told me exactly how many wafers he got from whom, what, how and so on. They can't source enough GaN, so I want to be clear here. This is great news for us and the JV and ST. It's great news for anybody who can produce GaN wafers in the short to medium term. There's insatiable demand so the customers are very highly motivated. Very, very highly motivated to qualify our products because they appreciate the fact -- I mean they look at ST it's profoundly different than anything else that they look at for GaN sourcing. It's Marco Monti over at the Mobile World Congress, he said, "Catania does 10,000 8-inch wafers a week. Singapore that they're going to be ramping, is 12,000 per day." So it's totally different scale and I don't anticipate any problem with that. Frankly, the entire problem is they want more sooner so it's -- that's a problem none the less to be honest.

這也是個好問題。我先強調一件可能不那麼顯而易見的事。中國那邊的本土 OEM，以及 Nokia 和 Ericsson 在當地的歐洲子公司，正面臨一場危機：他們無法取得足夠的 GaN 材料。我曾有機會與那邊某家頂尖基地台 OEM 的最高採購負責人一對一交流，他把他從誰那裡拿到多少晶圓、怎麼拿到、細節如何等等，都講得非常清楚。他們就是買不到足夠的 GaN，所以我在這裡要講清楚。這對我們、JV 以及 ST 來說是好消息。對任何能在短到中期內生產 GaN 晶圓的人來說，都是好消息。需求是無止境的，所以客戶的動機非常強。非常、非常強烈地想要認證我們的產品，因為他們明白——我的意思是，他們看 ST 的 GaN 供應能力，跟他們看到的其他任何 GaN 來源都截然不同。在世界行動通訊大會（Mobile World Congress）上，ST 的 Marco Monti 說過：「Catania 每週做 10,000 片 8 吋晶圓。而他們將要擴產的 Singapore，是每天 12,000 片。」這完全是不同的規模，所以我不預期會有任何問題。坦白說，整個問題在於他們希望更快拿到更多供應——老實說，這終究也是個問題。

And you guys have actually got good success with your front-end modules, your FEM modules. Is this part of the JV partnership as well for them to be a channel for your FEM solutions which oftentimes kind of go hand-in-hand with your PA solutions?

而且你們的前端模組（front-end modules），也就是 FEM 模組，其實已經取得不錯的成功。這也會是 JV 合作的一部分嗎？也就是讓他們成為你們 FEM 解決方案的通路——因為 FEM 解決方案往往會和你們的 PA 解決方案相輔相成？

Yes. So it's a very good question. So we kept it clean and easy right now for the whole purpose of export control and so on, not that FEMs are export controlled but right now it's focused on the PAs. There's no question that there are transmit/receive front end modules where the GaN PA is integrated into the FEM and yes, the JV would be the logical channel that we would use to bring those FEMs whether it's transmit/receive or transmit only, because it's different flavors of the same stuff, but yes. So that's for China and the thing I'll emphasize here is half the basestation market, even the Chinese OEMs will tell us to expect half the market outside of China. Because the service providers stick handle share. Obviously, no one wants to end up with a monopolists supply condition in China so there's very healthy business for FEMs that we do with the logical candidates outside of China. But it's -- for the whole 5G build-out it just -- it couldn't be better for us. Our FEM position is almost as strong as our GaN on Silicon position. Probably about half of the bill of materials of a transmit side but it's still half of a very large number is still a very large number. Great question.

是的。這是個非常好的問題。目前我們把架構維持得乾淨、簡單，主要是出於出口管制等因素——不是說 FEM 受出口管制，但目前重點先放在 PA。毫無疑問，有些收發（transmit/receive）的射頻前端模組會把 GaN PA 整合進 FEM；是的，JV 會是我們用來導入這些 FEM 的合理通路，不論是收發一體或僅發射，因為本質上是同一類技術的不同版本，但答案是肯定的。以上是針對中國；我還要強調的是，基地台市場有一半——甚至中國 OEM 也會告訴我們——預期有一半市場在中國以外。因為電信服務商會主導並分配市占。顯然，沒有人希望最後在中國形成供應壟斷的局面，所以我們在中國以外、與合理的合作對象之間，FEM 也會有非常健康的業務。但就整個 5G 建設而言——對我們來說再好不過了。我們的 FEM 競爭地位幾乎和我們的 GaN on Silicon 地位一樣強。它大概占發射端物料清單（BOM）的一半；但「很大的數字的一半」仍然是很大的數字。好問題。

Our next question comes from Richard Shannon with Craig-Hallum.

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

I'll follow-up on a question on Goertek as well. You referred to the uncertainty in accounting of the $8 million payment accruing by quarter. The rest of the payments I'm assuming those will be recorded revenues and would be 100% gross margin? Or if not, could you help us understand that structure?

我也想追問一個關於 Goertek 的問題。你提到按季度累計的 800 萬美元付款在會計處理上存在不確定性。其餘的付款我假設會認列為營收，且毛利率會是 100%？如果不是，能否幫助我們理解其結構？

So the $134.6 million would be consideration would be revenue. The $8 million, even if we do not consolidate, which is not a foregone conclusion I should be clear, we're working through that with our auditors. I wouldn't expect the $8 million would be treated as revenue. I would hope the spirit of it is contra-OpEx but if not contra-OpEx, maybe consulting fees like when we had the Autoliv agreement. Did I answer the question?

所以 1.346 億美元將作為對價並認列為營收。那 800 萬美元，即使我們不進行合併（我需要說清楚，這並非既定結論），我們正在與審計師一起釐清。我不預期這 800 萬美元會被視為營收。我希望其精神是作為營業費用的抵減項（contra-OpEx）；但如果不是抵減營業費用，可能會像我們之前與 Autoliv 的協議那樣，作為顧問費。我有回答到你的問題嗎？

You did answer the question. I'll follow up with another gross margin question specifically on the manufacturing of the GaN on Silicon PAs. Here obviously you're still in the process of getting equipment both for you and your partner but any update on the thought process of gross margins there? Certainly you've talked about that as in line to better will your longer-term gross margin goals of 60%. What's your current expectation there?

你有回答到。我再追問一個毛利率問題，特別是關於在矽基板上製造 GaN 功率放大器（PA）。顯然你和你的合作夥伴仍在取得設備的過程中，但對那裡的毛利率思考有沒有任何更新？你們當然談過這將有助於達成你們長期 60% 的毛利率目標。你目前的預期是什麼？

That's so we have locked in cost structure from our ST arrangement through the cross license. We've got a very strategic relationship with ST in terms of capacity and costs through enabling them and lots of other markets that they care to address that we don't care to address. And we actually have 5-year price targets from Chinese, the most aggressive the aspirational price targets. And I'll just leave it that we can meet those price targets with the ST cross structure and be accretive to our corporate targets. So it's one of the great benefits of manufacturing and mainstream silicon semiconductor factory. It's fantastic cost structure.

我們透過與 ST 的交叉授權安排，已鎖定成本結構。我們與 ST 在產能與成本方面有非常具策略性的關係，透過賦能他們去切入許多他們關心、但我們不打算切入的其他市場。而且我們也取得了來自中國方面的 5 年價格目標，其中最具進取性的、也就是理想化的價格目標。我就先說到這裡：我們可以在 ST 的交叉架構下達成那些價格目標，並且對我們的公司整體目標具有增益（accretive）。所以這是於主流矽半導體晶圓廠製造的一大好處。成本結構非常出色。

Thank you. And this concludes the question-and-answer session. I'll now like to turn the conference back over to management for closing remarks.

謝謝。以上結束問答環節。現在我想把電話會議交回管理層作結語。

Very good. Before closing out today's call, I'd like to mention several upcoming investor events that we'll be attending this quarter. We'll be at the Cowen TMT conference on May 29 in New York, the Stifel conference, June 11, in Boston and the Piper Jaffray Symposium on June 13, also in Boston. If you'd like to arrange a meeting at any of these events, please email us at ir@macom.com. That concludes today's remarks. We look forward to providing further updates on next quarter's earnings call. Operator, you may now disconnect the call.

很好。在結束今天的電話會議之前，我想提及本季我們將參加的幾個即將到來的投資人活動。我們將出席 5 月 29 日在紐約舉行的 Cowen TMT 會議、6 月 11 日在波士頓舉行的 Stifel 會議，以及 6 月 13 日同樣在波士頓舉行的 Piper Jaffray 研討會。如果您希望在上述任何活動安排會面，請寄信至 ir@macom.com 與我們聯繫。以上為今天的結語。我們期待在下季的財報電話會議上提供進一步更新。接線員，您現在可以中斷通話。

Ladies and gentlemen, this concludes today's conference. Thanks for your participation and have a wonderful day.

各位女士、先生，今天的電話會議到此結束。感謝各位的參與，祝您有美好的一天。