格芯 (GFS) 2021 Q3 法說會逐字稿

Thank you for standing by and welcome to GlobalFoundries Third Quarter 2021 Earnings Conference Call. (Operator Instructions) Please be advised that today's conference may be recorded. (Operator Instructions)

感謝各位稍候，歡迎參加格羅方德（GlobalFoundries）2021 年第三季財報電話會議。（接線員指示）請注意，今日的會議可能會被錄音。（接線員指示）

I would now like to hand the conference over to your host, Head of Investor Relations, Sukhi Nagesh. Please go ahead.

現在我想將會議交給本次會議主持人——投資人關係主管 Sukhi Nagesh。請開始。

Thank you, Latif, and good afternoon, everyone, and welcome to GlobalFoundries Third Quarter 2021 Earnings Conference Call, our first as a public company. On the call with me today are Tom Caulfield, our CEO; and Dave Reeder, our CFO.

謝謝你，Latif。各位下午好，歡迎參加格羅方德 2021 年第三季財報電話會議，這也是我們成為上市公司後的首次財報電話會議。今天與我一同出席的有我們的執行長 Tom Caulfield，以及我們的財務長 Dave Reeder。

A short while ago, we released GF's third quarter financial results press release, which is available on our website at investors.gf.com, along with today's accompanying slide presentation. This call is being recorded, and a replay will be made available on our Investor Relations web page.

稍早我們已發布 GF 第三季財務結果新聞稿，可於我們的網站 investors.gf.com 取得，並附有今日的簡報投影片。本次電話會議將被錄音，並會在我們投資人關係網頁提供重播。

During this call, we will present both IFRS as well as non-IFRS financial measures. The most directly comparable IFRS measures and reconciliations for our non-IFRS measures are available in today's press release and accompanying slides. I would remind you that our financial measures are unaudited.

在本次電話會議中，我們將同時呈報 IFRS 以及非 IFRS 的財務衡量指標。與我們非 IFRS 指標最直接可比的 IFRS 指標及其調節表，已載於今日的新聞稿與隨附投影片中。提醒各位，我們的財務數據尚未經審計。

Certain statements on today's call may be deemed to be forward-looking statements. Such statements can be identified by terms such as believe, expect, intend, anticipate and may. You should not place undue reliance on forward-looking statements. Actual results may differ materially from these forward-looking statements, and we do not undertake any obligation to update any forward-looking statements we may make today. For more information about factors that may cause actual results to differ materially from forward-looking statements, please refer to the press release we issued today as well as risks and uncertainties described in our SEC filings, including in the section under the caption Risk Factors in our financial prospectus filed with the SEC on October 29, 2021, in connection with our IPO.

今日電話會議中的某些陳述可能被視為前瞻性陳述。此類陳述可由 believe、expect、intend、anticipate 以及 may 等用語辨識。您不應過度依賴前瞻性陳述。實際結果可能與這些前瞻性陳述存在重大差異，我們亦不承擔更新今日所作任何前瞻性陳述的義務。關於可能導致實際結果與前瞻性陳述出現重大差異的因素，請參閱我們今日發布的新聞稿，以及我們向美國證券交易委員會（SEC）提交的文件中所述之風險與不確定性，包括我們於 2021 年 10 月 29 日就首次公開募股（IPO）向 SEC 提交之財務招股說明書中「Risk Factors」標題下的章節。

We will begin today's call with Tom providing a summary update on our end markets, capacity expansion and technologies. Following which, Dave will provide details on our third quarter financial results and fourth quarter guidance. We will then open the call up for questions. (Operator Instructions)

我們將以 Tom 對終端市場、產能擴張與技術的摘要更新作為開場。接著，Dave 將提供第三季財務結果的細節以及第四季指引。之後我們將開放提問。（接線員指示）

With that, I will now turn the call over to Tom for his prepared remarks.

接下來，我將把電話會議交給 Tom 進行事先準備的發言。

Thank you, Sukhi. Well, welcome, everyone, to our first earnings conference call as a public company. Our IPO on October 28 was an important milestone for GlobalFoundries. It was a culmination of over a decade of work to build an at-scale global semiconductor foundry with strong technology differentiation.

謝謝你，Sukhi。各位好，歡迎參加我們成為上市公司後的首次財報電話會議。我們於 10 月 28 日完成 IPO，這是格羅方德的重要里程碑。這是十多年努力的成果，旨在打造一家具規模的全球半導體晶圓代工廠，並擁有強勁的技術差異化。

In 2018, we drove a fundamental change in our strategy to become a more relevant and more importantly, a more vital contributor to this industry by becoming a profitable and sustainable business. We strengthened the management team. We refocused our investments in R&D and CapEx to differentiated feature-rich solutions. We forged stronger customer partnerships, and we streamlined our manufacturing footprint and cost structure. Our strategic initiatives are now well in place, and we anticipate driving profitable growth.

2018 年，我們推動策略上的根本性轉變，讓我們在這個產業中更具相關性，更重要的是，成為更關鍵的貢獻者——也就是成為一家獲利且可持續的企業。我們強化了管理團隊。我們將研發與資本支出（CapEx）投資重新聚焦於具差異化、功能豐富的解決方案。我們與客戶建立更緊密的合作夥伴關係，並精簡製造版圖與成本結構。我們的策略性舉措如今已全面到位，並預期將推動獲利成長。

While we are seeing the initial results of this in 2021, we are really at the beginning of this journey. We expect this will become a more -- this will become more apparent in 2022 and beyond as our revenue will continue to grow with the capacity investments we are making. And as a result of this increasing scale, we expect even faster growth in our margins and earnings. We have significant business visibility and certainty with customer long-term agreements. This gives us confidence that the fundamentals of our business will continue to improve at a rapid pace over the next 3 to 5 years.

雖然我們在 2021 年已看到初步成果，但我們其實才剛踏上這段旅程的起點。我們預期在 2022 年及之後會更加明顯，因為隨著我們正在進行的產能投資，我們的營收將持續成長。由於規模擴大，我們預期毛利率與獲利將以更快速度成長。我們透過與客戶簽訂長期協議，具備高度的業務能見度與確定性。這讓我們有信心，在未來 3 到 5 年內，我們的業務基本面將持續以快速步伐改善。

Now while we are excited and proud of our IPO, we recognize it is ultimately just the first step in a much longer journey. We take our responsibility and duty to create long-term value for investors, our customers, our employees and our many other stakeholders with a heightened sense of urgency. We'll do this by continuing to focus on creating innovative solutions for our customers by partnering closely with them, embracing the diversity of our team and delivering services and products that allow our customers to win in the markets they serve.

儘管我們對 IPO 感到興奮與自豪，但我們也明白，這終究只是更長旅程的第一步。我們以更強烈的急迫感，承擔為投資人、客戶、員工以及眾多其他利害關係人創造長期價值的責任與義務。我們將透過持續專注於為客戶打造創新解決方案、與客戶緊密合作、擁抱團隊多元性，並交付能讓客戶在其所服務市場中勝出的服務與產品來達成。

Now moving on to our third quarter. We are pleased to report a quarter of strong top line and profitability growth, demonstrating the continued success of our strategy. Third quarter revenue grew 5% quarter-on-quarter, driven by higher wafer output than the -- and the continued improvement in mix as our differentiated solutions become a larger portion of our total business.

接著談第三季。我們很高興報告本季營收與獲利能力皆強勁成長，展現我們策略的持續成功。第三季營收季增 5%，主要受更高的晶圓產出所帶動，並且隨著我們具差異化的解決方案在整體業務中占比提高，產品組合也持續改善。

Third quarter adjusted earnings per share came in at $0.07. Now David will provide more details on the financials in just a moment. But first, let me give a summary of third quarter revenue by our end markets.

第三季調整後每股盈餘為 0.07 美元。David 稍後將提供更多財務細節。不過在此之前，我先簡要說明第三季按終端市場劃分的營收概況。

First, in our smart mobile end market, which comprise roughly 50% of our third quarter revenue, we saw strong year-over-year growth of roughly 45%, driven mostly by the continued ramp of our single-source design wins in the growing markets such as 5G RF front-end modules for mobile handsets, for image sensors and WiFi 6. Our differentiated technologies continue to do well in the 5G sub-6 gigahertz market. We are still, though, in the early innings of the industry transition to 5G with market segments -- I'm sorry, with market estimates of doubling of 5G handsets to nearly 500 million units this year.

首先，在智慧行動（smart mobile）終端市場方面，該市場約占我們第三季營收的 50%。我們看到約 45% 的年增成長，主要由我們單一來源（single-source）設計導入案的持續放量所帶動，涵蓋成長中的市場，例如行動手機用的 5G RF 前端模組、影像感測器與 WiFi 6。我們的差異化技術在 5G sub-6 GHz 市場持續表現良好。不過，我們仍處於產業轉向 5G 的早期階段；市場預估今年 5G 手機出貨將倍增至接近 5 億支。

Next, our communications infrastructure and data center end market, which constituted approximately 17% of our third quarter revenue, saw over 30% year-over-year growth, driven by a combination of share gains by our customers, continued strength in the enterprise data center, cellular infrastructure and RF transceiver markets.

接下來是通訊基礎設施與資料中心終端市場，約占我們第三季營收的 17%。該市場年增超過 30%，成長動能來自多項因素的組合，包括我們客戶的市占提升、企業級資料中心的持續強勁需求，以及行動通訊基礎設施與 RF 收發器市場的強勢表現。

Moving on to home and industrial IoT end market. Third quarter revenue was roughly 13% of the total and grew approximately 37% year-over-year. The growth was driven by a combination of higher ASPs as well as the ramp of IoT products from some of our key customers for applications such as digital TVs, WiFi and secure contactless transactions. In addition, we saw broad-based growth for MCUs in the quarter.

再來是家用與工業物聯網（IoT）終端市場。第三季營收約占總營收的 13%，年增約 37%。成長來自 ASP 提升以及部分關鍵客戶之 IoT 產品放量，應用包含數位電視、WiFi 與安全的非接觸式交易。此外，本季 MCU 亦呈現廣泛性的成長。

Touching next on automotive. Revenue in this end market was approximately 6% of our total third quarter revenue and grew almost 4x from the year prior period. The strong growth in our automotive end markets was driven by the ramp of new designs that have been in development and qualification over the past years. GF's automotive products are now going into a variety of automotive uses such as in-vehicle comfort, safety, sensing and battery management solutions in EVs.

接著談汽車。該終端市場營收約占我們第三季總營收的 6%，較去年同期成長近 4 倍。汽車終端市場的強勁成長，主要由過去數年持續開發與認證的新設計開始放量所帶動。GF 的車用產品目前正導入多種車用情境，例如車內舒適性、安全、感測，以及電動車（EV）的電池管理解決方案。

The chip shortage in the auto industry has accelerated demand for many of our customers who have entered into long-term agreements with GF to ensure supply continuity for their new products that ramp over the next 3 to 5 years. We are very excited about our strong traction in the automotive end market.

汽車產業的晶片短缺加速了許多客戶的需求，這些客戶已與 GF 簽訂長期協議，以確保其未來 3 到 5 年放量的新產品供應連續性。我們對於在汽車終端市場取得的強勁進展感到非常振奮。

In our compute end market, revenue was roughly 70 -- 7% of total and declined year-over-year as expected as some of our customers designed to continue to transition to smaller nodes. We continue to forecast a decline in this end market for the first half of 2022 and then see stabilization as an improvement in the second half of 2022 from ramps from newer high-margin customer designs. We expect, however, the decline in revenue in this end market to be more than compensated with growth in other end markets.

在運算（compute）終端市場方面，營收約占總營收的 7%，且如預期般年減，因為部分客戶持續將設計轉移至更小的製程節點。我們仍預期 2022 年上半年該終端市場將下滑，之後在 2022 年下半年，隨著較新且高毛利的客戶設計開始放量，將出現改善並趨於穩定。不過，我們預期該終端市場的營收下滑，將可由其他終端市場的成長所充分抵銷。

Next, I'd like to provide a brief update on our ongoing capacity expansion plans. Overall, our global installed capacity will increase approximately 4% from 3Q to 4Q, which is approximately 12% increase from the fourth quarter of last year. Our installed capacity in our Fab 1 facility in Dresden, Germany is increasing output by approximately 16% from 3Q to 4Q this year, and this also represents about a 15% expansion at that facility from a year ago. All of this expansion in capacity is in support of customer demand for our differentiated technologies such as 22FDX, 20-nanometer ISP and our BiCMOS technology.

接下來，我想就我們持續進行的產能擴張計畫提供簡要更新。整體而言，我們全球已裝機產能將自第三季到第四季增加約 4%，相當於較去年第四季增加約 12%。我們位於德國德勒斯登（Dresden）的 Fab 1 廠區已裝機產能，今年自第三季到第四季的產出將提升約 16%，這也代表該廠區較一年前擴張約 15%。上述所有產能擴張，皆是為了支援客戶對我們差異化技術的需求，例如 22FDX、20 奈米 ISP，以及我們的 BiCMOS 技術。

Also, construction on our Phase 1 module expansion in Singapore remains on track with equipment slated to go in the facility in the second half of 2022 to support first production out in the first half of 2023. In addition to our ongoing capacity expansion, we continue to make strong progress in enhancing our differentiated technologies. For instance, in 3Q, we completed automotive Grade 1 qualification of our 22FDX RF and millimeter wave platform, including reference IPs for complex analog and RF blocks; a complete ecosystem of design services, including IP providers, EDA vendors and turnkey services. This feature-rich platform is targeting automotive smart sensors and processors.

此外，我們在新加坡第一期模組擴建的施工仍按計畫進行，設備預計將於 2022 年下半年進駐廠房，以支援 2023 年上半年首次量產。除了持續的產能擴張之外，我們也在強化具差異化的技術方面持續取得重大進展。例如在第三季，我們完成了 22FDX RF 與毫米波平台的車規 Grade 1 認證，其中包含用於複雜類比與 RF 模組的參考 IP；以及完整的設計服務生態系，涵蓋 IP 供應商、EDA 廠商與交鑰匙服務。這個功能豐富的平台鎖定汽車智慧感測器與處理器。

For example, Bosch is a lead customer for ADAS radar SoCs, and they're leveraging our 22FDX platform for their next-generation radar systems. Our technology team also delivered the first functional resistive RAM bit cell, an enhanced version of FDX we referred to as 22FDX+, which is targeted for Tier 1 customers seeking next-generation wireless secure transaction capability in mobile, IoT and automotive end markets. These achievements are truly differentiated and are allowing our customers to win with ultra willpower in analog, RF and MCU designs.

例如，博世（Bosch）是 ADAS 雷達 SoC 的主要客戶，他們正運用我們的 22FDX 平台打造其下一代雷達系統。我們的技術團隊也交付了第一個可運作的電阻式 RAM（ReRAM）位元單元，這是我們稱為 22FDX+ 的 FDX 強化版本，目標客戶為 Tier 1 廠商，尋求在行動裝置、IoT 與車用終端市場中具備下一代無線安全交易能力。這些成果確實具差異化，並讓我們的客戶能以超強的意志力在類比、RF 與 MCU 設計上勝出。

In silicon photonics, our 45CLO platform delivered first customer prototypes that demonstrated in 8 lambda, 32-gigabit per second optical link with extremely low bid errors. We are the technology leader in silicon photonics as we are the only provider of integrated CMOS, RF SOI and optical devices in a monolithic solution. This unique capability will drive a whole new upgrade of connectivity in data centers over the next decade.

在矽光子方面，我們的 45CLO 平台交付了首批客戶原型，展示了 8 lambda、每秒 32Gb 的光鏈路，且位元錯誤率極低。我們是矽光子技術領導者，因為我們是唯一能以單晶整合方案提供整合式 CMOS、RF SOI 與光學元件的供應商。這項獨特能力將在未來十年推動資料中心連線能力的全新升級。

So to summarize, we are seeing strong growth from our customers in the end markets we serve, and we are prudently and in partnership expanding our capacity to serve their needs and making great progress in accelerating our differentiated technologies for the future.

總結來說，我們看到所服務終端市場中的客戶需求強勁成長；同時我們也以審慎方式並與合作夥伴共同擴充產能以滿足其需求，並在加速我們面向未來的差異化技術方面取得重大進展。

With that, let me turn the call over to David to provide the financial details for the third quarter and also provide you our guidance for the fourth quarter. David?

接下來，我把電話交給 David，請他提供第三季的財務細節，並給各位第四季的展望指引。David？

Thank you, Tom. And let me also express my excitement for GF's IPO. What an incredible milestone for the company and an important funding mechanism for our future capacity expansion. Now on to our third quarter.

謝謝你，Tom。我也想表達我對 GF IPO 的興奮之情。這對公司而言是極其重要的里程碑，也是我們未來產能擴張的重要資金機制。接著談第三季。

Our third quarter results came in at or above the high end of our -- of the financial ranges we provided last month in our prospectus. Our third quarter revenue was approximately $1.7 billion, which increased 5% sequentially, driven by higher wafer shipments and mix. We shipped approximately 609,000 300-millimeter equivalent wafers in the quarter, which was up about 2.5% on a sequential basis.

我們第三季的結果落在、或高於我們上個月在招股說明書中提供之財務區間的高端。第三季營收約為 17 億美元，季增 5%，主要由更高的晶圓出貨量與產品組合所帶動。本季我們出貨約 609,000 片 300 毫米等效晶圓，季增約 2.5%。

Wafer revenue from our end markets accounted for approximately 92% of total revenue. Non-wafer revenue, which is typically between 5% and 10% of total revenue, accounted for approximately 8% of revenue for the quarter. As a reminder, non-wafer revenue includes reticles, nonrecurring engineering multiproduct wafers and other foundry services.

來自終端市場的晶圓營收約占總營收的 92%。非晶圓營收通常占總營收的 5% 至 10%，本季約占 8%。提醒一下，非晶圓營收包括光罩（reticles）、一次性工程費（NRE）的多產品晶圓，以及其他晶圓代工服務。

For the remainder of the call, including fourth quarter guidance, I will reference adjusted metrics. Our adjusted metrics exclude stock-based compensation.

在接下來的說明（包含第四季指引）中，我將引用調整後指標。我們的調整後指標排除以股票為基礎的薪酬。

For the third quarter, we delivered adjusted gross profit of $306 million, which translates into approximately 18% adjusted gross margin. The 155-basis-point sequential improvement was primarily driven by better fixed cost absorption and modestly improved ASPs.

第三季我們的調整後毛利為 3.06 億美元，約當 18% 的調整後毛利率。毛利率季增 155 個基點，主要由更佳的固定成本吸收，以及 ASP 小幅改善所帶動。

R&D expense, excluding stock-based compensation, was approximately $111 million, about $10 million lower than the previous quarter. The sequential decline was primarily due to lower fab technology start-up costs and increased customer-funded NRE. Excluding stock-based compensation, the sequential decline in R&D was mostly offset by a $9 million increase in SG&A expense. This increase was primarily due to higher employee costs and IPO-related expenses.

研發費用（排除以股票為基礎的薪酬）約為 1.11 億美元，較前一季少約 1,000 萬美元。季減主要因晶圓廠技術啟動成本降低，以及客戶出資的 NRE 增加。排除以股票為基礎的薪酬後，研發費用的季減大多被 SG&A 費用增加 900 萬美元所抵銷；該增加主要來自較高的員工成本與 IPO 相關費用。

Total operating expenses, excluding stock-based compensation, for the third quarter were approximately $225 million and roughly flat compared to the previous quarter. We delivered adjusted operating profit of approximately $81 million for the quarter, which translates into 5% adjusted operating margin, a sequential improvement of 224 basis points.

第三季總營業費用（排除以股票為基礎的薪酬）約為 2.25 億美元，與前一季大致持平。本季調整後營業利益約為 8,100 萬美元，約當 5% 的調整後營業利益率，季增 224 個基點。

Third quarter net interest expense was approximately $27 million, and we incurred a tax expense of approximately $22 million in the quarter. We also had approximately $2 million of other income primarily related to a gain on asset sales. The combination of these results delivered third quarter adjusted net income of approximately $34 million, which is $0.07 of adjusted earnings per share on a basic share count of about 500 million shares.

第三季淨利息費用約為 2,700 萬美元，且本季認列所得稅費用約 2,200 萬美元。我們也有約 200 萬美元的其他收益，主要與資產出售收益相關。綜合上述結果，第三季調整後淨利約為 3,400 萬美元；以約 5 億股的基本股數計算，調整後每股盈餘為 0.07 美元。

We delivered record third quarter adjusted EBITDA of approximately $505 million. EBITDA grew $39 million sequentially on $80 million of incremental revenue growth, an almost 50% fall through.

我們第三季調整後 EBITDA 創下紀錄，約為 5.05 億美元。在營收增加 8,000 萬美元的帶動下，EBITDA 季增 3,900 萬美元，增量利潤落入率接近 50%。

Let me now provide some key balance sheet and cash flow metrics. Cash flow from operations for the quarter was approximately $1.1 billion and included approximately $574 million of customer prepayments. Gross CapEx for the quarter was about $392 million or roughly 23% of revenue. We ended the quarter with a little over $1 billion in cash and cash equivalents, an increase of more than $200 million from the previous quarter.

接著提供幾項資產負債表與現金流的關鍵指標。本季營運活動現金流約為 11 億美元，其中包含約 5.74 億美元的客戶預付款。本季資本支出（CapEx）總額約為 3.92 億美元，約占營收的 23%。季末我們的現金及約當現金略高於 10 億美元，較前一季增加超過 2 億美元。

Now as you know, we raised approximately $1.5 billion from our IPO at the end of October. And as a result, we are very well capitalized. We expect to use the majority of the proceeds from the IPO for our capacity expansion plans to meet robust customer demand.

如各位所知，我們在 10 月底的 IPO 募得約 15 億美元。因此，我們的資本結構非常充裕。我們預期將把 IPO 募集資金的大部分用於產能擴張計畫，以滿足強勁的客戶需求。

Next, let me provide you with our outlook for the fourth quarter. We expect revenue to be between $1.8 billion and $1.83 billion. We expect gross profit to be between $335 million and $350 million. And adjusted gross profit is expected to be between $344 million and $359 million. We expect operating profit to be between $55 million and $75 million. And adjusted operating profit is expected to be between $92 million and $112 million.

接下來提供第四季展望。我們預期營收介於 18 億至 18.3 億美元之間。我們預期毛利介於 3.35 億至 3.50 億美元之間；調整後毛利預期介於 3.44 億至 3.59 億美元之間。我們預期營業利益介於 5,500 萬至 7,500 萬美元之間；調整後營業利益預期介於 9,200 萬至 1.12 億美元之間。

Excluding share-based compensation, for the fourth quarter, we expect R&D to increase moderately on a sequential basis, primarily due to higher fab technology start-up costs. SG&A is also expected to increase moderately on a sequential basis, primarily due to expenses related to our IPO and becoming a public company. We expect net income to be between $13 million and $33 million and adjusted net income to be between $50 million and $70 million. On a basic share count of approximately 525 million, we expect basic earnings per share for the fourth quarter to be between $0.02 and $0.06 and adjusted basic earnings per share to be between $0.09 and $0.13.

排除以股份為基礎的薪酬後，第四季我們預期研發費用將較前一季溫和增加，主要因晶圓廠技術啟動成本較高。SG&A 也預期將較前一季溫和增加，主要因與 IPO 以及成為上市公司相關的費用。我們預期淨利介於 1,300 萬至 3,300 萬美元之間，調整後淨利介於 5,000 萬至 7,000 萬美元之間。以約 5.25 億股的基本股數計算，我們預期第四季基本每股盈餘介於 0.02 至 0.06 美元之間，調整後基本每股盈餘介於 0.09 至 0.13 美元之間。

For the fourth quarter, we expect adjusted EBITDA to be between $510 million and $530 million. And finally, we expect share-based compensation to be approximately $9 million in cost of revenue, $2 million in R&D and $26 million in SG&A for a total share-based compensation of approximately $37 million.

第四季我們預期調整後 EBITDA 介於 5.10 億至 5.30 億美元之間。最後，我們預期以股份為基礎的薪酬約為：營收成本 900 萬美元、研發 200 萬美元、SG&A 2,600 萬美元，合計約 3,700 萬美元。

With that, we can open up the call for Q&A. Operator?

接下來我們可以開放 Q&A。接線員？

(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。

Congratulations on the strong results and guide on your first quarter as a public company. At the time of your filing, your long-term agreements covered about $20 billion of forward revenues, which when you combine that with current purchase orders, it covered about 85% of your revenues over the next 3 years.

恭喜你們作為上市公司第一季就交出強勁的成績與指引。在你們提交申報文件時，你們的長期協議涵蓋約 200 億美元的未來營收；若再加上目前的採購訂單，已涵蓋未來 3 年約 85% 的營收。

Since then, obviously, the supply-demand gap in the industry has gotten more severe and I think further motivating your customers to lock in assurance of supply over multiple years. Can you guys just give us an update on the $20 billion of LTA, what does that number look like today? Does it extend beyond 2024? And is it balanced across all of your end market segments?

自那之後，很明顯產業的供需缺口變得更嚴重，我認為這也進一步促使你們的客戶在多年期內鎖定供應保障。你們能否更新一下這 200 億美元的 LTA：目前這個數字看起來如何？是否延伸到 2024 年之後？以及在你們所有終端市場區隔之間是否分布均衡？

So first, Harlan, thank you for joining us. I'll let David start with that, and maybe I'll add some color at the end.

所以首先，Harlan，感謝你加入我們。我先讓 David 從這裡開始，最後我可能再補充一些說明。

Harlan, since our road show, we have made progress on our LTAs. I think at the time of road show, we were talking about roughly $2.5 billion of customer prepayments and fund access fees. And today, that number's a little bit north of $3 billion of commitments from customers with respect to prepayments and access fees. So some good progress there.

Harlan，自從我們的路演以來，我們在 LTA 上已取得進展。我記得在路演當時，我們談到客戶預付款與基金使用費（fund access fees）合計大約 25 億美元。而今天，這個數字在客戶就預付款與使用費所作出的承諾方面，已略高於 30 億美元。所以這方面有不錯的進展。

With respect to revenues, we've also continued to make progress on our LTA revenues where we've made good progress. We're north of $20 billion now. We're pleased to report that we're north of $20 billion now. We're not going to provide more color than that at this time other than we're very excited about the trajectory of the business.

就營收而言，我們在 LTA 營收方面也持續取得進展，進度良好。我們現在已超過 200 億美元。我們很高興報告，我們現在已超過 200 億美元。除了我們對業務走勢非常振奮之外，目前不會提供比這更多的細節。

Tom, anything you'd add?

Tom，你有什麼要補充的嗎？

Yes, I would just add. These long-term agreements are very important for us because of the visibility it gives to our business, but we always need to keep a certain amount of flexibility because some of our customers may need more than what they've signed up for. And so we can't have 100% of our business in any given year signed to a long-term agreement. We want to have -- leave some flexibility to be able to respond to our customers.

有的，我想補充一下。這些長期協議對我們非常重要，因為它們為我們的業務提供可見度；但我們也必須保留一定程度的彈性，因為部分客戶可能需要超出他們所簽約的量。因此，我們不可能在任何一年把 100% 的業務都簽成長期協議。我們希望保留一些彈性，以便能夠回應客戶需求。

So as your customers' customers figure out how to prevent a similar supply-demand dislocation, like the one that we're currently going through, they're thinking and rethinking their supply chain strategies. I think your recent partnership with Ford sort of encompasses this, right? They talked about potentially securing supply for their chip suppliers, but they also talked about potentially doing some in-house chip design as well.

所以當你們客戶的客戶在思考如何避免類似我們目前正在經歷的供需錯配時，他們正在思考並重新思考其供應鏈策略。我認為你們最近與 Ford 的合作某種程度上就涵蓋了這點，對吧？他們談到可能為其晶片供應商確保供應，但也談到可能進行一些內部晶片設計。

You already have other end customers, Cisco, Microsoft, Amazon, as direct customers. Do you guys see this trend continuing? In other words, more cloud titans, more auto OEMs, more consumer device OEMs coming to the table to talk to you? And how many of these discussions are you having? And how do you see this benefiting both GlobalFoundries and the industry longer term?

你們已經有其他終端客戶，例如 Cisco、Microsoft、Amazon，作為直接客戶。你們是否看到這個趨勢會持續？換句話說，會有更多雲端巨頭、更多汽車 OEM、更多消費性裝置 OEM 來與你們洽談嗎？你們正在進行多少這類討論？你們如何看待這將在長期上同時讓 GlobalFoundries 與整個產業受益？

Well, let me start with a high-level view of this is today, product companies differentiated by integrating technology with firmware and software. And now they're seeing that there's real leverage. Others have done this already very successfully. But there's leverage owning some of the silicon design as they integrate the technology. And I think that's the trend you're starting to see. Some of the more famous players in the mobility space kind of patented that technique, and others are following it.

我先從高層次的觀點談起：如今，產品公司是透過將技術與韌體、軟體整合來實現差異化。現在他們看到，當他們在整合技術時，擁有部分矽晶設計（silicon design）會帶來真正的槓桿效益。其他公司早已非常成功地做到這點。我認為這就是你開始看到的趨勢。在行動（mobility）領域，一些更知名的玩家某種程度上已把這種做法「制度化」，而其他人正在跟進。

Taking that down to, I think you're seeing this in automotive, as semiconductor is going to become the defining line for differentiation in the automotive experience, I believe you're seeing automotive companies want to become, what I would call, more silicon aware.

再把這點落到更具體的層面，我認為你在汽車領域也看得到：隨著半導體將成為汽車體驗差異化的分界線，我相信你會看到汽車公司希望變得更「懂矽」（silicon aware）。

And that's exactly what Ford is doing in their MOU they signed with GF. They want to first make sure that they understand the supply-demand dynamics. They want visibility to supply. They want to make sure they influence the technology road maps for foundries, so that the features that we create align with their needs. And then also making sure just like our customers, that whether they design the product or not, making sure that they reserve capacity to make sure they can build their -- the vehicles they want. And that's the essence of what I think you're going to see -- that's the essence of that agreement and that you'll see more of that in the industry and more of that from GF.

而這正是 Ford 在與 GF 簽署的 MOU 中所做的事。他們首先希望確保自己理解供需動態；他們希望對供應有可見度；他們希望確保能影響晶圓代工廠的技術路線圖，讓我們打造的功能與他們的需求一致。接著也要確保，就像我們的客戶一樣，不論他們是否自行設計產品，都要確保預留產能，以確保他們能生產他們想要打造的——車輛。我認為這就是你將會看到的——這就是該協議的核心，你也會在產業中看到更多這樣的情況，以及從 GF 這邊看到更多這樣的合作。

David, anything you'd add to that?

David，你有什麼要補充的嗎？

Well said, tom. We're excited about the opportunity.

Tom 說得很好。我們對這個機會感到非常興奮。

Our next question comes from Ross Seymore of Deutsche Bank.

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

I'll echo the congratulations on the first quarter out of the gate. Tom, I just want to ask a high-level question to you about customer behavior. Investors are just a little concerned about the sustainability demand. We know that you have a great batch of LTAs and prepayments, great visibility for next year, and those deals are set in stone, so to speak. But customer behavior on lead times expedites those sorts of things, at this point in the cycle, can tend to get a little bit more volatile. So I want to see if there's any update that you've seen or any changes you've seen in how your customers are acting on the demand side.

我也要呼應對第一季開局表現的祝賀。Tom，我想從高層次問你一個關於客戶行為的問題。投資人對需求的可持續性有些擔憂。我們知道你們有一批很棒的 LTA 與預付款，對明年的可見度很高，而且那些交易可以說是板上釘釘。但在這個景氣循環的此刻，客戶在交期、加急（expedites）等方面的行為，往往會變得更波動一些。我想了解你是否看到任何最新情況，或你是否看到客戶在需求端的行為有任何變化。

I think we -- first of all, thanks for joining and Ross, for your question. I think we sit in a very unique position as a foundry where we get to see what's going on in a lot of the end markets, in particular, the 5 we serve. And I have to tell you, not anything noticeable from our seat from what we talked about in the road show. We are still dealing with shortages of what we can supply to our key customers in 2022 and figuring out how we could produce more, how do we do the right balance and allocations. So there's nothing that we see in the industry for 2022 that would suggest the frothiness that you described is changing.

我想我們——首先，感謝你加入，也謝謝 Ross 的提問。我們作為晶圓代工廠處在一個非常獨特的位置，能看到許多終端市場正在發生的事情，特別是我們服務的 5 個市場。我必須告訴你，從我們的角度來看，與我們在路演時談到的相比，沒有任何明顯的變化。我們在 2022 年仍在面對供應不足，無法滿足關鍵客戶的需求，並在思考如何能生產更多、如何做出正確的平衡與配給。因此，就 2022 年而言，我們在產業中沒有看到任何跡象顯示你所描述的那種「過熱」正在改變。

I have to tell you, though, I would welcome a little bit of the softening of demand just a little bit to make some of the closure of that gap. And so what we're doing, I think, more importantly, is the investments we're making over 2021 and 2022 to close that supply-demand gap for the industry. I hope that helps.

不過我必須說，我其實會歡迎需求稍微放緩一點點，好讓那個缺口能更容易收斂。因此，我認為更重要的是，我們在 2021 年與 2022 年所做的投資，目的是為整個產業縮小供需缺口。希望這能回答你的問題。

It definitely does. One then as a follow-up for David was on the gross margin side of things. I know you talked about mix and scale, which are kind of good things but relatively obvious as you grow. How do we think about the growth -- gross margin trajectory going forward? Any more color on the 1.5 points you got 3Q to 4Q and then going forward beyond that?

確實有幫助。接著我想追問 David 一個關於毛利率的問題。我知道你談到產品組合與規模，這些是好事，但隨著成長也相對直觀。我們該如何看待未來毛利率的走勢？能否對你們從第三季到第四季提升的 1.5 個百分點，以及之後的走勢，再多提供一些說明？

Absolutely, Ross. And before we move to that question, let me just add one comment and expand upon Tom's answer from the last one. Look, we've signed about 25 long-term customer agreements. We signed even several since the road show, which, as you know, was only a month ago. And we've added more than $500 million of customer prepayments and access fees that are committed since the road show. And so we see some pretty strong demand in front of us. We are capacity limited, and we are working diligently every day to get new tooling in and get factories ramped and online, so that we can produce more wafers for our customers, which is a good lead-in related to the gross margin question that you just asked.

當然可以，Ross。在我們轉到那個問題之前，讓我先補充一點，並延伸 Tom 對上一題的回答。你看，我們已簽署大約 25 份長期客戶協議。自路演以來我們甚至又簽了幾份——如你所知，路演也才一個月前。而且自路演以來，我們新增了超過 5 億美元、已承諾的客戶預付款與使用費。因此，我們看到眼前有相當強勁的需求。我們受到產能限制，並且每天都在努力導入新設備、讓工廠爬坡並上線，以便為客戶生產更多晶圓——這也正好銜接到你剛才問的毛利率問題。

Gross margin going forward, we talked about 2021 really being a year in which depreciation moderates and we really start to take advantage of some of the fixed cost absorption as we tool out this manufacturing footprint that has been untooled since our pivot. And so as we start to tool out those facilities, the biggest one being Dresden, as we tool those facilities out, we're actually getting better cost absorption than we expected. That's the positive news about the gross margin in third quarter. It's also part of the reason why we're positive on the trajectory of gross margin in fourth quarter as well.

關於未來毛利率，我們談到 2021 年確實是一個折舊趨於緩和的一年，而我們也開始真正受益於固定成本吸收（fixed cost absorption），因為我們在這個製造版圖上進行設備配置（tool out）；在我們策略轉向（pivot）之後，這些產能先前並未完成設備配置。當我們開始為這些廠區配置設備時，其中最大的一個是德勒斯登（Dresden），隨著我們為這些廠區配置設備，我們實際上獲得了比預期更好的成本吸收。這是第三季毛利率的正面消息；這也是我們對第四季毛利率走勢持正面看法的部分原因。

Our next question comes from Vivek Arya of Bank of America Securities.

我們的下一個問題來自美銀證券（Bank of America Securities）的 Vivek Arya。

Congratulations to the team on going public. Tom, you described some of the customer prepayments that's kind of very strong endorsement about the visibility. I'm wondering what kind of dialogue you're having with the government, right, whether it's in the U.S. here with the chip side or whether it's with the governments across Europe and in Singapore, because there is a lot of motivation to onshore and bring on more sovereign capacity. And I'm wondering what kind of dialogue you're having, when do you start to see the benefit of that?

恭喜團隊完成上市。Tom，你提到一些客戶預付款，這對可見度而言是非常強的背書。我想知道你們與政府之間正在進行什麼樣的對話——不論是在美國這邊與晶片相關的政策，或是在歐洲各國以及新加坡的政府，因為各方都有很強的動機要把產能在地化、建立更多主權產能。我想了解你們正在進行什麼樣的對話，以及你們何時開始看到這些帶來的效益？

Well, let me start with our first expansion that we're doing in Singapore. As we spoke about in the road show, that's a $4 billion Phase 1 expansion. And it has significant government support, as we've talked about in the road. So governments are already participating when I call the new economic models to create the kind of capacity we need in this industry, where it's customers, foundries like GF and governments that participate to create the right economics to add capacity.

好，那我先從我們在新加坡正在進行的第一個擴產談起。如同我們在路演中提到的，那是一項40億美元的第一期擴產。而且它獲得了相當可觀的政府支持，正如我們在路演中談到的。因此，當我所稱的新經濟模式正在形成時，各國政府已經在參與，以打造我們在這個產業所需要的產能——也就是由客戶、像GF這樣的晶圓代工廠，以及政府共同參與，建立正確的經濟條件來新增產能。

The U.S. continues to make progress. There's now a fab bill that talks about 25% rebates on equipment. That's working its way through. Still very positive that you see we'll get funded. I think the timing in the U.S. is -- there's a couple of things that need to get out of the way to clear the hurdle like the debt limit and getting the build back better, kind of funded and clear the deck to get the chips bill funded.

美國方面持續取得進展。現在已有一項晶圓廠法案，內容提到對設備提供25%的回饋（rebates）。這項法案正在推進中。整體仍然非常正面，你可以看到我們將會獲得資金。我認為美國的時程是——有幾件事情需要先排除障礙，例如債務上限，以及讓「重建更好（Build Back Better）」之類的方案獲得資金並把相關事項釐清，才能清出空間讓晶片法案獲得資金。

And I think that is more targeted. The best crystal ball I have is Q1 of next year. I'd be pleasantly surprised if it got pulled into this year, but we're getting down to the short strokes here, as we -- first day of December is tomorrow.

而且我認為那會更具針對性。我手上最好的「水晶球」判斷是明年第一季（Q1）。如果能提前到今年，我會很驚喜，但我們現在已經進入最後階段了，因為——明天就是12月的第一天。

And then Europe continues to have high goals and ambitions to go fund capacity onshore, and we're waiting for more movement in funding of what's called IPCAI 2. Nothing new to report on that front from our road show just a month ago.

接著歐洲方面，仍然有很高的目標與企圖心，要在本土資助產能，我們正在等待所謂IPCAI 2的資金投入有更多進展。相較於我們一個月前路演時的說法，這方面沒有新的進展可報告。

And maybe I would add -- I'd add on that, just that in all of those regions, GF is uniquely positioned to be part of that partnership to create the capacity.

另外我也許可以補充——在所有這些地區，GF都具備獨特的定位，能成為合作夥伴的一部分，共同創造產能。

Do you have a follow-up, Vivek?

Vivek，你有追問嗎？

Yes, Sukhi. So for the follow-up, one more on gross margins. So the pricing environment seems very strong, right? It's clear to see your sales growth is outpacing your capacity, right, and wafer shipment growth. I'm curious, Dave, as you look at the next several quarters, how should we think about that interplay between how pricing and mix and utilization will help to drive your gross margins and help to kind of close the gap between where you guys are and versus where your peer group is?

有的，Sukhi。那我追問一題，關於毛利率。定價環境看起來非常強勁，對吧？很明顯你們的銷售成長正在超過你們的產能、以及晶圓出貨量的成長。我想請教Dave，當你看未來幾個季度時，我們應該如何理解定價、產品組合與產能利用率之間的互動，如何推動你們的毛利率，並幫助縮小你們與同業之間的差距？

Sure. Look, it does remain, I would say, a robust pricing environment. When we saw the demand that was in front of us, Vivek, as we kind of talked about in the road show, earlier this year and, in fact, even late 2020, we really wanted to focus on 3 things as a company to get us to our long-term model. We wanted to focus on certainty of demand. We wanted to focus on durability of the end markets and the customers that we serve, and we wanted to focus on profitability, in no particular order, those 3.

當然。你看，我會說定價環境確實仍然強勁。當我們看到眼前的需求時，Vivek，正如我們在路演中談到的，在今年稍早、甚至早在2020年末，我們作為一家公司非常希望聚焦三件事，讓我們走向長期模型。我們想聚焦在需求的確定性；我們想聚焦在終端市場與我們服務客戶的持久性；以及我們想聚焦在獲利能力——不分先後，這三項。

And so as we made progress on our LTAs throughout the course of the year, we started to see that 2021 for us, as I mentioned, was really a year of moderation of depreciation and fixed cost absorption. And we had a large percentage of our business that's single-sourced with those customers. And so as we signed and contemplated those LTAs, we really baked in a lot of the pricing improvements into 2022 and beyond.

因此，當我們在一年當中推進長期協議（LTA）時，我們開始看到，對我們而言，2021年——如我提到的——主要是折舊趨於溫和以及固定成本吸收改善的一年。而我們有很大比例的業務是由這些客戶單一來源供應（single-sourced）。所以當我們簽署並評估這些LTA時，我們其實已把許多定價改善「烘焙」進2022年及之後。

And so 2021, you're seeing the margin expand in a methodical and predictable way really based on moderation of depreciation and fixed cost absorption. And then when you look further out into time and you start to see those LTAs really start to pick up in the 2022 time frame, that's when you really start to see the pricing impacts.

因此在2021年，你看到毛利率以有系統且可預期的方式擴張，主要是基於折舊趨緩與固定成本吸收改善。然後當你把時間拉得更遠，開始看到這些LTA在2022年的時間框架真正開始發力時，你就會真正看到定價帶來的影響。

Tom, anything you'd add to that?

Tom，你要補充嗎？

Yes. I'd just say there was a real methodology to that as well with our customer engagement was to give our customers the chance to understand where these pricings will come from and give them a chance to respond for what they want to do in the marketplace.

是的。我只想說，這當中也有一套很明確的方法論；在我們與客戶互動時，我們會讓客戶有機會理解這些價格將從何而來，也讓他們有機會回應，決定他們在市場上想怎麼做。

Our next question comes from Joe Moore of Morgan Stanley.

下一題來自摩根士丹利（Morgan Stanley）的Joe Moore。

Congratulations on your first quarter. I wonder if you could talk about the smartphone business. You guys seem to be showing quite a bit of strength there. It's one area where, at least from the RF customers, there was a little bit of a mix. Q4 outlook is still a very good 2022 outlook. So maybe if you could just kind of talk to the prospects for growing that business over time.

恭喜你們第一季的表現。我想請你們談談智慧型手機業務。你們在那裡似乎展現出相當強的動能。這是其中一個領域——至少從RF客戶來看——產品組合有一些變化。第四季展望仍然非常好，2022年的展望也非常好。所以也許你們可以談談，這項業務隨時間推進的成長前景。

Yes. So I think for me, the smart mobile device is the poster child for feature-rich technology like we make, whether it's secure pay transaction, audio functions, the power management ICs, the chips that control the touchscreen or, as you're talking about, for connectivity, the front-end modules. You'll -- I think you'll see puts and takes amongst all the players, but we have a significant share because of our differentiated technology with all the players in that front-end market, the front-end module market with our ADAS [W] and RF SOI technologies.

是的。對我而言，智慧行動裝置是我們所做的「功能豐富」技術的最佳代表，不論是安全支付交易、音訊功能、電源管理IC、控制觸控螢幕的晶片，或如你所提到的連線用前端模組。你會——我想你會看到各家業者之間有得有失（puts and takes），但由於我們在前端市場、前端模組市場具備差異化技術，並且透過我們的ADAS [W]與RF SOI技術，我們在所有玩家中擁有相當可觀的市占。

So what we're seeing is maybe some puts and takes across the customers, but the overall volumes, we're not seeing any change. In fact, we're still trying to catch up to all the demand.

所以我們看到的是，客戶之間可能有一些此消彼長，但整體出貨量我們沒有看到任何變化。事實上，我們仍在努力追趕所有需求。

David, anything you'd add to that?

David，你要補充嗎？

No, I think it's well said, Tom. We're positive on the trajectory of that business.

沒有，我覺得Tom說得很好。我們對這項業務的走勢持正面看法。

Great. And just as a follow-up on that segment, is the visibility and the kind of supply constraint different in smartphone versus other markets? Or does it feel just as constrained to you as kind of the other end markets are?

很好。那針對這個部門再追問一下：在智慧型手機相較於其他市場，你們的能見度以及供應受限的情況是否不同？還是對你們而言，它感覺跟其他終端市場一樣受限？

A lot of the solutions we provide to our customers play in multiple markets. And so it's kind of squeezing the balloon one end, and it pops at the other end. So I think it's pretty broad-based, actually.

我們提供給客戶的許多解決方案會同時應用在多個市場。因此有點像擠壓氣球的一端，另一端就會鼓起來。我認為其實是相當廣泛、全面性的。

Our next question comes from Chris Danely of Citi.

下一題來自花旗（Citi）的Chris Danely。

And I'll add to the conga line of congratulations. Just another question on end markets to follow up Joe's question. So to be a little more specific on the sequentials, why was mobile flat sequentially? Was that pretty much in line with expectations? And then also, why was the comm business up so much sequentially? And then, any color you can give on guidance by end markets for Q4 would be great.

我也加入一長串的恭喜。再問一題終端市場的問題，延續Joe的提問。更具體談一下環比（sequential）：為什麼行動（mobile）環比持平？這大致符合預期嗎？另外，為什麼通訊（comm）業務環比成長那麼多？還有，如果你們能提供第四季按終端市場拆分的指引（guidance）色彩會很有幫助。

Dave, why don't you take that one, and I'll add some comments?

Dave，你先回答這題，我再補充一些看法？

Sure. Look, smart mobile is -- as you know, it represents about half of our total revenue. And we were expecting that business -- after good sequential growth from Q1 to Q2, we were actually expecting that business -- it was up modestly. We're expecting that business to be kind of relatively flat on a quarter-to-quarter basis on a sequential.

好的。你看，智慧行動（smart mobile）——如你所知——約占我們總營收的一半。我們原本預期這項業務——在第一季到第二季有不錯的環比成長之後——我們其實預期這項業務——它是小幅成長——我們預期它在季度之間的環比會相對持平。

And so the rest of the businesses, however, when we look at automotive up in a meaningful way, comms, as you mentioned, infrastructure data center up in a meaningful way and then, of course, home and industrial IoT also up in a meaningful way, if you think about the macro trends across those markets, those are all markets that we expected to grow in a pretty meaningful way for us sequentially, and they did.

不過其餘業務方面，當我們看到汽車業務有顯著成長；你提到的通訊、基礎建設與資料中心也有顯著成長；當然，家用與工業物聯網（IoT）同樣有顯著成長。若你從這些市場的總體趨勢來看，這些都是我們預期會在環比上對我們帶來相當明顯成長的市場，而結果也確實如此。

Tom, anything that you'd add?

Tom，你要補充嗎？

Yes, I'd add. Yes, I'd like -- there is some timing between when the cyclicality of the smart mobile market and when inventory start to build, and there's a little bit of that in play there as well.

是的，我補充一下。是的，我想——智慧行動市場的循環性與庫存開始建立的時間點之間存在一些時序差異，這裡也有一點這種因素在發揮作用。

Great. And for my follow-up, do you think that the supply chain is starting to get a handle on the shortages, especially in the automotive and the industrial space? Or do you think it's still getting a little bit worse out there?

很好。那我追問：你認為供應鏈是否開始掌握並改善短缺狀況，特別是在汽車與工業領域？還是你覺得外部情況仍在變得更糟？

That's a great question. I think, broad-brush, say better or worse that there's pockets where there's maybe better planning and there's better opportunity. I would say that over the course of this year, we've closed the gap, but it really started with a significant gap between the supply and the demand. And I think we've made as an industry as we have record shipments in all manufacturing, we've closed that gap, but nowhere near where it needs to be closed. I think that's the closest I could come to given a broad-based answer to that.

這是個很好的問題。我想若用較概括的方式說是變好或變糟，其實有些區塊可能規劃更好、機會也更好。我會說在今年的過程中，我們已經縮小了差距，但一開始供給與需求之間存在相當大的缺口。我認為作為一個產業，隨著我們在所有製造領域都有創紀錄的出貨，我們確實縮小了那個缺口，但離需要縮小到的程度還差得很遠。我想這是我能給出的最接近、較全面的回答。

David, would you add anything to that?

David，你會對此再補充什麼嗎？

No, I think that's well said.

不，我覺得說得很好。

Our next question comes from Chris Caso of Raymond James.

我們的下一個問題來自 Raymond James 的 Chris Caso。

The first question is a follow-up on some of your earlier comments on pricing. And what it sounds like from your comments and just to confirm my understanding that it sounds like there's a natural tailwind for pricing as we go into 2022 as the LTAs kick in that were signed at higher prices. But as a follow-on to that is just a question of, what the trajectory of pricing in the industry do you believe is longer term? I guess a question on how much of the improved pricing that we're seeing now is cyclical versus structural for your part of the industry.

第一個問題是針對你們先前一些關於定價的評論做追問。從你們的評論聽起來，並且只是想確認我理解無誤：隨著我們進入 2022 年，先前以較高價格簽署的長期協議（LTA）開始生效，定價似乎會有一個自然的順風。但接著延伸的問題是，你們認為產業定價的長期走勢會是如何？我想問的是，我們現在看到的定價改善，有多少是週期性的、又有多少是結構性的（就你們所處的產業環節而言）。

Yes. Let me -- I'll give the industry perspective. Dave, I always ask you for your thoughts on it. I think it really comes back to the fundamentals. This is an industry that's, call it, $0.5 trillion industry. It took 50 years to get to become $0.5 trillion industry. And you pick your analyst, it's going to double in the next 8 to 10 years, kind of a 5x acceleration it creates.

是的。我先——我先從產業角度來談。Dave，我一向也會請你分享你的看法。我認為這其實回到基本面。這是一個——姑且說——5,000 億美元規模的產業。它花了 50 年才發展到 5,000 億美元。你隨便挑一位分析師，他都會說未來 8 到 10 年會翻倍，等於是帶來約 5 倍的加速。

The economic model that created this $0.5 trillion industry is not the economic model that will allow us to double in the next 8 to 10 years. And as a result, there has to be better ways of funding this capacity. And one of which is higher ASPs to be able to afford that type of investment. So I don't think this is a moment in time. I think this is part of the growth trajectory of our industry to be able to afford the capacity and do it in a way where it makes economic sense for the offtakers of that capacity and foundries like GF that have to create that capacity.

促成這個 5,000 億美元產業的經濟模型，並不是能讓我們在未來 8 到 10 年翻倍的經濟模型。因此，必須有更好的方式來為這些產能提供資金。其中之一就是更高的平均售價（ASP），才能負擔那種投資規模。所以我不認為這只是某個時間點的現象。我認為這是我們產業成長軌跡的一部分，為了能負擔產能、並以對產能承購方（offtakers）以及像 GF 這樣必須建置產能的晶圓代工廠都具經濟合理性的方式來進行。

David, what would you add to that?

David，你會對此再補充什麼？

Yes. Just a couple of things really specific to GF. I agree and echo all of Tom's comments. But specific to GF, you've seen some modest -- I would say, modest price improvement from sequentially Q1 to Q2, Q2 to Q3. As Tom mentioned in his script, capacity sequentially Q3 to Q4 goes up about 4%. But if you look at our revenue guide at the midpoint, we're guiding up about 7%; at the high end, about 8%; at the low end, about 6%. And so that implies that we're mixing up and ASP upping our business sequentially from Q3 to Q4.

會的。就 GF 而言，我補充幾點更具體的。我同意並呼應 Tom 的所有評論。但就 GF 來看，你已經看到一些溫和——我會說是溫和的——價格改善，從 Q1 到 Q2、Q2 到 Q3 逐季改善。如 Tom 在講稿中提到的，產能從 Q3 到 Q4 逐季增加約 4%。但如果你看我們營收指引的中位數，我們指引約成長 7%；高端約 8%；低端約 6%。這意味著從 Q3 到 Q4，我們的產品組合在改善，且 ASP 也在逐季上升。

And so I think when I look out into time, I look at these LTAs, I look at the duration on the LTAs probably centered up around 4 years for us. I look at the pricing in those LTAs, which is really pretty flat. So kind of step function up and then pretty flat as those LTAs ramp kind of throughout 2022 and even parts of 2023. This looks like a long-term trend to me, and that's speaking specific to the data that I gave you.

因此，當我往後看，我會看這些 LTA，也會看 LTA 的期間——對我們而言大概集中在 4 年左右。我再看這些 LTA 的定價，其實相當平坦。也就是先有一個階梯式上升，然後隨著這些 LTA 在 2022 年、甚至 2023 年部分期間逐步放量，價格大致維持平穩。就我提供的這些數據來看，這對我而言像是一個長期趨勢。

Got it. Helpful. As a follow-up, if you can make some commentary on CapEx, where it is now and where it's been going. It looks like it's obviously up a lot year-on-year. From my numbers, it looks like it's actually down a bit sequentially. You talk about where CapEx needs to run over the next couple of quarters in order to hit your customer shipment goals and maybe speak generally about the availability of equipment, is the availability of equipment adequate to be able to support the ramp that you need to execute on?

了解，很有幫助。追問一下，如果可以請你們評論一下資本支出（CapEx），目前的水準以及其走勢。看起來很明顯年增很多。以我的數字來看，逐季似乎還略有下降。你們能否談談未來幾季 CapEx 需要維持在什麼水準，才能達成你們對客戶出貨的目標？也請大致談談設備供應的可得性——設備供應是否足以支援你們需要執行的擴產爬坡？

Sure. So if you recall from the road show, we were talking about our CapEx over '21 and '22 roughly being about $2 billion in 2021, roughly $4.5 billion in 2022, again, all supported by those LTAs that we've talked a bit about on this phone call. For 2021, we'll probably come in closer to about $1.9 billion, plus or minus $100 million or so, depending on what comes in between now and the end of the month from an equipment perspective. And then we're still looking good to roughly that $4.5 billion number for 2022.

當然。如果你還記得我們在路演時提到，2021 年與 2022 年的 CapEx 大致是：2021 年約 20 億美元、2022 年約 45 億美元——同樣都是由我們在這通電話中談到的那些 LTA 所支撐。就 2021 年而言，我們可能會更接近約 19 億美元，上下浮動約 1 億美元左右，取決於從現在到月底之間，設備交付方面還會進來多少。至於 2022 年，我們仍然看起來能達到約 45 億美元的水準。

So from a CapEx perspective, I would say we're largely on plan to what we spoke about in the road show. And so no real meaningful changes there with respect to CapEx.

所以從 CapEx 的角度來看，我會說我們大致符合路演時所談的計畫。因此在 CapEx 方面沒有任何真正重大的變化。

Tom, anything to add?

Tom，有什麼要補充的嗎？

No, I'd add 2 things. One, our long-term model is after we get through this essentially acceleration of capacity to meet our customers' needs in their partnership, we'll get to that 20% of revenue in time. And then there was a little bit of a question on do we have the access to the equipment. When we started the -- our expansion in our plans, it was late Q4 of 2020, early Q1 of 2021. A little bit ahead of the kind of the acceleration of all the equipment needs. So the combination of we were early into block our slots with our equipment suppliers in this rapid expansion the whole industry is doing as well as prudent planning of when those tools would come online, it gives us the confidence that when we look forward in how we're managing our business, we won't be gated by equipment deliveries.

沒有，我補充兩點。第一，我們的長期模型是：在我們完成這一波為了滿足客戶需求、並與客戶合作而進行的產能加速擴張之後，隨著時間推進，我們會回到 CapEx 約占營收 20% 的水準。第二，剛才也有一點是在問我們是否能取得足夠的設備。當我們開始——規劃擴產時，大約是在 2020 年 Q4 末、2021 年 Q1 初，時間點比整個產業設備需求全面加速稍微早一些。因此，結合我們較早就向設備供應商鎖定產能時段（block our slots），以及對這些機台何時上線的審慎規劃，讓我們有信心：展望未來，在我們管理業務的方式下，不會因設備交付而受限。

Our next question comes from Tristan Gerra of Baird.

我們的下一個問題來自 Baird 的 Tristan Gerra。

Congratulations on going public. A question about your end market mix longer term. I think you've mentioned in the past that smartphones and IoT are pretty much in line with corporate average gross margin, and the rest is higher margin. You've talked about the contribution of depreciation in ASPs and gross margin. Longer term, when do mix shift within your end markets and notably the higher-margin infrastructure, automotive start really contributing to your gross margin? And also, where could we see automotive as a percent of revenue in a few years from now from the 6% that you're exposed to today?

恭喜你們上市。我有一個關於你們長期終端市場組合的問題。我想你們過去提到，智慧型手機與 IoT 的毛利率大致與公司平均一致，其餘則是較高毛利。你們也談過折舊在 ASP 與毛利率中的貢獻。長期來看，你們終端市場內部的組合轉移，特別是較高毛利的基礎建設、車用，何時會開始對毛利率做出更明顯的貢獻？另外，從目前你們曝險約 6% 的營收占比來看，幾年後車用可能會到營收的多少百分比？

David?

David？

Yes. So what we've spoken about is that our end market mix longer term, our personal computing segment will become a smaller part of our business as a percentage of total revenue going forward. I think you're starting to see some of that on a sequential basis, and you're certainly seeing some of that on a year-over-year basis. So that's a trend that we expect to continue over time as we start to reach our longer-term model, let's call it, the 2025-ish time frame.

好的。我們談過的是，從長期來看，我們的終端市場組合中，個人運算（personal computing）部門占總營收的比重會逐步下降。我認為你已經開始在逐季的變化中看到一些跡象，也確實在年對年的比較中看到這點。因此，隨著我們逐步達到長期模型——大概可說是 2025 年左右的時間框架——我們預期這個趨勢會隨時間持續。

Automotive, for us, you've seen grow in a very meaningful way this year off a very small base, granted, but we have some very nice design wins. You've seen some of those publicly announced in the marketplace. You've seen some of the partnerships that have also been announced. And so that's a segment for us. That could get to a level of, call it, 10% to 15% of business longer term. That's something that could grow for us in a really meaningful way.

車用方面，我們今年在一個很小的基期上——這點沒錯——已經出現非常顯著的成長，而且我們拿下了一些非常不錯的設計導入（design wins）。你也看到其中一些已在市場上公開宣布。你也看到一些合作夥伴關係同樣已對外宣布。因此，這對我們而言是一個有機會的領域。長期來看，這個部門可能達到——姑且說——占業務 10% 到 15% 的水準。這對我們來說可能會有非常有意義的成長。

Comms, infrastructure and data center, home and industrial IoT and smart mobile devices, those are all areas where we have real franchises. So whether we're talking about our franchise in connectivity and SOI, whether you're talking about some of the things that we're able to do uniquely in power management, whether you're talking about longer-term silicon photonics or silicon germanium. Those are some areas where we have some real differentiation of the company, and those market segments will start to grow as those design wins are employed.

通訊、基礎建設與資料中心、家用與工業 IoT，以及智慧行動裝置，這些都是我們擁有真正既有優勢（franchises）的領域。所以不論我們談的是在連結性與 SOI 的既有優勢，或是我們在電源管理上能夠獨特做到的一些事情，或是更長期的矽光子（silicon photonics）或矽鍺（silicon germanium）。這些都是我們公司具備實質差異化的一些領域，而隨著這些設計導入開始被採用，這些市場區隔將會開始成長。

Tom, anything you'd add?

Tom，你有什麼要補充的嗎？

Yes. I'd add a little bit about what's really shifting in the industry. Maybe 10 years ago, the #1 criteria for a designer was digital performance, then power, then maybe connectivity. It's kind of flipped around now. Power is the #1 element to solve for, then connectivity and then there's plenty of digital compute. And I think that plays to our strength across these end markets.

有的。我補充一點關於產業真正正在轉變的地方。大概 10 年前，設計者的第一優先考量是數位效能，其次是功耗，再來可能是連結性。現在幾乎反過來了：功耗是第一個要解決的要素，其次是連結性，而數位運算能力則已相當充足。我認為這正好符合我們在這些終端市場的優勢。

Now on PC, that's a little bit different. That's all about digital compute, and that's why that market kind of moves away from us, which is fine, because the markets we play in, in IoT and industrial business and automotive and the mobile handset, it's all about power management. And so that's really where I think our differentiation, our leadership as a company comes into play. Not to mention next generation of connectivity in the data center and our leadership, I call bleeding edge in silicon photonics, is where we're positioned.

而在 PC 端，情況就有點不同。那完全是數位運算，因此那個市場某種程度上就逐漸遠離我們，這也沒關係，因為我們所參與的市場——在 IoT、工業業務、車用以及行動手機——核心都在於電源管理。因此我認為，這正是我們的差異化與公司領導地位真正發揮作用的地方。更不用說資料中心的下一代連接，以及我們在矽光子領域的領先地位——我稱之為最前沿（bleeding edge）——這就是我們所處的位置。

Do you have a follow-up, Tristan?

Tristan，你有追問嗎？

Yes. For my follow-up, wanted to get a sense perhaps of your growth for the various nodes, whether we're looking at 12-nanometer and higher nodes or any areas that you see growing faster. And how does that compare with the industry average? Just trying to get a sense of your mix by node in terms of growth prospect.

有。我的追問是，想了解你們在不同製程節點的成長情況，例如 12 奈米及以上的節點，或是你們看到哪些領域成長更快。以及這與產業平均相比如何？我只是想掌握你們按節點來看，在成長前景上的組合情況。

We don't think about our business that way, and we don't talk about the business that way. When we pivoted the company, it was about platforms that are enabled with markets and applications. There's a single device type in mind. For us, it's what are the features we're creating and what platform that make the most sense. And that's how we think about investments, that's what we think about driving our business, and that's what drives our top line growth in these markets. So we really don't have a view of that granularity, looking for in a node basis.

我們不會用那種方式來思考我們的業務，也不會用那種方式來談論業務。當我們推動公司轉型時，重點在於以市場與應用所驅動的平台。腦中想的是單一的裝置類型。對我們而言，重點是我們正在打造哪些功能，以及哪個平台最合理。這就是我們思考投資的方式、推動業務的方式，也正是驅動我們在這些市場營收成長的因素。因此我們確實沒有那麼細的觀點，不會以節點為基礎去看那種粒度。

David, anything you'd add to that?

David，你有什麼要補充的嗎？

No, that's absolutely right, Tom. And when I look at our sequential performance, what I see is -- I see feature-rich CMOS sequentially, let's call it, flat. It's actually slightly down, but I see good growth in FDX, good growth in RF SOI, good growth in SiGe. Even, I would say, very small numbers, but even a little bit of revenue in silicon photonics. And so I look at those areas, all areas where we have some real meaningful differentiation in the market, and I'm quite encouraged.

沒有，Tom，這完全正確。而當我看我們的連續（季度）表現時，我看到的是——我看到功能豐富的 CMOS 連續來看，我們就稱之為持平。其實是略微下滑，但我看到 FDX 有不錯的成長、RF SOI 有不錯的成長、SiGe 也有不錯的成長。甚至我會說，雖然數字很小，但矽光子也有一點點營收。因此我看的是這些領域——都是我們在市場上具備真正有意義差異化的領域——我感到相當鼓舞。

Our next question comes from Krish Sankar of Cowen.

下一個問題來自 Cowen 的 Krish Sankar。

(inaudible) calling on behalf of Krish. First off, I'd like to congratulate you guys as well on the strong results. Yes. The first question I had was related to your resin fab. And I guess, for the, I guess, the headroom there for driving much of the revenue growth in the coming quarters. So just curious, like can you provide any color on what the -- roughly what the utilization rates might be currently at Dresden? And also, for the new 22FDX process that was qualified for automotive platforms, is that just a new process that's being offered at that fab? Or is it like a new line that's been built out and, hence, might be driving a lot of the near-term revenues?

（聽不清）我代表 Krish 提問。首先，也想恭喜各位取得強勁的成果。是的。我第一個問題與你們的 Dresden 晶圓廠有關。我想，未來幾季推動營收成長的空間（headroom）可能很大。所以想請教，你們能否提供一些資訊——目前 Dresden 的產能利用率大概是多少？另外，針對已通過車用平台認證的全新 22FDX 製程，那只是該廠提供的一個新製程嗎？還是說你們新建了一條產線，因此可能會在短期內帶來大量營收？

So David, why don't you talk about the Dresden build-out, and I'll talk about the 22FDX?

David，你先談談 Dresden 的擴建，我再來談 22FDX？

Sure. Look, you're right on highlighting Dresden, it's an important fab for us. They're all important fabs for us. But from a build-out perspective, it's a fab that has the ability and the capacity once fully tooled to be able to deliver about 850,000 wafers a year, and that's up from about 300,000 tooled wafers per year in 2020.

好的。你點出 Dresden 很重要是對的，這對我們而言是一座重要的晶圓廠。其實每一座廠對我們都很重要。但就擴建角度而言，這座廠在設備完全到位後，具備能力與產能可在一年內交付約 85 萬片晶圓，這相較於 2020 年每年約 30 萬片已配置設備的晶圓產能有所提升。

And so we are tooling Dresden, as you know, from the road show conversations. We're not quite doubling capacity on a year-over-year basis there but getting close. And so that's a factory that is continuing to bring tools in, get them installed, ramp them to production and ship product for our customers that are clamoring for it. And so we're quite pleased with how Dresden is performing.

因此我們正在為 Dresden 配置設備，如你從路演對談中所知。我們在那裡的產能並未做到年對年幾乎翻倍，但已經很接近了。所以這座工廠持續導入設備、完成安裝、爬坡到量產，並為那些迫切需要產品的客戶出貨。因此我們對 Dresden 的表現相當滿意。

From a utilization perspective, I would say, at the enterprise level, we are above 100% utilization as an enterprise, as a GlobalFoundries entity. And I can say that's true for almost every factory, over 100% utilization right now.

就產能利用率而言，我會說，以公司整體層級來看，我們的利用率超過 100%——以 GlobalFoundries 這個實體而言。我也可以說，幾乎每一座工廠目前都超過 100% 的利用率。

Tom, anything that you'd add?

Tom，你有什麼要補充的嗎？

Yes, I'd add 2 things. One I'd add is, look, it makes sense as you're trying to add capacity as fast as possible. You add it where you already have the brick-and-mortar and a building that's facilitated but needs the equipment. And that's why we're able to ramp very quickly there, as you noted.

有，我補充兩點。第一點是，你想要盡可能快地增加產能時，把產能加在你已經有廠房與建築本體、而且基礎設施已就緒但仍需要設備的地方，是合理的。這也是為什麼如你所提到的，我們能在那裡非常快速地爬坡。

With regards to 22FDX, that is one of the key technologies that we've built in Dresden for our Fab 1 facility. So that's -- the real story about 22FDX is that was a homegrown technology, which is a great example of how we think about technology. We don't say, well, here's a node, what do we do with it? We started with application in mind. We said we wanted the lowest power technology possible. So that's why it's built on SOI technology. We first enabled it with back [pipes], so we can make the power even lower. Then we embedded RF and embedded memory to make it the platform of choice for any SoC that wanted to be connected, wanted the ultimate leadership technology in power per device -- minimalization of power per device.

至於 22FDX，那是我們在 Dresden 的 Fab 1 廠區所建置的關鍵技術之一。所以——22FDX 的真正故事在於，那是一項自研技術，這是我們如何思考技術的絕佳例子。我們不會說，嗯，這裡有個節點，我們要拿它做什麼？我們是從應用出發。我們說我們想要可能最低功耗的技術。所以它才建立在 SOI 技術之上。我們先用背面（back [pipes]）把它啟用，讓功耗更低。接著我們嵌入 RF 與嵌入式記憶體，使其成為任何想要連接、想要在每個裝置功耗上具備極致領先技術——把每個裝置的功耗降到最低——的 SoC 的首選平台。

That was a technology that was developed in Dresden, and we're building out in Dresden and that we're winning a lot of our connectivity business. And part of that expansion David talked about in Dresden is to enable that platform with capacity.

這項技術是在 Dresden 開發的，我們也正在 Dresden 擴建，並且在那裡贏得了大量連接相關業務。David 剛才談到的 Dresden 擴產，其中一部分就是為了讓這個平台具備足夠的產能。

Perfect, Tom. Just a real quick follow-up, Dave, on -- in terms of government incentives in Q3 and maybe any expectations for Q4. Like how much in terms of incentives and grants might have been embedded in COGS and OpEx?

很好，Tom。我再很快追問一下，Dave，關於——Q3 的政府補助，以及對 Q4 是否有任何預期。大概有多少補助與補貼被計入銷貨成本（COGS）與營業費用（OpEx）？

Very, very little in Q3. And so when we look out, the single largest area for us with government partnership right now is actually the Fab 7H in Singapore, which Tom referenced earlier. That is an area where, right now, we have an ongoing, very strong partnership with the Singaporean government to bring that economic development there to Singapore. But in terms of benefit from that relationship, very, very little. Nothing that's really material in Q3, but we do expect it to be more meaningful in the future.

Q3 非常、非常少。因此當我們往前看，目前我們與政府合作夥伴關係中最大的一塊，其實是在新加坡的 7H 廠，也就是 Tom 先前提到的。那是我們目前與新加坡政府維持持續且非常強健合作的領域，目的是把那裡的經濟發展帶到新加坡。但就這段關係帶來的效益而言，目前非常、非常少。Q3 沒有任何真正具重大性的項目，但我們確實預期未來會更有意義。

Our next question comes from Raji Gill of Needham & Company.

下一個問題來自 Needham & Company 的 Raji Gill。

Congratulations on a strong quarter out of the gate. So that's great to see. A couple of questions, if I may. The first question on the $500 million increase in customer prepayments access fees since the road show, it's about a 20% increase, can you maybe elaborate further on what you're seeing there in terms of customers paying you upfront to get access to capacity? What's been -- what has changed over the last 6 weeks? Any details there?

恭喜你們一開始就交出強勁的一季，看到這點很棒。我有幾個問題，如果可以的話。第一個問題是關於自路演以來，客戶預付款／產能保留費（access fees）增加了 5 億美元，約增加 20%。你們能否更進一步說明你們看到的情況：客戶為了取得產能而提前付款，背後反映了什麼？過去 6 週發生了什麼變化？有任何細節嗎？

And just for my follow-up, the capacity increases that you saw this quarter and you're talking about another 4% in going into Q4, coupled with a 3% increase in ASPs, how do we think about the split between capacity expansion versus ASP growth when we're looking at calendar '22? Any kind of flavor there, directionally, data would be helpful.

另外我的追問是，本季你們看到的產能提升，以及你們談到 Q4 還會再增加 4%，再加上 ASP 提升 3%，當我們看 2022 年曆年時，該如何思考產能擴張與 ASP 成長之間的拆分？任何方向性的說明或數據都會有幫助。

Sure. So let me take those, and then Tom, I'll throw it over to you and see if you have anything to add. So look, what's changed in the last 6 weeks, really, we just continue to execute. We've been engaged with customers, well, given the size and the scope of our single-source business for a long time, as you know. And as we've continued to work those LTAs, that queue of LTAs, we have customers that are willing to put their balance sheet to work and provide prepayments and access fees for us to bring online and commit capacity to them, both today as well as in the future.

好的。我先回答這些，然後 Tom，我再把問題丟給你，看看你是否要補充。你看，過去 6 週真正改變的是，我們只是持續執行。我們一直與客戶保持互動——如你所知，考量到我們單一來源（single-source）業務的規模與範圍，我們已經這樣做很長一段時間了。隨著我們持續推進那些 LTA（長期協議），也就是那一批 LTA 的排隊進度，我們看到有客戶願意動用他們的資產負債表，提供預付款與產能保留費，讓我們能把產能上線並承諾給他們，不僅是現在，也包括未來。

Remember, we are capacity constrained right now. And so new agreements that we're signing, they're really signing up for new capacity that will come online in the future. So that's the future capacity that we're adding specific for those customers.

請記住，我們目前受到產能限制。因此，我們正在簽署的新協議，實際上是為未來將上線的新產能進行預訂。所以，那是我們為這些客戶新增的、專門對應的未來產能。

So in terms of what changed, I'd just say we're continuing to execute. We're working down the queue of engagements with customers, working with them in close partnership to bring them the product that they need.

至於有哪些變化，我只能說我們持續在執行。我們正依序推進與客戶的合作案，並與他們緊密合作夥伴關係，為他們帶來所需要的產品。

Yes. And I would just add, David, that the time frame from the time you begin to carve out the nonbinding MOU to creating a definitive agreement is roughly 2 to 3 months, almost a quarter. And so what you're seeing that took place since the last quarter is the signing of LTAs that are already well along in closing the final Ts and Cs.

是的。我也想補充一下，David，從開始擬定不具約束力的 MOU（合作備忘錄）到形成具約束力的最終協議，時間大約是 2 到 3 個月，接近一個季度。因此，你看到自上個季度以來發生的事情，是那些在敲定最終條款與條件（Ts & Cs）方面已經進入後期的 LTA（長期協議）完成簽署。

And then -- and look, in terms of your question about 2022, I'll defer the question about the split between volume, mix and ASP for 2022 until maybe a conversation in February. But I think what you're seeing on a sequential basis where I mentioned earlier that capacity is growing about 4% sequentially. But yet, we're guiding at the midpoint up about 7% sequentially. I think if you're looking to model something, you can maybe take some of those recent proof points and expand those out into the future if you're looking for guidance.

另外——就你關於 2022 年的問題而言，我會把 2022 年在出貨量、產品組合與 ASP（平均售價）之間的拆分問題，延到可能 2 月的一次對話再談。不過我認為，你在環比上看到的是：如我先前提到，產能環比大約成長 4%；但我們在指引的中位數卻是環比成長約 7%。如果你想建立模型、尋求一些指引，你或許可以把近期的一些佐證點納入，並將其外推到未來。

Our next question comes from Mehdi Hosseini of SIG.

下一個問題來自 SIG 的 Mehdi Hosseini。

A couple of follow-ups. Did I hear you correct that all your facilities are currently running at full utilization rate?

有幾個追問。我是否聽對了：你們所有的廠房目前都以滿載的稼動率在運行？

So let me be very clear on that. Our facilities for what its tools for today are running white-hot. They're making as many wafers as possible. Our customers are counting on us to do that. The capacity, every day changes as we add more tools, right, to create more capacity. And as soon as that tool -- those tools are qualified and able to make more wafers, we will. So utilization, for all intents and purposes, is 100-plus percent, and we see that remaining for quite some time.

我把這點講清楚。我們現有工具所對應的產能，廠區目前都在滿負荷運轉，火力全開，盡可能多地生產晶圓。我們的客戶也指望我們這麼做。產能每天都在變化，因為我們會新增更多設備（tools），對吧，以創造更多產能。而一旦那些設備完成認證、具備生產更多晶圓的能力，我們就會立刻投入生產。所以就所有實際目的而言，稼動率是 100% 以上，而且我們預期這種狀況會維持相當長一段時間。

Okay. And then in terms of your technology platforms comparing like SOI to CMOS and FinFET, is there a big margin difference? Should we think about the mix also has an impact to your gross and operating margin profile?

好的。那麼就你們的技術平台而言，例如把 SOI 與 CMOS、FinFET 相比，毛利差異大嗎？我們是否應該認為產品組合也會影響你們的毛利率與營業利益率輪廓？

David, why don't you start, and I'll add a little color?

David，你先回答，我再補充一些說明，好嗎？

Yes. Look, I think you could look across any of our technology platforms, and you could find products that have higher margins and lower margins. So whether you're looking at SOI with RF SOI and FDX, whether you're looking at feature-rich CMOS, FinFET, SiGe, silicon photonics, I think the real question is how differentiated is that the product that, that customer is engaging with GF, how differentiated is that product, and the margin structure tends to follow.

是的。你看，我認為在我們任何一個技術平台上，你都能找到高毛利與低毛利的產品。所以不論你看的是 SOI（包含 RF SOI 與 FDX），或是功能豐富的 CMOS、FinFET、SiGe、矽光子，我覺得真正的問題在於：客戶與 GF 合作的那個產品有多差異化？產品差異化程度越高，毛利結構通常也會隨之反映。

Tom, would you add anything?

Tom，你要補充嗎？

Yes. And again, a little bit, think of it this way. Every one of these platforms have a number of different features. If a solution is only using one feature, there's a different ASP associated with that than if they kind of leverage the RF connectivity, they're going to put embedded memory. It adds complexity, but they have value that our customers need that they get rewarded for and we get rewarded for.

是的。再補充一點，可以這樣想：每一個平台都有多種不同特性。如果一個解決方案只用到其中一個特性，對應的 ASP 會不同；但如果他們進一步利用 RF 連接能力、加入嵌入式記憶體，就會增加複雜度，但也帶來我們客戶所需要的價值——他們因此獲得回報，我們也獲得回報。

So I think of it more feature-rich, the solution we sell independent of the platform at time. The more differentiated it is, the more we create value for our customers for them to capture, for us to capture.

所以我更傾向把它視為：我們所銷售的解決方案本身是否功能更豐富，而不單取決於當下的平台。差異化越高，我們為客戶創造的價值越大，客戶能捕捉到的價值越多，我們也能捕捉到更多。

All right. I just want to better understand because I'm under assumption that like with SOI, the wafer or raw material cost is much higher than FinFET, but you're able to get the premium on wafer ASP that would make it competitive with other platforms. That's what I was trying to better understand.

好的。我只是想更了解，因為我原本的假設是：像 SOI 的晶圓或原材料成本比 FinFET 高很多，但你們能在晶圓 ASP 上取得溢價，使其與其他平台具競爭力。我想釐清的是這一點。

Yes. Look, no one is going to pay for an extra cost substrate if it doesn't come with the value that's required to do the application. So it gets factored into the differentiation and the value we create using that specialty substrate.

是的。你看，如果額外成本的基板沒有帶來完成該應用所需的價值，沒有人會願意為它付費。所以這會被納入差異化與我們利用該特殊基板所創造的價值之中。

Got it.

了解。

Do you have a follow-up, Mehdi?

Mehdi，你還有追問嗎？

Yes. And then just for purpose of modeling, as I'm under the assumption that with these long-term agreements, there's a large proportion of those agreements that are coming in earlier during that 2- to 4-year period. Should I also assume that your OpEx growth should be moderate? So therefore, your operating margin expansion is pretty much driven by top line growth and the margin improvement that comes with better fixed cost absorption?

有。再就建模的目的來說，我的假設是：在這些長期協議中，有相當大比例的訂單量會在那個 2 到 4 年期間的較早階段就進來。我是否也應該假設你們的 OpEx（營運費用）成長會較溫和？因此，你們的營業利益率擴張主要由營收成長，以及更好的固定成本吸收所帶來的毛利改善所驅動？

That's right. If you recall, our long-term financial model, we've got revenue growth about 8% to 12%, gross margin at 40%, operating income margin, about 25%, which that implies then that OpEx is about 15%. And I think if you look at OpEx today, you'll see it's kind of rough and top around that level. And so that's something that we expect that to be relatively flat and to scale maybe modestly with revenue but certainly less than the rest of the group.

沒錯。如果你還記得我們的長期財務模型：營收成長約 8% 到 12%，毛利率 40%，營業利益率約 25%，這意味著 OpEx 約 15%。我想如果你看今天的 OpEx，你會發現大致就在那個水準附近並在那裡見頂。因此，我們預期它會相對持平，可能隨營收小幅擴張，但肯定會低於其他項目的增幅。

At this time, I'd like to turn the call back over to Sukhi Nagesh for closing remarks.

此時，我想把電話交回給 Sukhi Nagesh 作結語。

All right. Great. Thank you, Latif. Well, thank you, everyone, for joining us on our Q3 2021 earnings call. We look forward to meeting with many of you over the course of the quarter. And if you have any follow-ups, please feel free to reach out to us. Thank you.

好的。非常好。謝謝你，Latif。也謝謝各位參加我們 2021 年第三季（Q3 2021）財報電話會議。我們期待在本季度期間與各位中的許多人會面。如果你們有任何後續問題，歡迎隨時與我們聯繫。謝謝。

And this concludes today's conference call. Thank you for participating. You may now disconnect.

今天的電話會議到此結束。感謝各位參與。您現在可以掛線。