希捷科技 (STX) 2023 Q4 法說會逐字稿

Good day, and welcome to the Seagate Technology Fourth Quarter and Fiscal Year 2023 Conference Call. (Operator Instructions) Please note this event is being recorded. I would now like to turn the conference over to Shanye Hudson, Senior Vice President, Investor Relations. Please go ahead.

大家好，歡迎參加希捷科技2023財年第四季及全年電話會議。 （操作員指示）請注意，本次會議正在錄製中。現在，我想將會議交給投資者關係高級副總裁Shanye Hudson。請開始。

Thank you. Good afternoon, everyone, and welcome to today's call. Joining me are Dave Mosley, Seagate's Chief Executive Officer; and Gianluca Romano, our Chief Financial Officer. We posted our earnings press release and detailed supplemental information for our June quarter and fiscal year 2023 results on the Investors section of our website.

謝謝。大家下午好，歡迎參加今天的電話會議。希捷執行長戴夫·莫斯利和財務長詹盧卡·羅馬諾也與我一同出席。我們在網站的「投資者」板塊發布了收益新聞稿以及2023財年第二季和第三季業績的詳細補充資訊。

During today's call, we'll refer to GAAP and non-GAAP measures. Non-GAAP figures are reconciled to GAAP figures in the earnings press release posted on our website and included in our Form 8-K that was filed with the SEC. We've not reconciled certain non-GAAP outlook measures because material items that may impact these measures are out of our control and/or cannot be reasonably predicted. Therefore, a reconciliation to the corresponding GAAP measures is not available without unreasonable efforts.

在今天的電話會議上，我們將討論 GAAP 和非 GAAP 指標。非 GAAP 數據已與我們網站上發布的收益新聞稿中的 GAAP 數據進行了調節，並包含在我們提交給美國證券交易委員會 (SEC) 的 8-K 表格中。我們尚未調整某些非 GAAP 展望指標，因為可能影響這些指標的重大事項超出了我們的控制範圍和/或無法合理預測。因此，如果不付出不合理的努力，我們將無法獲得與相應 GAAP 指標的調整結果。

Before we begin, I'd like to remind you that today's call contains forward-looking statements that reflect management's current views and assumptions based on information available to us as of today and should not be relied upon as of any subsequent date. Actual results may differ materially from those contained in or implied by these forward-looking statements as they're subject to risks and uncertainties associated with our business.

在開始之前，我想提醒您，今天的電話會議包含前瞻性陳述，這些陳述反映了管理層基於截至今日所掌握資訊的當前觀點和假設，不應在任何後續日期依賴這些陳述。實際結果可能與這些前瞻性陳述中包含或暗示的結果有重大差異，因為它們受與我們業務相關的風險和不確定性的影響。

To learn more about the risks, uncertainties and other factors that may affect our future business results, please refer to the press release issued today and our SEC filings, including our most recent annual report on Form 10-K and quarterly report on Form 10-Q as well as the supplemental information, all of which may be found on the Investors section of our website.

欲了解可能影響我們未來業務成果的風險、不確定性和其他因素的更多信息，請參閱今天發布的新聞稿和我們向美國證券交易委員會提交的文件，包括我們最新的 10-K 表年度報告和 10-Q 表季度報告以及補充信息，所有這些文件均可在我們網站的“投資者”部分找到。

As always, following our prepared remarks, we'll open the call for questions. I'll now hand the call over to Dave for opening remarks.

像往常一樣，在準備好發言稿之後，我們將開始提問環節。現在，我將把演講時間交給戴夫，請他致開幕詞。

Thank you, Shanye, and hello, everyone. We appreciate you being here with us on today's call. Amid a tough business environment, Seagate delivered fourth quarter revenue of $1.6 billion, while holding non-GAAP operating margin at 3.5% to narrow our non-GAAP loss per share to $0.18. These results demonstrate financial leverage and our focus on returning to profitability.

謝謝Shanye，大家好！感謝您參加今天的電話會議。在艱困的商業環境下，希捷第四季營收達16億美元，非公認會計準則下的營業利潤率維持在3.5%，將非公認會計準則下的每股虧損收窄至0.18美元。這些業績反映了我們強大的財務槓桿，也反映了我們致力於恢復獲利能力的決心。

Importantly, we continue to generate positive free cash flow, achieving $168 million for the quarter and $626 million for the fiscal year, reinforcing Seagate's solid operational and financial execution amid the business backdrop we have seen in fiscal 2023. This past fiscal year has been shaped by macro and end market conditions that have tested our resilience in addition to our financial performance. I'm proud of our team's perseverance and execution.

重要的是，我們繼續產生正向自由現金流，本季實現1.68億美元，本財年實現6.26億美元，這鞏固了希捷在2023財年業務環境下穩健的營運和財務執行力。過去一個財年，宏觀市場和終端市場環境的變化不僅考驗了我們的財務表現，也考驗了我們的韌性。我為我們團隊的毅力和執行力感到自豪。

As a result of our proactive actions, we've lowered our cost structure by more than $350 million on an annualized basis. We've enhanced balance sheet flexibility taking out nearly $800 million of debt, funded largely by monetizing nonmanufacturing facility assets. We reduced production output by approximately 25% compared with peak volume in order to drive better supply/demand dynamics and enhance profitability as the markets recover. And all of these accomplishments were made while delivering on our 30-plus terabyte HAMR product development and qualification milestones with volume ramp on track to begin in early calendar 2024.

透過採取積極主動的措施，我們將成本結構按年率降低了超過3.5億美元。我們增強了資產負債表的靈活性，並消除了近8億美元的債務，這些債務主要透過非製造設施資產的貨幣化來籌集資金。為了在市場復甦之際推動更好的供需平衡並提高獲利能力，我們將產量與高峰相比減少了約25%。所有這些成就都是在我們實現30多TB HAMR產品開發和認證里程碑的同時取得的，產量預計在2024年初開始成長。

Looking ahead, we expect the macro and end market conditions that we have flagged throughout fiscal 2023 to continue weighing on demand through at least December. Consistent with our track record of managing what is within our control, we are taking additional measures to weather the near-term business environment. First, we are adjusting pricing, which we believe will help to ensure a healthy industry supply chain for our customers over the long term. And second, we are carefully managing our manufacturing capacity through a build-to-order approach for certain areas of the business to align our future production output with customer qualifications and demand plans.

展望未來，我們預計，2023財年期間我們一直關注的宏觀和終端市場狀況將至少持續到12月，進而對需求造成壓力。秉承我們管理可控制範圍內的過往記錄，我們正在採取額外措施來應對短期業務環境。首先，我們正在調整定價，我們相信這將有助於確保長期為客戶打造健康的產業供應鏈。其次，我們正在透過訂單生產的方式，對某些業務領域進行謹慎的產能管理，以確保我們未來的產量與客戶資格和需求計劃保持一致。

It will take time to bring production capacity and associated support resources back online. Therefore, it is crucial that we balance lead times for mass capacity products with our ability to re-ramp production. We expect these efforts will enable Seagate to efficiently manage supply for our customers as demand improves.

恢復產能和相關支持資源需要時間。因此，我們必須平衡大容量產品的交付週期和恢復產能的能力。我們預計，隨著需求的回升，這些努力將使希捷能夠有效地為客戶管理供應。

I'll now share some perspectives on near-term demand factors, in particular in the mass capacity markets, starting with our business in China. As expected, sales in China improved sequentially off of the March quarter lows, driven by increased demand in the VIA markets and certain regional cloud and enterprise OEM customers. While these trends are pointing in the right direction, sales are still well below historical levels.

現在，我將分享一些關於近期需求因素的觀點，尤其是在大容量市場，首先從我們在中國的業務開始。如預期，受威盛市場以及某些區域雲端運算和企業OEM客戶需求成長的推動，中國市場的銷售額較3月季度的低點環比有所回升。雖然這些趨勢正在朝著正確的方向發展，但銷售額仍然遠低於歷史水準。

Since the strict COVID protocols were eased last December, the pace of economic recovery in China has been uneven. For now, we are forecasting sales into China to remain relatively stable for the balance of the calendar year. We are encouraged to see Chinese authorities begin to inject more stimulus to reaccelerate economic growth. Based on customer input, economic improvement is expected to catalyze e-commerce, drive cloud-related ad revenue and spur new smart city projects. Seagate's VIA and nearline products are positioned to benefit from these anticipated improvements in the end market demand.

自去年12月嚴格的新冠疫情防治措施放鬆以來，中國經濟復甦步伐不一。目前，我們預測今年全年中國市場的銷售額將保持相對穩定。我們很高興看到中國政府開始推出更多刺激措施，以加速經濟成長。根據客戶回饋，經濟復甦預計將促進電子商務發展，推動雲端相關廣告收入成長，並刺激新的智慧城市專案。希捷的VIA和近線產品有望從這些預期的終端市場需求成長中受益。

Outside of China, nearline demand from enterprise OEM customers has remained soft. CIOs continue to operate under tighter budgets in response to near-term macroeconomic uncertainties. Ongoing efforts to optimize existing workloads, both on-prem and in the cloud are helping enterprise customers defer mass storage deployments. These trends have slowed the pace of inventory absorption among most of our U.S. cloud customers. We have significantly reduced shipments to several large cloud customers in order to accelerate inventory absorption and protect our financial returns.

在中國以外，來自企業OEM客戶的近線需求持續疲軟。由於近期宏觀經濟的不確定性，資訊長們的預算持續收緊。持續優化本地和雲端現有工作負載的努力正在幫助企業客戶推遲大規模儲存部署。這些趨勢減緩了我們大多數美國雲端客戶的庫存吸收速度。為了加快庫存吸收並保障我們的財務回報，我們已大幅減少了對幾家大型雲端客戶的出貨量。

We project it will take another couple of quarters for inventory levels to normalize. We also believe the timing for demand recovery could be affected by spending priorities focused on accelerating the build-out of compute-intensive AI infrastructure. While AI and ML have been around for quite some time, generative AI has quickly emerged as the next megatrend. This megatrend makes us as excited as ever about the long-term growth drivers for Seagate.

我們預計庫存水準還需要幾個季度才能恢復正常。我們也認為，加速建立運算密集型人工智慧基礎設施的支出重點可能會影響需求復甦的時機。雖然人工智慧和機器學習已經存在了相當長一段時間，但生成式人工智慧已迅速崛起，成為下一個大趨勢。這一大趨勢讓我們對希捷的長期成長動力充滿期待。

In addition to the ongoing migration of workloads to the cloud, which we believe is far from over, gen AI is expected to be a catalyst for data creation, underpinning future demand for mass capacity storage.

除了工作負載向雲端的持續遷移（我們認為這還遠未結束）之外，人工智慧有望成為資料創建的催化劑，以支撐未來對大容量儲存的需求。

In today's earliest stages of gen AI development, you're seeing the necessary first steps of building and training of AI models. These efforts require significant investment in compute architectures, and we're seeing that investment ramp today. The next stage of development will yield enterprise-specific use cases that leverage trained AI models to convert data into value-enhancing applications. As this phase plays out, cost-effective mass storage will be critical and we see HDD as a long-term beneficiary.

在當今人工智慧發展的早期階段，我們正見證著建構和訓練人工智慧模型的必要第一步。這些工作需要在運算架構上進行大量投資，而我們目前正看到這方面的投資不斷增加。下一階段的發展將催生企業特定的用例，這些用例利用經過訓練的人工智慧模型將資料轉化為增值應用程式。隨著這一階段的推進，經濟高效的大容量儲存將至關重要，而我們認為 HDD 將是長期的受益者。

We are already seeing examples of content creators generating high-definition images from text, which is growing in use, as evidenced by 4 of the top services [totaling] 20 million images each day. Development is well underway to create data-intensive videos and animations simply from voice commands. The adoption of AI-generated video bodes well for mass capacity storage. For context, today, nearly 4 million videos are uploaded daily to YouTube and their file sizes can be thousands of times larger than a single image.

我們已經看到一些內容創作者利用文字生成高清圖像的例子，這種做法的使用率正在不斷增長，四大頂級服務每天共生成2000萬張圖片就證明了這一點。僅透過語音指令即可建立資料密集型視訊和動畫的開發也正在順利進行中。人工智慧生成影片的普及預示著大容量儲存的良好發展。例如，如今每天有近400萬個影片上傳到YouTube，其檔案大小可能是單張圖片的數千倍。

Additionally, we believe that predictive AI will lead to advances in many fields, including science and health care. For example, predictive AI can be used to analyze large collection of medical images, many created by gen AI, to provide insight for early detection, prevent disease progressions and develop patient-specific treatments.

此外，我們相信預測性人工智慧將推動包括科學和醫療保健在內的許多領域的進步。例如，預測性人工智慧可用於分析大量醫學影像（其中許多影像由新一代人工智慧創建），為早期發現、預防疾病進展和開發針對特定患者的治療方案提供洞見。

Mass capacity data storage will remain both an enabler and beneficiary of these trends working in harmony with flash and DRAM memories to bring AI applications to bear. While flash will continue to feed high-performance compute engines, mass capacity HDDs will remain the most cost-efficient storage media to house the enormous volumes of data being generated and used for predictive analytics.

大容量資料儲存將繼續成為這些趨勢的推動者和受益者，並與快閃記憶體和 DRAM 記憶體協同工作，推動 AI 應用的發展。雖然快閃記憶體將繼續為高效能運算引擎提供支持，但大容量 HDD 仍將是最具成本效益的儲存介質，用於儲存產生並用於預測分析的大量資料。

Even in today's unsustainably low NAND pricing environment, HDDs are still roughly 5x more cost-efficient and comparable flash solutions on a per bit basis, and we do not project that gap to close in the next decade. This level of conviction is due in large part to our leading technology road map. Seagate is leveraging magnetic recording technology innovations such as HAMR across our mature 10-disk HDD platform, positioning us very well to meet increasing demand, including from data-intensive AI applications.

即使在當今NAND價格低得難以為繼的環境下，HDD的單位成本效率仍然比同類快閃記憶體解決方案高出約5倍，而且我們預計未來十年內這一差距不會縮小。這種信心很大程度源自於我們領先的技術路線圖。希捷正在我們成熟的10碟HDD平台上充分利用HAMR等磁力記錄技術創新，這使我們能夠充分滿足日益增長的需求，包括來自數據密集型AI應用的需求。

Given the current business climate, Seagate is focused on product development and execution and helping our customers qualify next-generation drive capacities. We are executing well on both fronts. We continue to deliver areal density and TCO advancements through our conventional PMR hard disk drive products. Development efforts on what may be our last PMR product are nearing completion and will extend drive capacities into the mid- to upper 20 terabyte range.

鑑於當前的商業環境，希捷專注於產品開發和執行，並幫助客戶獲得下一代硬碟的容量認證。我們在這兩方面都表現良好。我們將繼續透過傳統的PMR硬碟產品來提升面密度和整體擁有成本 (TCO)。我們最後一款PMR產品的開發工作已接近尾聲，並將把硬碟容量提升到20TB的中高階水準。

As mentioned earlier, our 30-plus terabyte product launch plan is fully intact and initial customer qualifications are progressing well. We are on track to begin volume ramp in early calendar 2024. We are also preparing qualifications with a broader number of customers, including testing for lower capacity drives targeting VIA and enterprise OEM workloads. While there's always work to do, I am pleased with the progress the product development teams have made during fiscal year 2023.

如前所述，我們30多TB的產品發布計畫已全面完成，初始客戶認證工作進展順利。我們預計在2024年初開始量產。我們也正在為更廣泛的客戶進行認證做準備，包括針對VIA和企業OEM工作負載的低容量硬碟進行測試。雖然工作總是有的，但我對產品開發團隊在2023財年的進展感到滿意。

Overall, Seagate has consistently demonstrated the ability to quickly adjust and execute at a high level on every factor within our control. We remain diligent in managing through near-term business conditions. At the same time, I'm excited by the tremendous long-term opportunities ahead brought about by existing and emerging megatrends underpinned by data. Cost-effective mass capacity storage is a critical enabler and Seagate is poised to deliver with strong technology road maps and improving financial leverage.

整體而言，希捷始終展現出快速調整和高效執行的能力，能夠掌控我們掌控的每一個因素。我們始終保持謹慎的管理，以應對短期業務狀況。同時，我對數據驅動的現有和新興大趨勢所帶來的巨大長期機會感到興奮。經濟高效的大容量儲存是關鍵的推動因素，希捷已準備好憑藉強大的技術路線圖和不斷提升的財務槓桿來實現這一目標。

I'll stop there and hand it over to Gianluca.

我就此打住，並將其交給詹盧卡。

Thank you, Dave. Amidst the current industry down cycle, we delivered results that demonstrate our ability to maintain disciplined production output and strong cost control measures.

謝謝你，戴夫。在目前產業下行週期中，我們所取得的業績證明了我們有能力維持嚴格的生產產出和強而有力的成本控制措施。

For the June quarter, we reported revenue of $1.6 billion, in line with our recent public commentary, and non-GAAP loss of $0.18 per share, slightly better than the midpoint of our guided range. Total hard disk drive revenue declined 14% sequentially to $1.4 billion, with shipment of 91 exabytes. Mass capacity revenue declined 20% sequentially to $1 billion, reflecting softer demand in the cloud nearline market, partially offset by an improvement in the VIA business.

截至6月當季，我們報告營收為16億美元，與我們近期的公開評論一致；非公認會計準則每股虧損0.18美元，略高於我們預期區間的中位數。硬碟總營收季減14%，至14億美元，出貨量為91EB。大容量儲存營收季減20%，至10億美元，反映出雲端近線市場需求疲軟，但VIA業務的改善部分抵銷了這項影響。

Shipment into the mass capacity markets totaled 75 exabytes compared with 104 exabytes in the March quarter. Mass capacity shipments as a percentage of total HDD exabytes was roughly 82% compared to the March quarter 88%. The expected slowdown in our cloud business led to nearline shipments of 55 exabytes, down 37% sequentially.

大容量市場出貨量總計75EB，而3月季為104EB。大容量出貨量佔HDD總EB出貨量的百分比約為82%，而3月季為88%。由於雲端業務預期放緩，近線出貨量為55EB，季減37%。

As Dave outlined earlier, we anticipate that it will take a couple more quarters for CSPs to consume inventory and demand to improve. Specific to the VIA market, revenue was up sequentially from what we believe was a demand drop in the March quarter. We expect a relatively stable VIA demand environment in the second half of the calendar year, although still below the run rate before the downturn.

正如Dave之前所述，我們預計CSP還需要幾個季度才能消化庫存，需求也會回升。具體到VIA市場，我們認為3月季度的需求有所下降，但營收季增。我們預計下半年VIA需求環境將相對穩定，但仍低於經濟低迷前的運行率。

Within the legacy market, revenue was $401 million, up 8% sequentially, reflecting higher demand for mission-critical products. The client and consumer markets were down slightly quarter-over-quarter, typical for a June quarter. Finally, revenue for non-HDD business decreased 15% sequentially to $218 million. As anticipated, current IT spending behavior in light of economic uncertainties impacted our enterprise system business.

傳統市場收入為4.01億美元，較上季成長8%，反映出對關鍵任務產品的需求增加。客戶端和消費者市場較上季略有下降，是6月季度的典型情況。最後，非硬碟業務收入季減15%，至2.18億美元。如預期，鑑於經濟不確定性，目前的IT支出行為影響了我們的企業系統業務。

We expect this headwind and the purchase timing for a couple of large customers to result in a meaningful decline in our system revenue in the September quarter. Importantly, we shipped our first HAMR-based Corvault system for revenue as planned during the June quarter. We expect broader availability of these Corvault systems by the end of calendar 2023.

我們預計，這一不利因素以及幾家大客戶的採購時機將導致我們9月季度的系統收入大幅下降。重要的是，我們已於6月季度按計劃交付了首款基於HAMR的Corvault系統，並實現了營收成長。我們預計到2023年底，這些Corvault系統將會得到更廣泛的應用。

Moving to our operational performance. Non-GAAP gross profit in the June quarter was $313 million, down 10% sequentially, reflecting lower revenue and a less favorable mix, partially offset by lower underutilization costs. Non-GAAP gross margin expanded 80 basis points sequentially, approaching 20%. Underutilization costs decreased to approximately $40 million in the June quarter as we temporarily increased production to build long lead time components.

談到我們的營運表現。 6月當季非公認會計準則毛利潤為3.13億美元，環比下降10%，這反映了收入下降和產品組合不佳，但部分被未充分利用成本的降低所抵消。非公認會計準則毛利率較上季成長80個基點，接近20%。由於我們暫時增加了產量以生產長交付週期組件，6月當季未充分利用成本降至約4000萬美元。

Based on the current business outlook that Dave outlined, we expect higher underutilization costs in the September quarter. However, we expect gross margin to modestly increase as our cost structure improvements are fully realized and we implement pricing adjustments in certain markets during the second half of calendar 2023.

根據Dave概述的當前業務前景，我們預計9月份季度的未充分利用成本將有所上升。然而，隨著成本結構優化的全面實施，以及我們將在2023年下半年在某些市場實施價格調整，我們預計毛利率將小幅上升。

We reduced non-GAAP operating expenses to $258 million, down $91 million year-over-year and $24 million sequentially. The year-over-year decline reflects our cost structure improvement actions to date, lower variable compensation as well as disciplined cost management. We expect non-GAAP OpEx in the September quarter to be similar to the June quarter, reflecting nearly full realization of cost savings from restructuring efforts announced in April, offsetting an increase related to the recent sales and leaseback of 3 nonmanufacturing sites.

我們將非公認會計準則營運支出降至2.58億美元，年減9,100萬美元，季減2,400萬美元。同比下降反映了我們迄今為止的成本結構改進措施、較低的浮動薪酬以及嚴格的成本管理。我們預計9月份季度的非公認會計準則營運支出將與6月份季度持平，這反映了4月份宣布的重組措施帶來的成本節約幾乎全部實現，抵消了近期3個非製造基地售後回租帶來的成本增長。

Moving on to the balance sheet and cash flow. We ended the June quarter with liquidity level of approximately $2.3 billion, including our revolving credit facilities of $1.5 billion, which was lowered by $250 million as previously disclosed in the credit agreement amendment. Inventory was down $60 million quarter-over-quarter at $1.14 billion and we will continue to focus on reducing our own inventory in the coming quarters.

再來看看資產負債表和現金流。截至6月當季，我們的流動性水準約為23億美元，其中包括15億美元的循環信貸額度，根據先前信貸協議修訂案的披露，該額度減少了2.5億美元。庫存季減6000萬美元，至11.4億美元，未來幾季我們將繼續致力於減少自身庫存。

Capital expenditure totaled $50 million in the June quarter, down 7% sequentially. Fiscal year '23 CapEx was $316 million or roughly 4% of revenue. We target lower CapEx in fiscal year '24 as we maintain our focus on shifting to a build-to-order factory loading, balancing supply to demand and still supporting our innovation-driven product road map.

6月當季資本支出總計5000萬美元，季減7%。 23財年資本支出為3.16億美元，約佔營收的4%。我們計劃在24財年降低資本支出，因為我們將繼續專注於轉向按訂單生產的工廠，平衡供需，並繼續支持我們以創新為導向的產品路線圖。

Free cash flow generation was $168 million for the quarter and $626 million for the fiscal year, reflecting solid operational execution and efficiency in working capital management. We expect free cash flow to be lower in the September quarter as we incur a majority of the cash payments associated with the restructuring charges announced in April. We used $145 million for the quarterly dividend and exited the quarter with 207 million shares outstanding. There were no share repurchases during the quarter.

本季自由現金流為1.68億美元，本財年為6.26億美元，反映了穩健的營運執行力和高效的營運資本管理。我們預計9月份季度的自由現金流將有所下降，因為我們的大部分現金支出與4月份宣布的重組費用有關。我們使用了1.45億美元作為季度股息，本季末流通股數為2.07億股。本季沒有股票回購。

Our debt balance equity in the quarter was $5.5 billion, down $507 million quarter-over-quarter, primarily reflecting the 2023 notes retirement. This debt reduction was mostly funded through the sales and leaseback of the 3 facilities noted earlier. Additionally, we raised $1 billion in new notes and used the proceeds to prepay $450 million of existing term loans and retired the March 2024 notes of $500 million. As a result of these actions, we now have less than 12% of total debt coming due over the next 2 fiscal years.

本季我們的債務餘額為55億美元，季減5.07億美元，主要反映了2023年票據的到期償還。此次債務削減主要透過前面提到的3筆融資工具的售後回租籌集。此外，我們發行了10億美元的新票據，並用所得款項預付了4.5億美元的現有定期貸款，並到期償還了5億美元的2024年3月到期的票據。由於這些措施，我們目前有不到12%的總債務將在未來兩個財年到期。

Interest expense in the June quarter was $84 million and is expected to increase slightly for the September quarter, reflecting a full quarter of the debt restructuring actions that I just described. Adjusted EBITDA for the last 12 months totaled $974 million, resulting in a net debt leverage ratio of 4.8x, well below our amended credit agreement terms.

6月當季利息支出為8,400萬美元，預計9月當季將略有增加，這反映了我剛才描述的整個季度的債務重組行動。過去12個月的調整後EBITDA總計9.74億美元，淨負債槓桿率為4.8倍，遠低於我們修訂後的信貸協議條款。

Turning to our September quarter outlook. For the hard drive business, we expect incremental improvements in mass capacity to offset declines in the legacy market. For non-HDD, revenue for the system business is expected to decrease sequentially due in part to demand timing among a couple of our large system customers, as noted earlier.

回顧我們9月季度的展望。對於硬碟業務，我們預期大容量將逐步提升，以抵銷傳統市場的下滑。對於非硬碟業務，預計系統業務的收入將環比下降，部分原因是我們幾家大型系統客戶的需求時間安排，如前所述。

With that as context, we expect September quarter revenue to be in the range of $1.55 billion plus or minus $150 million. At the midpoint of our revenue guidance, we expect non-GAAP operating margin to be in the low to mid-single-digit range, which includes the proactive measure we discussed today, partially offset by higher underutilization costs in the range of $70 million. And we expect a non-GAAP loss per share in the range of $0.16 plus or minus $0.20.

基於此，我們預計9月當季營收將在15.5億美元上下浮動1.5億美元。根據我們營收預期的中位數，我們預計非公認會計準則（Non-GAAP）營業利潤率將處於低至中等個位數區間，這已包含我們今天討論的主動措施，但部分抵消了約7000萬美元的未充分利用成本。我們預計非公認會計準則每股虧損將在0.16美元上下浮動0.20美元。

I will now turn the call back to Dave for final comments.

現在我將把電話轉回給戴夫，請他發表最後的評論。

Thanks, Gianluca. I'm proud of the perseverance and agility as we continue to proactively respond to the market environment we have faced over this past year. Entering fiscal 2024, we are leaner, our balance sheet is healthier and our product road map is stronger. All of these factors enhance our ability to weather near-term market conditions, deliver financial leverage and position the company to capture attractive long-term opportunities for mass capacity storage.

謝謝，詹盧卡。我為我們在過去一年中持續積極應對市場環境的毅力和敏捷性感到自豪。邁入2024財年，我們的業務更加精簡，資產負債表更加健康，產品路線圖也更加完善。所有這些因素都增強了我們應對短期市場狀況、發揮財務槓桿作用的能力，並使公司能夠抓住極具吸引力的長期大容量儲存機會。

As we conclude our prepared remarks, I would like to thank our employees, suppliers, customers and shareholders once again for their support. Thanks for joining us, and let's open up the call for questions.

在我們結束準備演講之際，我想再次感謝我們的員工、供應商、客戶和股東的支持。感謝各位的參與，現在開始提問環節。

(Operator Instructions) The first question today comes from Krish Sankar with TD Cowen.

（操作員指示）今天的第一個問題來自 TD Cowen 的 Krish Sankar。

This is Eddy for Krish from TD Cowen. I understand you won't guide for December, but curious to know your thoughts on December nearline exabyte shipments. Like do you see them growing sequentially from September, more like flat or maybe slightly down? I'm asking because in case we remain at current revenue of $1.6 billion per quarter into next year, I'm wondering if there would be a risk to the dividend, in case, especially you guys continue to invest in the business.

我是 TD Cowen 的 Krish 的 Eddy。我知道你們不會發布 12 月份的業績指引，但我好奇地想知道你們對 12 月份近線 EB 出貨量的看法。您認為 12 月的近線 EB 出貨量會比 9 月環比成長，還是持平或略有下降？我之所以問這個問題，是因為如果我們明年的季度營收仍保持在目前的每季 16 億美元水平，我想知道股息是否會面臨風險，尤其是在你們繼續投資業務的情況下。

Yes. Thanks, Eddy. So we aren't guiding for December. I think we expect at this point to hold the line on overbuilding. And so that's one of the reasons we're being very cautious. But there will be some -- there's an indication of some natural flow-through at the cloud service providers, for example. So we expect it to slowly start trending back up from here. I don't expect a big hockey stick rebound. I mean I could be wrong on that, but -- and we do expect it to slowly go back up mainly because we're not pushing in a bunch of our own inventory.

是的，謝謝，Eddy。所以我們沒有為12月制定預期。我認為我們目前預計會控製過度建設。這也是我們非常謹慎的原因之一。但會有一些——例如，有跡象顯示雲端服務供應商會有一些自然的產能流出。所以我們預計它會從現在開始緩慢回升。我不認為會出現大幅反彈。我的意思是，我可能錯了，但是——我們確實預計它會緩慢回升，主要是因為我們沒有大量囤積自己的庫存。

And relative to the dividend, we're planning for all this in all of our plans right now. And I think the way we execute this plan, we should be just fine and we want to continue to be the person for all of our stakeholders that we've always been before. We've got laser focus on that right now.

至於股息，我們目前在所有計劃中都已考慮到了這一點。我認為，按照我們執行這項計劃的方式，應該不會有問題，我們希望繼續像以往一樣，為所有利害關係人服務。我們現在正全神貫注於此。

The next question comes from Wamsi Mohan with Bank of America.

下一個問題來自美國銀行的 Wamsi Mohan。

I was wondering, Dave, if you could flush out a little bit 2 of the points that you made in your prepared remarks, one on adjusting pricing strategy in certain markets. What specifically are you doing in these markets? And then secondarily, the reduced production output through this build-to-order and lower CapEx comments that you made. How exactly will this build-to-order work? Is this specifically for mass capacity? What kind of lead times would you require? And are your customers? And what do you expect follow-through from customers and also from a competitive standpoint?

戴夫，我想問您在準備好的發言中提到的兩點。第一點是關於調整特定市場的定價策略。您具體在這些市場做了哪些工作？第二點是關於透過按訂單生產和降低資本支出來減少產量。這種以訂單生產具體是如何運作的？是專門針對大規模產能的嗎？您需要什麼樣的交付週期？您的客戶也是這樣嗎？從競爭的角度來看，您希望客戶能夠如何跟進？

Thanks, Wamsi. Yes, it is complicated because there's a number of different markets in play here. So for example, a distribution channel or some of the classic markets we have consumer client server, you just don't move things very quickly. But some places, it's relatively easier to raise prices and have that flow through into the system.

謝謝，Wamsi。是的，這很複雜，因為涉及多個不同的市場。例如，分銷管道或一些我們擁有消費者客戶伺服器的經典市場，你的行動速度不會很快。但在某些地方，提高價格並讓這些價格流入系統相對容易。

On mass capacity drives, generally speaking, we do the price increases through transitions of products. So we'll say this is the changing economics, we need to make sure that we get paid for that. So as we go through the next product generation, there's still a TCO benefit to the end customer to drive through that product transition, but the economics have to change, largely so that we can go back and pay our suppliers and keep everybody healthy.

對於大容量硬碟，我們通常會透過產品轉型來提高價格。因此，我們會說這是經濟情勢的變化，我們需要確保我們能從中獲得回報。因此，隨著我們推出下一代產品，推動產品轉型仍然會為最終客戶帶來總體擁有成本 (TCO) 效益，但經濟狀況必須改變，這主要是為了讓我們能夠回頭支付供應商的費用，確保所有人的健康。

There's been such a profound downturn in demand that we have to make sure that the entire supply chain is treated properly. But exactly what you said, different markets are behaving very differently and accept things very differently. And this is the direction we're pushing at this point.

需求下滑如此嚴重，我們必須確保整個供應鏈得到妥善處理。正如您所說，不同市場的行為方式和接受方式截然不同。而這正是我們目前正在努力的方向。

Yes. On the cost side, Wamsi, and on the production that is, of course, driving underutilization. We had a little bit higher production in the June quarter, mainly on the wafer side and a little bit on the media, but I would say more on the wafer. We will not need to do that level of production right now. So the underutilization charges were declined in the June quarter from almost $80 million to $40 million. It's probably going up a little bit in September. We estimated about $70 million. But of course, through the quarter, that could change a little bit.

是的。 Wamsi，成本方面，以及生產方面，當然，產能利用率不足也是原因之一。我們在第二季的產量略有提高，主要是晶圓方面，介質方面也略有提高，但我認為晶圓方面的產量更高。我們現在不需要達到這個產量水準。因此，第二季的產能利用率不足費用從近8,000萬美元下降到4,000萬美元。九月可能會略有上升，我們估計約為7000萬美元。當然，整個季度，情況可能會有所改變。

The next question comes from Timothy Arcuri with UBS.

下一個問題來自瑞銀的提摩西‧阿庫裡。

I also had a question on pricing and sort of how you think about it versus market share. We've heard examples of you raising prices 10% to 15%. It's obviously a little bit counterintuitive in such a soft market. So I guess 2 questions. One, how are customers reacting to that? And then also, are you willing to lose share? As you kind of look out, is there like a line in the sand on share that you're going to draw where you wouldn't raise pricing if you lost share? Or are you willing to lose share to get pricing reset to a more acceptable level?

我還想問一個關於定價的問題，以及您如何看待定價與市場份額的關係。我們聽過您漲價10%到15%的例子。在如此疲軟的市場環境下，這顯然有點違反直覺。所以我想問兩個問題。第一，客戶對此有何反應？第二，您願意失去市場佔有率嗎？您認為，在市佔率方面，您是否設定了一條底線，如果市場佔有率下降，您就不會漲價？或者，您願意為了將價格調整到一個更可接受的水平而失去市場份額？

Yes. Thanks, Tim. So yes, you are right. It's a weak demand environment, but I think also there's -- generally speaking, the whole industry is underwater. So -- I mean we have to make sure we protect ourselves, we protect the supply base as we talked about earlier. And it's different in different products across the market, obviously, the different markets, obviously.

是的，謝謝，提姆。所以，你說得對。這是一個需求疲軟的環境，但我認為——總體而言，整個產業都處於困境之中。所以——我的意思是，我們必須確保保護好自己，保護好供應基礎，就像我們之前提到的。顯然，不同市場的產品情況會有所不同。

How is it being accepted? There's -- you can imagine the entire panoply of different responses. Some people are worried about supply longer term. They do believe that supply will be short and demand will come back. And so -- and especially with the long, long lead times that there are on mass capacity products right now, I think it's something that people are very much in tune with. And so it's forcing good discussions. There's other people for various reasons that need to make tactical procurement decisions and they'll take whatever deals that are out in front of them. And I get that. I mean that happens from time to time in the business. That's one of the big reasons why we don't want to overbuild. We want to make sure we hold as much as we possibly can.

它是如何被接受的？你可以想像各種各樣的反應。有些人擔心長期供應問題。他們確實相信供應會短缺，需求會回升。所以，尤其是在目前大容量產品的交貨週期很長的情況下，我認為人們非常認同這一點。因此，這促使人們進行良好的討論。還有一些人出於各種原因需要做出戰術採購決策，他們會接受擺在他們面前的任何交易。我理解這一點。我的意思是，這種情況在行業中時有發生。這也是我們不想過度建設的重要原因。我們希望確保盡可能多地保有庫存。

And as we continue to play this forward, I think we'll see and adjust accordingly to make sure that we defend our products. I think the optimism for us is that we can actually go through the product transitions, get to the higher capacity points to provide a better TCO proposition, get components out of the supply chains and so on. And then we didn't dilute or extend the problem any by overbuilding or dilute the problem by continuing to overbuild and then lowering price. And that's what we have to hold on to.

隨著我們繼續推進這項工作，我認為我們會看到並相應地做出調整，以確保我們能夠捍衛自己的產品。我認為，我們樂觀地認為，我們能夠真正完成產品轉型，達到更高的產能點，從而提供更優的總體擁有成本 (TCO) 方案，並將零件從供應鏈中剝離出來等等。這樣，我們就不會透過過度建設來淡化或擴大問題，也不會透過繼續過度建設然後降低價格來淡化問題。這就是我們必須堅持的。

The next question comes from Erik Woodring with Morgan Stanley.

下一個問題來自摩根士丹利的艾瑞克‧伍德林。

Awesome. Dave, I just wanted to ask if you could flesh out the comment you made in your prepared remarks about the timing of the demand recovery being delayed by AI investments. Just in the context of you talking about market conditions weighing on demand through December, are you telling us we shouldn't expect sequential growth in capacity shipped until calendar '24? Just if you could help us connect some of these thoughts, that would be great.

太棒了。戴夫，我只是想問一下，您能否補充一下您在準備好的演講中提到的，由於人工智慧投資，需求復甦的時機被推遲。您剛才提到，市場狀況會影響到12月份的需求，您是想告訴我們，我們不應該預期到2024年，出貨量會較上季成長嗎？如果您能幫助我們理清這些想法，那就太好了。

Thanks, Erik. Yes. And I think I mentioned in one of the earlier comments that there will be slight capacity growth. But the way I read through most of the results that are already coming through in the cloud and in most of my conversations as well, I think it's fairly consistent that there's priorities being put on data center infrastructure, new data center buildout, power infrastructure that was already kind of -- things were biased that way instead of mass capacity storage, the mass capacity being something that would come later. And we all know about the great new capabilities that there are in some of the compute technologies and how those new technologies are being applied in new applications. So that's AI. And that's a high priority for cloud service providers and people doing stuff on-prem as well.

謝謝，Erik。是的。我之前在評論中提到過，容量會略有增長。但從我閱讀的雲端運算領域目前公佈的大部分結果以及我的大部分對話來看，我認為大家的優先事項基本一致，那就是數據中心基礎設施、新數據中心的建設以及電力基礎設施——這些已經有點偏向於大容量存儲，而大容量存儲是以後才會出現的東西。我們都知道一些計算技術擁有強大的新功能，以及這些新技術是如何應用於新應用的。這就是人工智慧。對於雲端服務供應商和本地部署人員來說，人工智慧也是高度優先考慮的。

We think data follows behind that. Predicting it is going to be really tough, but we do think that, that's going to be a bow wave. At some point, there will be competition for the data that actually feeds the AI engines, if you will. Right now, there's a big battle to get the AI engines up and running and show those applications as being meaningful, and I think that's an interesting space to watch. Behind it, there's going to be a lot of data that actually has to feed those AI engines every day, every day, every day, and we think that's mass capacity infrastructure net positive.

我們認為數據緊隨其後。預測這一點非常困難，但我們確實認為，這將是一股強勁的浪潮。在某種程度上，人們將對真正驅動人工智慧引擎的數據展開競爭。目前，一場關於啟動和運行人工智慧引擎並展示其應用價值的激烈競爭正在進行，我認為這是一個值得關注的有趣領域。在這背後，將有大量數據需要每天、每天、每天地驅動這些人工智慧引擎，我們認為這對大容量基礎設施來說將是一個淨利好。

The next question comes from Thomas O'Malley with Barclays.

下一個問題來自巴克萊銀行的托馬斯·奧馬利。

In the deck, you guys mentioned that you were extending your PMR platform into the mid- to upper 20 TB capacities. Could you talk about where your original plan was to extend into the mid-20s and where that is now and why you made that decision?

在演講中，你們提到正在將PMR平台的容量擴展到20TB左右。能否談談你們最初的計劃，以及目前的進展，以及為什麼做出這個決定？

Yes, Tom, there's no real change. We've been talking about it, I think, for about a year. We said we can get PMR into the mid-20s. And the product that we have now -- let me -- I'll talk about it very quickly here. We believe that with SMR variants that can get in the high 20s, the product is in the mid-20s, but it's highly, highly leveraged. It's almost identical electronics and mechanics that takes us to HAMR. So it's almost the same product. You swap out the heads and disks for HAMR and a couple of other component changes that have to happen.

是的，湯姆，沒有真正的改變。我想我們已經討論了大約一年。我們說過，我們可以把PMR的磁頭頻率控制在25%左右。至於我們現在的產品——讓我——我會在這裡快速介紹一下。我們相信，SMR的磁頭頻率可以達到25%左右，而我們的產品頻率在25%左右，而且它的利用率非常高。 HAMR的電子元件和機械結構幾乎完全相同。所以它幾乎是同一款產品。你需要將磁頭和磁碟換成HAMR，以及其他一些必須進行的元件變更。

So we're going to get a lot of leverage from that. That platform has been planned for a long, long time. We called it a 10-disk platform in the prepared remarks. So I have a lot of confidence that we can solution PMR anywhere that the customer wants to stay there in the mid-20s or we can go to HAMR variants with the same bag of parts or very, very similar bag of parts.

因此，我們將從中獲得巨大的優勢。該平台已經規劃了很長時間。我們在準備好的演講稿中稱之為10盤平台。因此，我非常有信心，我們可以在任何客戶希望保持在20多兆赫茲的範圍內解決PMR問題，或者我們可以使用相同或非常相似的零件來開發HAMR變體。

I do think that there's a lot of confusion out in the market, I'll say that this way about what capacity point, is it 24, 25, 26. Look, there's customers who need all kinds of different variants. And I think we confuse people, frankly, by kind of amalgamating all these things. All I will say is that I think those products that we have are very, very competitive in the market.

我確實認為市場上有許多困惑，例如容量點，像是24、25、26。你看，有些客戶需要各種不同的型號。坦白說，我們把所有這些東西混在一起，反而讓人感到困惑。我想說的是，我認為我們現有的產品在市場上非常有競爭力。

Right now, the issue is if I added -- I'm not going to try to force it into the market into a space that people don't need the product simply because it poisons the well for that product maybe 2 quarters out or 3 quarters out for when people do need it, and it would erode the value proposition for the HAMR products when they come. So we've got to run the business the way we are right now.

現在的問題是，如果我補充一點——我不會試圖把它強行推向一個人們不需要它的市場，因為這會毒害該產品的市場，尤其是在人們真正需要它的時候，可能還要等兩個季度或三個季度，而且這會削弱HAMR產品的價值主張。所以我們必須按照現在的方式經營業務。

Yes. Let me add to that just for clarity, when we say the mid-20s is in the CMR version. If you go to the SMR version of that drive, of course, we are in the mid- to high 20s.

是的。為了更清楚起見，我補充一下，我們說的25%左右是指CMR版本。當然，如果是SMR版本，也就是25%左右。

The next question comes from Aaron Rakers with Wells Fargo.

下一個問題來自富國銀行的 Aaron Rakers。

Aaron, sorry, we can't hear you.

亞倫，抱歉，我們聽不到你說話。

Yes, sorry about that. I just wanted to unpack some of the kind of considerations we should be thinking about in terms of gross margin as we move forward. Appreciating that you're not guiding explicitly. But I guess, first of all, if I look at the pricing discussion, based on the math, it looks like your mass capacity dollar per terabyte was up by about 10% sequentially. I guess as we look forward, should we be thinking that the ASP on that basis continues to trend higher? Or how much of that capacity shift have you already kind of imposed price increases upon?

是的，很抱歉。我只是想解釋一下，在未來的發展中，我們應該在毛利率方面考慮哪些因素。感謝您沒有給予明確的指導。但我想，首先，如果我看一下定價討論，根據計算結果，您的大容量每TB價格似乎環比上漲了約10%。展望未來，我們是否應該認為在此基礎上的平均售價會繼續走高？或者，您已經對多少容量轉移施加了價格上漲？

And then also, when we look at the capacity shift in total with a 25% reduction in production capacity, do we take prior peak capacity shipment levels and take 25% of that, and that should be a level for which you think you would get back to kind of that 30% gross margin level? I know there's a lot in that, but I'm just trying to think about some of the variables, looking out just not this current quarter but beyond.

另外，當我們考慮產能轉移時，如果產能減少25%，我們是否應該以之前的峰值產能出貨量水準為基礎，再減去25%，這樣您認為毛利率就能回到30%左右的水平？我知道這其中有很多因素，但我只是想考慮一些變量，不僅著眼於當前季度，也著眼於未來。

Thanks, Aaron. I'll take my cut, and then I'll give it to Gianluca to answer. I think as I look in at the tactics of what's going on right now, I think it's dangerous to say what changed quarter-over-quarter and extrapolate from there. You can draw straight lines, but I'm not sure that -- like whether we take a deal at the end of a quarter or not is a good way to think about what's going to happen over the long haul.

謝謝，Aaron。我會拿走我的分成，然後交給Gianluca來回答。我認為，當我審視目前的策略時，我認為說出季度環比變化並以此推斷是很危險的。你可以畫一條直線，但我不確定——例如我們是否在一個季度末達成協議，是否是思考長期發展的好方法。

We've said that pricing needs to go up or stabilize depending on the different capacity points, the product transitions that we're going through, whether or not suppliers still want to stay around for a certain product line. I mean there's a lot of different things to unpack there, to your point.

我們說過，價格需要上漲或穩定，這取決於不同的產能點、我們正在經歷的產品轉型，以及供應商是否仍願意繼續生產某個產品線。我的意思是，正如您所說，這裡面有很多不同的因素需要分析。

Relative to our manufacturing capacity, we talked about 25%. Of course, that's not perfect across every single factory we have, nor is it perfect across the supply chain. And we're only as good as our weakest link. We have to make sure those weakest links are still well cared for. And so that's largely what's driving that. The ability to go put that capacity back online would -- is severely hampered because it's not just the capital, if you will, that we haven't invested in enough for peak capacity, but it's also the people we'd have to bring back. I mean, unfortunately, our people and a lot of people in the supply chain have been dramatically taken down.

相對於我們的製造能力，我們談到了25%的產能。當然，這並不適用於我們所有的工廠，也並非適用於整個供應鏈。我們的表現取決於我們最弱的環節。我們必須確保這些最薄弱的環節仍然得到妥善維護。這就是我們採取這項措施的主要原因。恢復這些產能的能力將受到嚴重阻礙，因為我們不僅在資本方面（如果你願意這麼說的話）沒有投入足夠的資金來達到峰值產能，而且還需要召回一些員工。我的意思是，不幸的是，我們的員工以及供應鏈中的許多人都大幅裁員。

So do you want to add something, Gianluca?

那麼，詹盧卡，你還有什麼要補充的嗎？

Yes, Aaron, you were asking about the price per terabyte in particular in Q4. That is a result also of the mix. So we have a reduction in cloud. That, of course, is our higher capacity that usually has the lowest price per terabyte, but also the lowest cost per terabyte. And we have an increase in VIA. So mid-capacity drives with, again, higher price per terabyte and a little bit higher cost per terabyte. So there is a mix that is impacting. Of course, when the volume is not very high, just a few exabyte difference in the different segments can swing those price and cost per terabyte.

是的，Aaron，您問的是第四季每TB的價格。這也是產品組合的影響。我們的雲端業務有所減少。當然，我們的高容量產品通常每TB價格最低，每TB成本也最低。而VIA業務增加。中容量硬碟的每TB價格較高，每TB成本也略高。所以，這種產品組合是有影響的。當然，當容量不是很高時，不同細分市場中僅幾EB的差異就足以影響每TB的價格和成本。

In terms of capacity, Aaron, the 25%, I would not take that as the peak of where we were probably more than a year ago. We have done a lot of restructuring, a lot of adjustments. And as Dave said, we will need time to eventually rebuild that capacity also in terms of employees.

就產能而言，Aaron，那25%的產能，我不認為這是我們一年多前水準的頂峰。我們做了很多重組，很多調整。正如Dave所說，我們最終需要時間來重建產能，包括員工方面。

Yes. We don't need to get back to the full capacity in order to get back into our margin model because we've taken so many cost actions across the business. I think we've been very, very aggressive on that. Knowing that our footprint had to change, I think we'll be able to scale much better from a margin perspective as some of the demand comes back.

是的。我們不需要為了恢復利潤率模式而恢復滿載運轉，因為我們已經在整個業務範圍內採取了許多成本控制措施。我認為我們在這方面已經非常積極了。考慮到我們的佈局必須改變，我認為隨著部分需求的回升，從利潤率的角度來看，我們將能夠更好地擴展規模。

The next question comes from C.J. Muse with Evercore ISI.

下一個問題來自 Evercore ISI 的 C.J. Muse。

This is Kurt Swartz on for C.J. Hoping to delve into generative AI implications for both HDD and SSD demand. What is the perceived time line for storage demand uplift as data creation accelerates? And is there any concern regarding more workloads moving from HDD to flash within an AI data center paradigm?

我是 Kurt Swartz，代表 C.J. 進行訪談。希望深入探討生成式人工智慧對 HDD 和 SSD 需求的影響。隨著資料創建的加速，儲存需求預期成長的時間軸是怎麼樣的？在 AI 資料中心範式中，是否有人擔心更多工作負載會從 HDD 轉移到快閃記憶體？

Yes, thanks. I don't think of things that way that people in data centers are not making decisions like you would make. For example, if you're building your own PC, this or that or this combination. Instead, they know how to manage data across multiple tiers and geographies very, very well. And make sure that the right data gets to right place at the right time.

是的，謝謝。我不這麼認為，資料中心的人不會像你那樣做決定。例如，如果你要組裝自己的電腦，他們會選擇這個、那個或這個組合。相反，他們非常非常了解如何跨層級和跨地域管理數據，並確保正確的數據在正確的時間到達正確的位置。

There are some gen AI discussions that are very transactional and probably don't need a whole lot of HDD and then there's a lot that are very video-intensive and want data to be sequestered in a certain tank and -- so that nothing else can touch it. And we believe that's going to be that normal tier that you see today. So I think it's a win-win for memory and storage long term. Obviously, it's win for compute.

有些關於新一代人工智慧的討論非常注重事務性，可能不需要大量的硬碟；但也有很多討論涉及視訊密集型數據，希望將數據隔離在特定的容器中，這樣其他任何東西都無法觸及。我們相信這將成為您今天看到的那種常規層級。所以我認為從長遠來看，這對內存和存儲來說是雙贏的。顯然，這對計算來說也是雙贏的。

And I think as some of these new applications are lit up, I'm really excited to see it. We haven't seen it yet. I mean unless you start to look back at some traditional cloud service provider capabilities and say that all along that was AI. But I do think there's a whole host of new tools coming. It will be a very competitive space. And the data tiering infrastructure will become a critical part of that as we go forward.

我認為，隨著一些新應用的蓬勃發展，我對此感到非常興奮。我們還沒有看到這樣的發展。我的意思是，除非你回顧一些傳統雲端服務供應商的功能，並認為這些功能一直以來都是人工智慧。但我確實認為，將會有大量新工具湧現。這將會是一個競爭非常激烈的領域。隨著我們未來的發展，資料分層基礎設施將成為其中的關鍵組成部分。

The next question comes from Karl Ackerman with BNP.

下一個問題來自法國國家銀行的卡爾·阿克曼。

I have 2 questions. First, I may have missed this, but why did average capacity per drives for the mass capacity area of your business shrink this quarter? And then second, I guess, how should we think about mass capacity unit growth from here? And where do you believe we are in the replacement cycle and whether that should continue to elongate or whether that returns to some form of normal from a unit perspective?

我有兩個問題。首先，我可能忽略了這一點，但為什麼本季貴公司大容量業務的平均單驅動器容量會下降？其次，我們該如何看待未來大容量產品單元的成長？您認為我們目前處於更換週期的哪個階段？從產品單元的角度來看，更換週期會繼續延長，還是會恢復到某種正常水準？

Thanks, Karl. So the -- we are not overbuilding, especially the high-capacity drives are trying to push out into some of the cloud service providers. We're waiting until the inventory actually goes down. And so I think that's why -- and I made reference to this earlier, that's why some of the metrics may look weird, like you just pointed out on capacity per drive diving down. I don't think that's a long-term trend either. I think, of course, that's going to go back up.

謝謝，Karl。所以我們並沒有過度建設，尤其是那些試圖向一些雲端服務供應商推銷的大容量硬碟。我們正在等待庫存真正下降。所以我認為這就是為什麼——我之前提到過，這就是為什麼一些指標看起來很奇怪，就像你剛才指出的，每個硬碟的容量正在下降。我也不認為這是長期趨勢。當然，我認為這個數字還會回升。

Yes. The average capacity is down because of mix. So as I said before, in the June quarter, we shipped less to cloud and more to VIA. And of course, the capacity is much higher in the cloud business.

是的。由於產品組合的原因，平均產能有所下降。正如我之前所說，6月份季度，我們向雲端的出貨量減少，向VIA的出貨量增加。當然，雲端業務的產能也高得多。

I'm sorry, Karl, what was the second part of your question?

抱歉，卡爾，你的問題的第二部分是什麼？

Yes. Sorry about that. It's about how we should think about mass capacity unit growth from here? And where do you believe we are in the replacement cycle? I asked because it appears your mass capacity unit shipments peaked in the first quarter of 2022. So it's been about 5 consecutive quarters of decline. At the same time, the units are down over 2/3 in that same period. And so if we could just discuss perhaps the corollary between units and exabyte growth, I think it's relevant here, would be helpful.

是的，抱歉。問題是我們該如何看待今後的大容量儲存單元成長？您認為我們目前處於更換週期的哪個階段？我之所以問這個問題，是因為您的大容量儲存單元出貨量似乎在2022年第一季達到高峰。也就是說，已經連續五個季度下滑。同時，儲存單元數量也在同一時期下降了超過三分之二。所以，如果我們能討論一下儲存單元數量和EB成長之間的必然關係，我認為這與此相關，會很有幫助。

There are hundreds of -- sorry, hundreds of millions of cloud drives, if you will, in data centers around the world. And some of them are aging to your point. And if unit shipments go down -- exabytes can still go up even though unit shipments are going down because the ones that are aging off are lower capacity points, say, 4, 8, 12 terabytes whatever. But over time, it's much more efficient. Actually, it's more efficient power-wise and for a lot of other features, system-level features, to put the newer drives on.

全球各地的資料中心裡有數百個——抱歉，應該說是數億個雲端硬碟。正如您所說，其中一些正在老化。如果單位出貨量下降，EB 容量仍可能上升，因為老化的硬碟容量較低，例如 4TB、8TB 或 12TB。但隨著時間的推移，效率會更高。實際上，在功耗方面，以及在許多其他功能（系統級功能）方面，安裝較新的硬碟都更有效率。

So it's about all these things in balance. And I do believe that unit shipments are down largely because of this inventory digestion that we've been talking about. That will turn again. Will it go back to the peak of where it was 5 or 6 quarters ago and how fast will it go back? These are tough questions for our customers right now. It comes down to their prioritization of their spend. But all indications are in the latest numbers.

所以，關鍵在於所有這些因素之間的平衡。我確實認為，單位出貨量下降很大程度上是因為我們一直在談論的庫存消化。這種情況會再次好轉。它會回到五、六個季度前的峰值嗎？回升的速度有多快？這些對我們的客戶來說都是目前棘手的問題。最終取決於他們如何優先安排支出。但所有跡像都體現在最新的數據中。

And if I think about these latest numbers that the cloud service providers are announcing versus where they were 6 to 9 months ago, there is a very positive trend. And I think, to the questions about gen AI and other feature sets, we're not a peak cloud yet. So I do think it will come back fairly strongly from a unit's perspective when it does.

如果我把雲端服務供應商公佈的最新數據與6到9個月前的數據進行比較，就會發現一個非常正面的趨勢。而且我認為，關於新一代人工智慧和其他功能集的問題，我們還沒有達到雲端運算的頂峰。所以我認為，當雲端運算達到頂峰時，從業務部門的角度來看，它會以相當強勁的勢頭回歸。

The next question comes from Tristan Gerra with Baird.

下一個問題來自貝爾德的特里斯坦·傑拉 (Tristan Gerra)。

Couple of questions. The first one is about the production cost curve. And if you could talk about the percentage decline per year that we should expect in the medium term on a per bit basis? And then second question about -- is the percentage of production that's coming from China and also the true demand that you see from China as a percentage of your revenue?

幾個問題。第一個問題是關於生產成本曲線。您能否談談我們預計中期內每比特成本每年下降的百分比？第二個問題是──來自中國的產量佔比是多少？而您認為中國的實際需求佔您收入的百分比是多少？

I think the relative -- I can answer the China question first and then Gianluca can take the cost discussion. China is down quite a bit, but relatively as part of our portfolio, it's about the same relative numbers that it was even when the peak of our portfolio. So I wouldn't overly analyze China versus CSPs versus consumer spending and all these other things that have affected our revenue downturn.

我認為相對而言——我可以先回答中國的問題，然後詹盧卡可以討論成本問題。中國業務下滑幅度很大，但相對而言，作為我們投資組合的一部分，其相對數字與我們投資組合巔峰時期的水平大致相同。因此，我不會過度分析中國、電信業者、消費者支出以及其他所有影響我們收入下滑的因素。

I think it's a fairly consistent number across the entire portfolio. And we're looking for demand upswings in every one of those sections as well as we go forward, but we're also managing the business really carefully because we don't think it's going to come back in the next 6 months, like we said in our prepared remarks.

我認為整個投資組合的表現表現相當穩定。我們預期未來每個部門的需求都會上升，但我們也非常謹慎地管理業務，因為正如我們在準備好的發言稿中所說，我們認為未來6個月內需求不會回升。

The cost per terabyte, of course, we don't guide the full year. What I would say is in a down cycle, it depends a lot from the level of utilization that you have in your factories, so can vary a lot from one quarter to another quarter. In general, no, we are continuing to do our product road map. And as Dave said, we have a new PMR product coming out soon, higher capacity, that means lower cost per terabyte. And then we will have HAMR in just a couple of quarters. So all those new products will help us in reducing the cost per terabyte.

當然，我們不會提供全年的每TB成本預測。我想說的是，在經濟下行週期，這很大程度上取決於工廠的利用率，因此每季之間的差異可能很大。總的來說，我們會繼續推進產品路線圖。正如Dave所說，我們即將推出一款新的PMR產品，容量更大，這意味著每TB成本更低。之後，我們將在幾個季度內推出HAMR產品。所有這些新產品都將幫助我們降低每TB成本。

I like our chances in the cost per terabyte arena going forward. And I think as we continue to work yields and scrap on these new products and so on, have the highest capacity points out there, have the ability to take that areal density to lower capacity points as well, it helps our cost per terabyte immensely.

我看好我們未來在每TB成本上的機會。而且我認為，隨著我們繼續提升這些新產品的良率和報廢率，擁有最高的容量點，並有能力將面密度降低到更低的容量點，這將極大地幫助我們降低每TB成本。

The next question comes from Mehdi Hosseini with Susquehanna.

下一個問題來自 Susquehanna 的 Mehdi Hosseini。

Yes. A couple of follow-ups. Given your lower OpEx (inaudible) around to 160, how should I think about scaling, especially on the revenue side, before you would need to increase OpEx materially?

是的。還有幾個後續問題。考慮到你們的營運成本（聽不清楚）較低，大約160，在你們需要大幅增加營運成本之前，我該如何考慮擴大規模，尤其是在收入方面？

Yes. Mehdi, I think we can hold the line on where we are right now and get a lot of scale. I mean we've been through the product transition far enough on the PMR product and HAMR product that we can -- we're pretty much done with the development phases, all the experimentation and so on and so forth. So I don't think we need to grow OpEx until we build back quite a bit of revenue and so we'll get very efficient scale there. I appreciate the question.

是的。 Mehdi，我認為我們可以維持目前的水平，並實現大規模擴張。我的意思是，我們在PMR產品和HAMR產品上已經完成了足夠的產品轉型，我們基本上已經完成了開發階段、所有實驗等等。所以我認為在我們恢復相當多的收入之前，我們不需要增加營運支出，這樣我們就能實現非常有效率的規模擴張。我很感謝你的提問。

And then for Gianluca, should I assume that free cash flow would be down in September and then up in December?

那麼對於 Gianluca 來說，我是否應該假設自由現金流會在 9 月下降，然後在 12 月上升？

Well, in general, no, we have generated good free cash flow through all our quarters. I said in the prepared remarks, September would be a little bit lower because we have to pay for the restructuring charges that we executed in end of June and during July. But no, we also said we expect every quarter to be positive.

嗯，總體來說，我們所有季度都產生了良好的自由現金流。我在準備好的演講稿中提到，9月的業績會略低一些，因為我們必須支付6月底和7月份執行的重組費用。不過，我們也說過，我們預期每季的業績都會是正值。

(inaudible) negative, just sequentially, down in September and then rebounding in December on a sequential basis.

（聽不清楚）負面，只是連續下降，9 月下降，12 月連續反彈。

Better. Yes, better point. On the OpEx, Mehdi, I want to -- I just want to clarify, we don't have variable compensation right now in our numbers. So we could have a little bit of increase in the future if the business comes back stronger, just a little bit, not much.

更好。是的，說得更好。關於營運支出，Mehdi，我想澄清一下，我們目前的薪資數據中沒有浮動薪酬。所以，如果業務恢復得更好，未來我們的薪酬可能會略有增加，只是一點點，不會太多。

The next question comes from Mark Miller with the Benchmark Company.

下一個問題來自 Benchmark Company 的馬克米勒。

I'm just wondering if -- it sounds like the PMR is going further than I originally thought it would be. Is this a change? Or was that always in your plan?

我只是想知道——聽起來PMR的進展比我最初預想的要快得多。這是一個改變嗎？還是這一直就在你的計畫裡？

It was always in the plan. Yes. I think, Mark, we've kind of been focused on 16, 20, 24 and bringing on the technology that way. We believe we have great heads of media on PMR. It's getting towards the top end of the curve, right? So it's harder and harder to squeeze things out.

這一直在計劃中。是的。馬克，我想我們一直專注於16、20、24，並以此方式推進科技。我們相信我們在PMR領域擁有優秀的媒體負責人。它正在走向曲線的頂端，對吧？所以，擠出機會越來越難。

And we've worked really hard to make sure that the drive that was taking us there would also take us to HAMR. So we're leveraging as much as we possibly can and then only changing the areal density critical components that get us the areal density to get us into the 30s and building much more confidence on that new S curve actually as time goes on.

我們非常努力地確保，推動我們實現這一目標的驅動力也能幫助我們實現 HAMR。因此，我們盡可能地利用資源，只改變那些能夠將表面密度提升到 30 秒的關鍵元件，並隨著時間的推移，對新的 S 曲線建立更大的信心。

So I'm convinced there's going to be 40 terabyte drives in the world not too distant future. And we've talked about the capabilities of these things in the lab. We're ready to go 5 terabytes per disk. So teams incrementally more positive every time. So that's good. The teams are making progress.

所以我相信不久的將來世界就會出現40TB的硬碟。我們已經在實驗室裡討論過這些設備的性能了。我們準備將硬碟容量提升到5TB。所以團隊每次都越來越積極。這很好。團隊正在取得進展。

Is there any specific technology advance that's allowing you to push that now up to almost 30 terabytes?

是否有任何具體的技術進步使您可以將這一數字提升至近 30 TB？

Sorry, you mean beyond HAMR?

抱歉，您的意思是超越 HAMR 嗎？

No. I mean taking the PMR up to the high 20s in terms of terabytes, anything new in the technology and any breakthroughs or advancements?

不。我的意思是，如果將 PMR 提升到 20 多 TB 的話，這項技術有什麼新進展嗎？有什麼突破或進步嗎？

Well, we're always working new reader designs that are applicable to go to 30 and 40 terabyte drives as well. It's not only about the write technology, if you will. There's -- because PMR is implying the write technology -- that's write with a W-R-I-T-E, you know that. So it's not a HAMR just versus PMR. There's other technology things that are being worked -- technology vectors being worked.

嗯，我們一直在研究新的讀取器設計，這些設計也適用於 30 和 40 TB 的硬碟。這不僅僅關乎寫入技術。 PMR 指的是寫入技術，也就是用 W-R-I-T-E 進行寫入，你知道的。所以這不僅僅是 HAMR 與 PMR 的對比。還有其他技術正在研發中——各種技術載體也在研發中。

But mid-20s is CMR, high 20s has been SMR, and there's many different kinds of SMR variants for different customers. So I think you have to be a little bit careful with picking individual capacity points. And by the way, HAMR can use SMR as well on top of it. So when we get into the 30s, there might be that same kind of spread of capacity points because of how people are choosing to deploy HAMR in their data centers.

但是20多兆瓦時是CMR，20多兆瓦時是SMR，而且針對不同的客戶，SMR有很多不同的版本。所以我認為在選擇單一容量點時必須謹慎一些。順便說一句，HAMR也可以在其基礎上使用SMR。所以，當容量點達到30多兆瓦時，由於人們選擇在資料中心部署HAMR的方式不同，容量點的分佈可能也大致相同。

The last question today comes from Vijay Rakesh with Mizuho.

今天的最後一個問題來自瑞穗的 Vijay Rakesh。

Just a quick 2 questions. Just when you look at HAMR, obviously ramping well, any way of looking at what the exabyte mix could be as you look at calendar '24 or even calendar '25?

只想問兩個問題。 HAMR 顯然發展勢頭良好，您能預測一下 2024 年甚至 2025 年 EB 的佔比嗎？

Yes. I do think that as we -- and this is a delicate balance between the capacity we'll still have to make drives, the lead times associated with those big drives and the economic returns that we're going to be getting, so how many do we start now -- how many parts do we start now to be able to answer that. But theoretically, with HAMR allowing us to jump quite a bit in exabytes per box, if the demand is strong enough, then the exabyte growth will be strong, very strong.

是的。我確實認為，我們需要在硬碟的產能、這些大容量硬碟的交付週期以及我們將獲得的經濟回報之間取得微妙的平衡，那麼我們現在應該開始生產多少個零件才能回答這個問題呢？理論上，由於 HAMR 技術讓我們可以大幅提升每台硬碟的 EB 容量，如果需求夠強勁，那麼 EB 容量的成長將會非常強勁。

So a lot of this -- first order driver is still demand. We're going to answer that call with the components that we have. We're very much leaning into the product transition on HAMR because we have that much confidence. And I think the customer's qualification is going well and the customer interest is high because they see such a market TCO proposition as well. And if all that stuff lines up into good box demand, like somebody asked earlier, that means good exabyte growth. That's what we're all hoping for.

所以，其中很大一部分——首要驅動因素仍然是需求。我們將利用現有的組件來滿足這項需求。我們非常傾向於向HAMR進行產品轉型，因為我們對此充滿信心。我認為客戶的資質認證進展順利，客戶的興趣也很高，因為他們也看到了這樣的市場TCO（整體擁有成本）方案。如果所有這些因素都能形成良好的盒式需求，就像之前有人問到的那樣，那就意味著EB級容量的良好成長。這正是我們所有人所期待的。

Got it. And then on the data center inventory side, any way to kind of gauge how much -- how many weeks of inventory is still there? And how that compares to the prior sort of, I guess, in data center inventory or...

明白了。那麼，關於資料中心庫存方面，有什麼方法可以估算一下庫存還有多少——可以維持幾週？與之前的資料中心庫存相比怎麼樣…

Right. It's not an exact science. I would say we have evidence that it's going down, not only that's what people are telling us, our customers are telling us, but also we're not putting any in. So definitely, the Seagate inventory is going down. We're even managing our owned inventory down as well. I can't speak for the rest of the industry, but all indications are that things are getting healthier and -- but we don't think that it's going to be fixed overnight. We think it's going to take another couple of quarters before the upturn.

是的。這不是一門精確的科學。我想說，我們有證據表明庫存正在下降，不僅是人們告訴我們的，我們的客戶也告訴我們的，而且我們也沒有增加任何庫存。所以，希捷的庫存肯定在下降。我們甚至也在減少自有庫存。我不能代表整個行業，但所有跡像都表明情況正在好轉——但我們不認為這種情況會在一夜之間得到解決。我們認為還需要幾個季度才能好轉。

And that's something that I'm very cognizant of. We didn't project originally going into this downturn that this would be this long, but -- yes, and I don't think it's going to last forever either. I think it will turn north again here. We just basically said not until the end of the calendar year.

這是我非常清楚的。我們最初在陷入這場經濟衰退時並沒有預料到它會持續這麼久，但是——是的，而且我認為它也不會永遠持續下去。我認為它會再次轉向。我們基本上說要到今年年底才會出現這種情況。

This concludes our question-and-answer session. I would like to turn the conference back over to management for any closing remarks.

我們的問答環節到此結束。我想將會議交還給管理階層，請大家發表最後發言。

Thanks, Betsy. As you heard today, Seagate is continuing to take proactive actions to navigate this tough business environment over the near term. However, we remain excited as ever for the long-term opportunities for mass capacity storage, and it's brought about by existing and emerging trends that are underpinned by data like AI, like we talked about today.

謝謝，貝琪。正如您今天所聽到的，希捷正在繼續採取積極行動，以應對短期內嚴峻的商業環境。然而，我們仍然對大容量儲存的長期機會感到興奮，而這種機會是由現有和新興趨勢帶來的，這些趨勢以人工智慧等數據為基礎，就像我們今天所討論的那樣。

Our leading technology road map positions us really well to address future demand and enhance value for both our customers and for Seagate's financial performance. I'll close by thanking all of our stakeholders for their ongoing support. Thanks for joining us today.

我們領先的技術路線圖使我們能夠更好地滿足未來的需求，並為我們的客戶和希捷的財務表現提升價值。最後，我要感謝所有利害關係人一直以來的支持。感謝您今天的參與。

The conference has now concluded. Thank you for attending today's presentation. You may now disconnect.

會議現已結束。感謝您參加今天的演講。您可以斷開連線了。