惠普企業 (HPE) 2023 Q4 法說會逐字稿

Good afternoon, and welcome to the Fourth Quarter 2023 Hewlett Packard Enterprise Earnings Conference Call. My name is Gary, and I'll be your conference moderator for today's call. (Operator Instructions). As a reminder, this conference is being recorded for replay purposes.

下午好，歡迎參加 2023 年第四季惠普企業財報電話會議。我叫加里，我將擔任今天電話會議的主持人。 （操作員指令）。提醒一下，本次會議正在錄音以供重播。

I would now like to turn the presentation over to your host for today's call, Jeff Kvaal, Senior Director, Investor Relations. Please proceed.

現在，我想將演講交給今天電話會議的主持人、投資者關係高級總監 Jeff Kvaal。請繼續。

Thanks, Gary, and good afternoon, everyone. I'm Jeff Kvaal. I'd like to welcome you to our fiscal 2023 4th quarter earnings conference call with Antonio Neri, HPE's President and Chief Executive Officer; and Jeremy Cox, HPE's Senior Vice President, Controller and Interim Chief Financial Officer.

謝謝，加里，大家下午好。我是傑夫·克瓦爾。歡迎您參加我們 2023 財年第四季財報電話會議，與 HPE 總裁兼執行長 Antonio Neri 一起；以及HPE資深副總裁、財務長兼臨時財務長Jeremy Cox。

Before handing the call to Antonio, let me remind you that this call is being webcast. A replay of the webcast will be available shortly after the call concludes. We have posted the press release and the slide presentation accompanying the release on our HPE Investor Relations web page.

在將電話交給安東尼奧之前，請允許我提醒您，該通話正在進行網路直播。通話結束後不久將提供網路直播的重播。我們已在 HPE 投資者關係網頁上發布了新聞稿和隨附的幻燈片簡報。

Elements of the financial information referenced on this call are forward-looking and are based on our best view of the world and our businesses as we see them today. HPE assumes no obligation and does not intend to update any such forward-looking statements. We also note that the financial information discussed on the call reflects estimates based on the information available at this time and could differ materially from the amounts ultimately reported in HPE's annual Form 10-K for the fiscal year ended October 31, 2023. For more detailed information, please see the disclaimers on the earnings materials relating to forward-looking statements that involve risks, uncertainties and assumptions. Please refer to HPE's filings with the SEC for a discussion of these risks.

本次電話會議中引用的財務資訊部分是前瞻性的，基於我們對當今世界和我們業務的最佳看法。 HPE 不承擔任何義務，也不打算更新任何此類前瞻性聲明。我們還注意到，電話會議上討論的財務信息反映了基於​​目前可用信息的估計，可能與 HPE 截至 2023 年 10 月 31 日的財年的年度 10-K 表中最終報告的金額存在重大差異。請參閱 HPE 向美國證券交易委員會提交的文件，以了解有關這些風險的討論。

For financial information we have expressed on a non-GAAP basis, we have provided a reconciliation to the comparable GAAP information on our website. Please refer to the tables and slide presentation accompanying today's earnings release on our website for details. Throughout this conference call, all revenue growth rates, unless noted otherwise, are presented on a year-over-year basis and adjusted to exclude the impact of currency. Finally, Antonio and Jeremy will reference our earnings presentation in their prepared comments.

對於我們以非 GAAP 基礎表達的財務訊息，我們在網站上提供了與可比較 GAAP 資訊的對帳。有關詳細信息，請參閱我們網站上今天的收益報告中附帶的表格和幻燈片演示。在整個電話會議中，除非另有說明，所有收入增長率均按同比計算，並經過調整以排除貨幣的影響。最後，Antonio 和 Jeremy 將在他們準備好的評論中參考我們的收益報告。

And with that, let me turn it to you, Antonio.

現在，讓我把話題轉到你身上，安東尼奧。

Thanks, Jeff, and good afternoon and thank you for joining us today.

謝謝，傑夫，下午好，謝謝您今天加入我們。

Pardon me, this is the conference operator. The speaker's line appears to have dropped. Please remain on the line while we rejoin them.

對不起，我是會議接線生。發言人的電話線似乎斷了。我們將重新加入他們，請繼續保持通話。

Gary, can you hear us?

加里，你聽得到我們說話嗎？

Yes, sir, we can hear you now. Thank you.

是的，先生，我們現在能聽到您的聲音。謝謝。

All right. Thank you. Well, thanks, Jeff, and good afternoon and thank you for joining us today. In fiscal year 2023, HPE delivered record performance against non-GAAP financial metrics by capitalizing on strong momentum across our portfolio. Our steady execution has resulted in higher revenue, further gross margin expansion, larger operating profit and record-breaking non-GAAP diluted net earnings per share and free cash flow.

好的。謝謝。好吧，謝謝傑夫，下午好，謝謝您今天加入我們。 2023 財年，HPE 利用整個產品組合的強勁發展勢頭，在非 GAAP 財務指標方面取得了創紀錄的業績。我們的穩健執行帶來了更高的收入、進一步的毛利率擴大、更大的營業利潤和破紀錄的非公認會計準則每股攤薄淨收益和自由現金流。

Our growth engines in Intelligent Edge and HPC & AI as well as our HPE GreenLake platform are helping to accelerate our revenue and profit diversification. Importantly, HPE has achieved demonstrable success this year in our ongoing portfolio pivot to higher-growth, higher-margin areas aligned to the key market mega trends driving customer demand. I will focus my commentary today on our full year fiscal year 2023 results and allow Jeremy to expand on the fourth quarter and segment results.

我們的智慧邊緣、HPC 和 AI 以及 HPE GreenLake 平台的成長引擎正在幫助加速我們的收入和利潤多元化。重要的是，HPE 今年在持續調整產品組合至更高成長、更高利潤的領域方面取得了顯著的成功，這與推動客戶需求的關鍵市場大趨勢相一致。今天，我將重點放在我們 2023 財年全年業績，並讓 Jeremy 詳細介紹第四季和分部的業績。

I'm very proud of all that HPE delivered in fiscal year 2023 for our shareholders, our customers and our team members. Our performance demonstrates the relevance of our strategy, our portfolio differentiation and our strong execution. We delivered extraordinary innovation to customers, resulting in share gains in key markets and profitable growth for our company.

我為 HPE 在 2023 財年為我們的股東、客戶和團隊成員所取得的成績感到非常自豪。我們的業績證明了我們策略的相關性、我們投資組合的差異化以及我們強大的執行力。我們為客戶提供了非凡的創新，從而增加了關鍵市場的份額並實現了公司的獲利成長。

HPE generated our highest gross margin and our highest operating profit since I became CEO and our highest annual revenue in 4 years. Non-GAAP diluted net earnings per share and free cash flow were the largest ever in our company's history. We saw healthy sustained growth in revenue at $29.1 billion for the full year, an increase of 5.5% year-over-year in constant currency, our third straight year of revenue growth, [above the midpoint] of our full fiscal year 2023 guidance of 5%. Gross margins exceeded 35%. We also added substantially to our annualized revenue run rate, closing fiscal year 2023 with more than $1.3 billion in ARR. That represents a nearly $370 million increase from this time last year. Non-GAAP operating profit margin was up 20 basis points year-over-year to end the year at 10.8%. We have executed well on our addressable market opportunities and exercised prudence with our expenses and obligations.

HPE創造了我擔任執行長以來最高的毛利率和最高的營業利潤，並且創造了4年來最高的年收入。非公認會計準則攤薄每股淨收益和自由現金流是我們公司歷史上最高的。我們全年營收持續健康成長至 291 億美元，以固定匯率計算年增 5.5%，這是我們連續第三年實現營收成長，[高於] 2023 財年全年預測 5% 的中點。毛利率超過35%。我們也大幅提高了年化收入運行率，截止 2023 財年，年化收入運行率已超過 13 億美元。這比去年同期增加了近 3.7 億美元。非公認會計準則營業利益率年增 20 個基點，年底達到 10.8%。我們很好地把握了可獲得的市場機會，並審慎處理我們的開支和義務。

Non-GAAP EPS increased 6.4% year-over-year to $2.15. Improved profitability also means that we have significantly boosted free cash flow, which rose by about $400 million this year to $2.2 billion, a nearly 25% year-over-year increase. Both non-GAAP EPS and free cash flow results are record-breaking and well above our fiscal year 2022 SAM guidance.

非公認會計準則每股收益年增 6.4% 至 2.15 美元。獲利能力的提高也意味著我們的自由現金流大幅增加，今年增加了約 4 億美元，達到 22 億美元，較去年同期成長近 25%。非 GAAP EPS 和自由現金流結果都創下紀錄，遠高於我們 2022 財年的 SAM 指引。

In summary, HPE delivered an impressive set of results in fiscal year 2023. With the progress we have made this year and our confidence in generating further value for our shareholders, we are raising our dividend in 2024.

總而言之，HPE 在 2023 財年取得了令人印象深刻的業績。

HPE is more relevant than ever to our customers. We have deliberately aligned our strategy over the last few years to significant trends in the market around edge, hybrid cloud and AI. These growth engines align to our customers' interest and where they are targeting their IT spend. Even against an uncertain macroeconomic backdrop, we saw continued though uneven demand across our HPE portfolio with a significant acceleration in AI orders.

HPE 與我們的客戶比以往任何時候都更加密切相關。過去幾年來，我們刻意調整了我們的策略，使其與邊緣、混合雲和人工智慧等市場重大趨勢一致。這些成長引擎與我們客戶的興趣以及他們的 IT 支出目標一致。即使在不確定的宏觀經濟背景下，我們仍然看到 HPE 產品組合的需求持續但不均衡，AI 訂單顯著加速。

Demand in our AI solutions is exploding. We saw a significant uptick in customer demand in recent quarters for accelerated computing infrastructure and services. In Q4, orders for servers that include accelerated processing units, or APUs, represented 32% of our total server order mix, up more than 250% from the beginning of fiscal year 2023. APUs, which includes GPU-based servers orders across our business, represented 25% of our total server order mix in fiscal year 2023.

我們對人工智慧解決方案的需求正在激增。最近幾個季度，我們看到客戶對加速運算基礎設施和服務的需求大幅上升。在第四季度，包含加速處理單元 (APU) 的伺服器訂單占我們總伺服器訂單組合的 32%，比 2023 財年初增長了 250% 以上。

Our HPC & AI segment revenue grew 25% year-over-year in fiscal year 2023. We ended this fiscal year with the largest HPC & AI order book on record driven by $3.6 billion in company-wide APU orders. This tops what has been a historically large order book in the third quarter as demand accelerated for our supercomputing and AI solutions. We anticipate demand next fiscal year will remain very strong. We likely have a large order backlog until GPU supply is less constrained.

2023 財年，我們的 HPC 和 AI 部門營收年增 25%。由於對我們的超級運算和人工智慧解決方案的需求加速，這創下了第三季的歷史訂單量紀錄。我們預計下一財年的需求將保持強勁。在 GPU 供應不再那麼緊張之前，我們可能會有大量訂單積壓。

Two weeks ago, at the Supercomputing conference, we expanded our collaboration with NVIDIA to announce a turnkey preconfigured supercomputing solution for generative AI to streamline the model development process. The new solution speeds up the training and fine-tuning of AI models using private data sets. Designed for large enterprises, research institution and government organizations, it comprises HPE AI software, our industry-leading supercomputing and storage solutions, our HPE Slingshot Interconnect fabric and HPE services and the quad NVIDIA Grace Hopper GH200 Superchip. Later this week, at the HPE Discover Barcelona, we will further expand our NVIDIA partnership with new solutions created for enterprise customers.

在兩週前的超級運算大會上，我們擴大了與 NVIDIA 的合作，宣布推出用於生成式 AI 的交鑰匙預先配置超級運算解決方案，以簡化模型開發流程。新的解決方案利用私有資料集加速人工智慧模型的訓練和微調。它專為大型企業、研究機構和政府組織設計，包括 HPE AI 軟體、我們業界領先的超級運算和儲存解決方案、我們的 HPE Slingshot Interconnect 結構和 HPE 服務以及四核心 NVIDIA Grace Hopper GH200 超級晶片。本週晚些時候，在巴塞隆納 HPE Discover 大會上，我們將進一步擴大與 NVIDIA 的合作夥伴關係，為企業客戶打造新的解決方案。

Also this month, the University of Bristol announced that HPE has been selected to deliver the U.K.'s fastest supercomputer, thanks to our U.K. government investment of GBP 225 million intended to make the nation a world leader in AI. The Japan National Institute of Information and Communications Technology recently turned to HPE Cray XD supercomputers to develop an AI-based multilingual communication tool to process, translate and interpret texts and images for 17 languages.

本月，布里斯託大學還宣布，HPE 已被選中提供英國最快的超級計算機，這要歸功於英國政府投資 2.25 億英鎊，旨在使英國成為人工智慧領域的世界領先者。日本國家資訊通信技術研究所最近利用 HPE Cray XD 超級電腦開發基於 AI 的多語言通訊工具，以處理、翻譯和解釋 17 種語言的文字和圖像。

And we just announced a partnership with DarkByte, a provider of next-generation high-performance computing green data centers, to power its AI cloud service with HPE Cray supercomputers and our purpose-built machine learning software suite. With HPE-built supercomputers consistently ranked among the top 10 within the top 500 most powerful and sustainable supercomputers, our clear leadership in this space continues to position us well in the AI market.

我們剛剛宣布與下一代高效能運算綠色資料中心供應商 DarkByte 建立合作夥伴關係，利用 HPE Cray 超級電腦和我們專門構建的機器學習軟體套件為其 AI 雲端服務提供支援。 HPE 製造的超級電腦在 500 台最強大和最具可持續性的超級電腦中始終名列前十，我們在該領域的明顯領導地位使我們在 AI 市場上繼續佔據有利地位。

Our Intelligent Edge segment was the largest driver of our revenue and profit growth in fiscal year 2023, making up 18% of our overall revenue and 39% of the segment operating profit. Fiscal year 2023 revenue in this segment increased by 45% to $5.2 billion, while operating margins expanded more than 1,200 basis points year-over-year to 27.3%, demonstrating the relevance of our offering and the payoff of our investments over time. A critical part of our Intelligent Edge portfolio in the edge supply portfolio is to continue -- will continue to contribute meaningfully to profitable growth for our shareholders.

我們的智慧邊緣部門是我們 2023 財年營收和利潤成長的最大驅動力，占我們整體營收的 18% 和部門營業利潤的 39%。 2023 財年該部門的營收成長 45%，達到 52 億美元，而營業利潤率年增超過 1,200 個基點，達到 27.3%，證明了我們產品的相關性以及我們投資的長期回報。我們邊緣供應產品組合中智慧邊緣產品組合的關鍵部分是繼續為我們的股東的獲利成長做出有意義的貢獻。

The Compute segment is going through a cyclical period where customers are consuming prior investments, making this a price-competitive market for now. As we prepared to capture a greater share of AI-linked inferencing opportunities, we saw overall demand improved moderately in the second half and are encouraged in our outlook for this segment. Customer interest in servers with APUs is growing. We are investing in specialized sales resources that can enhance future growth. We also know we must keep our focus on capturing every unit in this business while keeping balancing our operating margin performance.

計算領域正在經歷一個週期性時期，客戶正在消耗先前的投資，這使得該市場目前處於價格競爭激烈的狀態。當我們準備好抓住更大份額的人工智慧相關推理機會時，我們看到下半年整體需求適度改善，我們對這一領域的前景充滿信心。客戶對搭載 APU 的伺服器的興趣日益濃厚。我們正在投資可以促進未來成長的專業銷售資源。我們也知道，我們必須集中精力佔領該業務的每個部門，同時保持營業利潤率表現的平衡。

The overall storage market has been sluggish this year, and on our -- even our uneven performance in this segment is in line with most of our peers. However, we are encouraged with 3 quarters of stable demand. We saw sequential improvement in Storage revenue in the fourth quarter. We continue to invest in our sales execution capabilities. We recently deployed a large specialized storage sales force, including a team devoted to growing our storage IP product mix. We expect our subscription-based offerings and differentiated IP products like HPE Alletra will continue to be sources of strong growth to enhance the profitability of this business in the year ahead.

今年整個儲存市場一直表現低迷，即使我們在這一領域的表現參差不齊，也與大多數同行一致。然而，我們對 3 個季度的穩定需求感到鼓舞。我們看到第四季儲存收入連續改善。我們將繼續投資我們的銷售執行能力。我們最近部署了一支大型專業儲存銷售隊伍，其中包括一支致力於擴大我們的儲存 IP 產品組合的團隊。我們預計，我們的訂閱制的產品和 HPE Alletra 等差異化 IP 產品將繼續成為強勁成長的源泉，從而在未來一年提升該業務的獲利能力。

We remain focused on advancing our position in hybrid cloud. We ended the year with 29,000 customers on our HPE GreenLake cloud platform. Customers require a hybrid by design IT estate. They are attracted to our cloud platform because of its experience, flexibility and cost. In the fourth quarter, we closed our largest HPE GreenLake for Private Cloud Enterprise deal to date. In addition, we saw more customer demand around our HPE GreenLake SaaS offerings across data protection, observability and sustainability services.

我們將繼續致力於提升我們在混合雲領域的地位。截至年底，我們的 HPE GreenLake 雲端平台已擁有 29,000 名客戶。客戶需要混合設計的 IT 資產。他們因為我們的雲端平台的體驗、靈活性和成本而被我們的雲端平台所吸引。第四季度，我們完成了迄今為止最大的HPE GreenLake私有雲企業交易。此外，我們發現客戶對 HPE GreenLake SaaS 產品在資料保護、可觀察性和永續性服務方面的需求越來越大。

Finally, our HPE Financial Services segment continues to deliver strategic sustainable solutions for customers, accelerating our strategy and helping to expand earnings for our shareholders. Financing volumes rose year-over-year with a major contribution from efforts to boost our as-a-service volumes through HPE GreenLake.

最後，我們的 HPE 金融服務部門持續為客戶提供策略性永續解決方案，加速我們的策略並協助擴大股東收益。融資量年增，主要得益於我們透過 HPE GreenLake 提升即服務量的努力。

Fiscal year 2023 was an important year for HPE, one we advanced our strategy and delivered record financial performance for our shareholders through focused execution and operating discipline. I'm confident in our ability to continue to deliver for our shareholders in fiscal year 2024 and beyond for 3 main reasons. First, the work we have done over the last several years to innovate and invest in the right places, places where we have the expertise and the ability to scale, has given us a portfolio that I believe stands apart from any other.

2023 財年對 HPE 來說是重要的一年，在這一年中，我們推進了策略，並透過專注的執行和營運紀律為股東帶來了創紀錄的財務表現。我相信我們有能力在 2024 財年及以後繼續為股東帶來回報，主要原因有三。首先，過去幾年來，我們在正確的地方（我們擁有專業知識和擴展能力的地方）進行創新和投資，這些工作為我們帶來了一個我認為與眾不同的投資組合。

Second we have weathered cyclical dynamics well. While others have had much more severe shifts to make, we have been able to stay the course in bringing our strategy to life because we have made operational improvements and have managed expenses in a disciplined way.

第二，我們很好地抵禦了周期性波動。當其他人必須做出更嚴峻的轉變時，我們卻能夠堅持實踐我們的策略，因為我們已經進行了營運改進並以嚴謹的方式管理費用。

And finally, we have been bold in undertaking transformation across the company. The changes we have made to our operating model to strengthen capabilities in our growth segment, focus on our road maps across the portfolio, create new customer experience and reach them in new ways will each pay off next year and beyond.

最後，我們大膽地在整個公司範圍內進行轉型。我們對營運模式所做的改變是為了增強我們成長部門的能力，專注於我們整個產品組合的路線圖，創造新的客戶體驗並以新的方式接觸他們，這些改變將在明年及以後帶來回報。

We are set up very well to navigate the current climate with and for our customers and to add significantly to the long-term profitability and value we create for our shareholders. While headwinds remain in certain segments of the IT market, HPE is positioned very well to continue our momentum, and we are laser focused in areas where we know we can do more to improve our performance. I hope you share my optimism in what HPE can achieve in the year ahead.

我們已經做好了充分的準備，可以與客戶一起應對當前的環境，並大幅提高我們為股東創造的長期獲利能力和價值。儘管 IT 市場的某些領域仍然存在阻力，但 HPE 仍處於非常有利的位置來繼續保持我們的發展勢頭，並且我們專注於我們知道可以做得更多以提高我們的績效的領域。我希望您和我一樣對 HPE 在未來一年的成就充滿信心。

Let me now invite Jeremy to discuss the fourth quarter and specific segment results. So Jeremy, over to you.

現在，我請傑瑞米討論第四季和具體部門的業績。那麼傑里米，就交給你了。

Thank you very much, Antonio. We closed FY '23 with another solid performance. Our Q4 results reflect our strategic focus to diversify our business towards higher-growth, higher-margin areas of the market. The company's transformation we have undertaken several years ago to pivot to edge and cloud is paying off. And with AI exploding in 2023, we see several promising indicators as we look further in 2024 despite some unevenness in some areas of the IT market where we participate.

非常感謝，安東尼奧。我們以另一次穩健的業績結束了23財年。我們的第四季業績反映了我們的策略重點，即將業務多元化至更高成長、更高利潤的市場領域。幾年前，我們進行的公司轉型轉向邊緣和雲端，正在取得成效。隨著人工智慧在 2023 年的爆發式成長，儘管我們所參與的 IT 市場某些領域存在一些不平衡，但展望 2024 年，我們看到了一些有希望的指標。

Q4 performance was highlighted by healthy revenues and gross margins. Revenue grew 5% sequentially in constant currency to $7.4 billion, hitting the midpoint of our Q4 revenue guidance range. Year-over-year revenue, however, was down 6% on a difficult compare to Q4 '22, during which significant order book consumption boosted our results. Q4 non-GAAP gross margin was 34.8%, which was up 170 basis points year-over-year, largely due to the increasing contribution of our Intelligent Edge segment.

第四季的業績亮點是良好的收入和毛利率。以固定匯率計算，營收季增 5% 至 74 億美元，達到我們第四季營收預期的中點。然而，與 22 年第四季相比，年收入下降了 6%，這一數字較為困難，在此期間，大量的訂單消費推動了我們的業績。第四季非公認會計準則毛利率為 34.8%，較去年同期成長 170 個基點，這主要歸功於我們智慧邊緣部門的貢獻不斷增加。

Let me remind you of the deliberate targeted investment decisions we made in Q4 that I flagged at SAM, which were funded with the better-than-expected OI&E performance in the same period. The impact of this investment, along with the strong seasonal growth in HPC & AI, resulted in a Q4 non-GAAP operating margin of 9.7%. This is down 60 basis points sequentially and 180 basis points year-over-year. We expect non-GAAP operating margins to quickly recover in Q1 '24.

讓我提醒您我在 SAM 上提到的我們在第四季度做出的深思熟慮的有針對性的投資決策，這些決策是由同一時期好於預期的 OI&E 表現資助的。此項投資的影響，加上 HPC 和 AI 的強勁季節性成長，導致第四季度非 GAAP 營業利潤率達到 9.7%。這比上一季下降了 60 個基點，比去年同期下降了 180 個基點。我們預計非 GAAP 營業利潤率將在 24 年第一季快速恢復。

This led to GAAP diluted net EPS to $0.49 and non-GAAP diluted net EPS to $0.52, which was at the high end of our Q4 guidance range of $0.48 to $0.52. Our Q4 free cash flow of $2.3 billion was record-breaking for the company.

這導致 GAAP 稀釋淨每股收益達到 0.49 美元，非 GAAP 稀釋淨每股收益達到 0.52 美元，這處於我們第四季度指導範圍 0.48 美元至 0.52 美元的高位。我們第四季的自由現金流達到 23 億美元，創下了公司的最高紀錄。

HPE also delivered an impressive full year performance in FY '23. We delivered revenue growth of 5.5% in constant currency, which was at the upper end of our guidance range of 4% to 6%. Currency was a 330 basis point headwind for the year. Non-GAAP diluted net EPS of $2.15 was also at the high end of our prior guidance and well above our initial guidance from SAM 2022 of $1.96 to $2.04. FY '23 free cash flow of $2.2 billion was well above our guidance of $1.9 billion to $2.1 billion.

HPE 在 23 財年全年業績也令人印象深刻。以固定匯率計算，我們的營收成長了 5.5%，處於我們 4% 至 6% 的預期範圍的上限。今年貨幣匯率面臨 330 個基點的阻力。非 GAAP 稀釋淨每股收益為 2.15 美元，也處於我們先前指引的高位，並遠高於我們對 SAM 2022 的初始指引 1.96 美元至 2.04 美元。 23財年自由現金流為22億美元，遠高於我們預期的19億美元至21億美元。

Let's now turn to our segment results. As a reminder, all revenue growth rates on this slide are in constant currency, and I will discuss the segments in our prior structure. The company is now operating according to our new structure. And as communicated at SAM, we plan to file restated historical financials for the new segments well before we report our Q1 '24 results.

現在我們來看看各分部業績。提醒一下，本投影片上的所有收入成長率均以固定匯率計算，我將按照我們先前的結構討論各個部分。公司現在按照我們的新架構運作。正如 SAM 所傳達的那樣，我們計劃在報告 2024 年第一季業績之前為新部門提交重述的歷史財務報表。

In Intelligent Edge, we grew revenue 40% year-over-year in Q4. Demand in our core product was down sequentially on customer digestion but grew in our software-centric solutions such as our HPE Aruba Central cloud management software and in our new TAM opportunities such as our SASE security software suite. Our operating margin of 29.5% was up more than 1,600 basis points year-over-year. While very pleased with this performance, we continue to expect a mid-20% operating margin over time, as communicated at SAM 2023. And finally, we're making progress on our order book and expect to be back close to historical normal levels by midyear 2024.

在智慧邊緣方面，我們第四季的營收年增了 40%。由於客戶消化，我們核心產品的需求持續下降，但以軟體為中心的解決方案（例如我們的 HPE Aruba Central 雲端管理軟體）和新的 TAM 機會（例如我們的 SASE 安全軟體套件）的需求卻有所增長。我們的營業利益率為 29.5%，比去年同期上升了 1,600 個基點以上。雖然對這一表現感到非常滿意，但我們仍預計隨著時間的推移，營業利潤率將達到 20% 左右，正如 SAM 2023 所傳達的那樣。

In HPC & AI, Q4 revenue grew 38% year-over-year and 41% sequentially. Q4 revenue results included only a modest amount of AI revenue. On a sequential basis, the continued strength in AI demand lifted our order book in the HPC & AI segment to more than $3 billion despite our significant revenue growth in the quarter. We continue to expect double-digit revenue growth in the segment over the next 3 years, and we are increasingly confident we will be above trend in FY '24 with GPU supply our gating factor. Our Q4 operating margin was 4.7%, up 120 basis points year-over-year and 550 basis points sequentially. While we have much more progress to make, this does illustrate the positive benefits of scale. The early stage of the AI market, tightness in certain key accelerators and long lead times in this segment means that the HPC & AI margins will fluctuate.

在 HPC 和 AI 領域，第四季營收年增 38%，季增 41%。第四季的收入結果僅包含少量的人工智慧收入。與上一季相比，儘管我們本季的營收大幅成長，但人工智慧需求的持續強勁推動我們在 HPC 和人工智慧領域的訂單量增加至 30 億美元以上。我們繼續預計未來 3 年該部門的收入將實現兩位數增長，並且我們越來越有信心，在 GPU 供應成為我們的門控因素的情況下，我們將在 24 財年實現高於趨勢的收入。我們第四季的營業利潤率為 4.7%，較去年同期成長 120 個基點，較上季成長 550 個基點。雖然我們還有很長的路要走，但這確實說明了規模的正面效益。人工智慧市場處於早期階段，某些關鍵加速器的緊張狀態以及該領域的較長交貨時間意味著 HPC 和 AI 利潤率將會波動。

Storage revenues fell 12% year-over-year on a difficult compare but rose 3% sequentially. Storage demand has now been flat to up for 3 straight quarters. HPE Alletra revenue grew over 50% year-over-year, and it will remain a robust growth contributor in FY '24 supported by growth in file and object storage. HPE Alletra is shifting our mix within Storage to higher-margin software subscription revenue, which is a key driver of our ARR growth. Q4 Storage operating margin of 8.1% was down 730 basis points year-over-year. Headwinds included the deferred revenue impact of the HPE Alletra subscription software, accelerated investment outpacing year-over-year revenue performance and a high mix of third-party products this quarter. Our investments in product portfolio give us confidence Storage will make progress through FY '24 toward our long-term operating profit margin target communicated at SAM 2023 of mid-teens.

與去年同期相比，儲存收入下降了 12%，但比上一季成長了 3%。儲存需求已連續三個季度持平或上升。 HPE Alletra 營收年增超過 50%，在文件和物件儲存成長的推動下，它仍將是 24 財年強勁的成長貢獻者。 HPE Alletra 正在將我們的儲存產品組合轉向利潤率更高的軟體訂閱收入，這是我們 ARR 成長的關鍵驅動力。第四季儲存營業利潤率為 8.1%，年減 730 個基點。不利因素包括 HPE Alletra 訂閱軟體的遞延收入影響、加速投資超過同比收入表現以及本季度第三方產品的高度混合。我們對產品組合的投資使我們有信心，Storage 將在 24 財年取得進展，朝著我們在 SAM 2023 中期提出的長期營業利潤率目標邁進。

Compute revenue was down 30% year-over-year on a difficult compare and down 1% sequentially. Deal elongation and customer digestion we discussed previously continued to be most prevalent in Compute. Declining AUPs from a record high in Q1 '23 was also a meaningful driver. However, 2 straight quarters of sequential improvement in demand lends further confidence to our FY '24 outlook. Our full year operating margin of 13.7% exceeded our target range of 11% to 13%, but the operating margin of 9.8% in Q4 was temporarily below the range. Given gross margins remain largely flat sequentially and the sequential demand increases, we continue to expect operating margins to be in the target range for FY '24.

相比之下，計算收入年減 30%，季減 1%。我們之前討論過的交易延長和客戶消化在計算領域仍然最為普遍。 AUP 從 23 年第一季的歷史高點下降也是一個重要的推動因素。然而，需求連續兩季持續改善進一步增強了我們對 24 財年前景的信心。我們全年的營業利潤率為 13.7%，超過了我們 11% 至 13% 的目標範圍，但第四季的營業利潤率為 9.8%，暫時低於該範圍。鑑於毛利率較上季基本持平且需求較上季增加，我們繼續預期 24 財年的營業利潤率將在目標範圍內。

HPE Financial Services revenues were flat year-over-year, and financing volume was $1.5 billion. Our operating margin of 8.9% was down 220 basis points year-over-year, reflecting rapid interest hikes that we are offsetting through pricing as well as asset management margins returning to normal as supply challenges ease. Our Q4 loss ratio remained steady at 0.5%.

HPE金融服務收入與去年同期持平，融資額為15億美元。我們的營業利益率為 8.9%，較上年同期下降 220 個基點，反映出利率快速上漲的影響，我們透過定價來抵銷這一影響，同時隨著供應挑戰的緩解，資產管理利潤率也恢復正常。我們第四季的損失率保持穩定在 0.5%。

Let's double-click on our portfolio pivot to higher-growth, higher-margin, recurring revenue and in particular, to Intelligent Edge. Even 3 years ago, the Intelligent Edge segment constituted just 10% of segment revenue. This year, it represented 18%. The operating profit trajectory is even more telling. The Intelligent Edge segment contributed approximately 14% of the segment's operating profit 3 years ago and was nearly 40% in FY '23.

讓我們雙擊我們的投資組合，轉向更高成長、更高利潤、經常性收入，特別是智慧邊緣。即使在三年前，智慧邊緣部門也僅佔部門收入的 10%。今年這一比例上升至18%。營業利潤走勢則更引人注目。三年前，智慧邊緣部門貢獻了該部門營業利潤的約 14%，而在 23 財年則達到近 40%。

In FY '24, we intend to give you a similar view combining our growth pillars of HPC & AI, hybrid cloud and Intelligent Edge segments. Within that, we expect the Intelligent Edge segment growth to moderate and the HPC & AI segment growth to accelerate.

在24財年，我們打算結合HPC和AI、混合雲和智慧邊緣部分的成長支柱，為您提供類似的觀點。其中，我們預期智慧邊緣部分的成長將放緩，而 HPC 和 AI 部分的成長將加速。

Robust demand in our hybrid cloud and as-a-service offerings illustrates the durability of our portfolio shift. ARR exceeded $1.3 billion in Q4. This represented 37% year-over-year growth, which is in line with our long-term CAGR target of 35% to 45%. Storage and Intelligent Edge software and services are the fastest-growing components of ARR.

我們對混合雲和即服務產品的強勁需求表明我們的投資組合轉變具有持久性。第四季的 ARR 超過 13 億美元。這代表著年成長 37%，符合我們 35% 至 45% 的長期複合年增長率目標。儲存和智慧邊緣軟體和服務是 ARR 成長最快的組成部分。

We continue to lift HPE GreenLake's value proposition with an increasing mix of higher-margin, recurring software and services revenue. Our software and services mix was 68% in the quarter. We expect this mix to increase into the upper 70% range by FY '26. This expansion is driven by the growth of subscription-based software with our products and the increased attach of our HPE operational services, which rose double digits in Q4.

我們繼續透過增加高利潤、經常性軟體和服務收入的組合來提升 HPE GreenLake 的價值主張。本季我們的軟體和服務組合佔 68%。我們預計到26財年這一比例將升至70%以上。這一擴張是由我們產品的訂閱式軟體的成長以及我們的 HPE 營運服務附加價值的增加所推動的，後者在第四季度實現了兩位數的成長。

The rising software and services mix is expanding our as-a-service margins. Our as-a-service orders grew 11% year-over-year in Q4 and finished the year up a solid 23% after 68% growth in FY '22. Our cumulative as-a-service TCV has now increased to nearly $13 billion. As ARR is now on a clear path to meaningful scale, we will simplify our as-a-service disclosure from ARR as-a-service orders and TCV to only ARR in FY '24.

不斷增長的軟體和服務組合正在擴大我們的即服務利潤率。我們的即服務訂單在第四季年增 11%，繼 22 財年成長 68% 之後，全年成長 23%。我們累積的即服務 TCV 現已增加到近 130 億美元。由於 ARR 現已明確走上有意義的規模化道路，我們將在 24 財年簡化我們的即服務揭露，從 ARR 即服務訂單和 TCV 簡化為僅有 ARR。

The explosion in AI demand is driving robust growth in our APU orders. Total APU orders, which include APUs in our Compute, Cray EX and Cray XD businesses, totaled $3.6 billion in FY '23, which is up from the over $3 billion we disclosed at SAM 2023. HPE Cray XD APU orders accelerated in Q3 and in Q4, reaching $2.4 billion for the year. The impressive APU server demand is also evident in our total server demand. APUs represented 25% of total server order dollars in FY '23, up from 10% in the prior year.

人工智慧需求的激增推動了我們的APU訂單強勁成長。總 APU 訂單（包括我們 Compute、Cray EX 和 Cray XD 業務中的 APU）在 23 財年總計 36 億美元，高於我們在 SAM 2023 上披露的 30 多億美元。令人印象深刻的 APU 伺服器需求也反映在我們的總伺服器需求中。 23 財年，APU 佔伺服器訂單總額的 25%，高於去年的 10%。

We will continue to disclose APU orders. However, orders can be easily north of $100 million each. Therefore, orders may be lumpy. But the key point is that we are capturing AI demand now and are well positioned for coming demand across the entire AI life cycle from training to tuning to inferencing.

我們將繼續揭露APU訂單。然而，每份訂單的價值很容易就超過 1 億美元。因此，訂單可能會不穩定。但關鍵在於我們現在正在抓住人工智慧需求，並且已經為整個人工智慧生命週期（從訓練到調整到推理）中未來的需求做好了準備。

Our strategy is delivering top line growth and gross margin expansion. Despite the challenging macro, ongoing product digestion and currency headwinds, we still produced one of our highest quarterly revenue figures since 2018 while sustaining our expanded gross margin profile. Our Q4 non-GAAP gross margin rose 170 basis points year-over-year to 34.8%. And our full year non-GAAP gross margin rose 140 basis points to 35.3%. That's a more than 500 basis point improvement from FY '18.

我們的策略是實現營業額成長和毛利率擴大。儘管面臨宏觀經濟挑戰、持續的產品消化和貨幣逆風，我們仍然創造了自 2018 年以來最高的季度收入之一，同時保持了擴大的毛利率。我們第四季的非公認會計準則毛利率年增 170 個基點至 34.8%。我們全年非公認會計準則毛利率上漲 140 個基點，達到 35.3%。這比 2018 財年提高了 500 多個基點。

Moving to free cash flow. In Q4, we generated $2.8 billion in cash flow from operations and $2.3 billion in free cash flow. Our $2.2 billion in free cash flow in FY '23 was above the high end of our guidance and up meaningfully from $1.8 billion in FY '22, primarily on improved earnings and net income conversion.

轉向自由現金流。第四季度，我們產生了 28 億美元的營運現金流和 23 億美元的自由現金流。我們23財年的22億美元自由現金流高於我們預期的高位，並且比22財年的18億美元有顯著增長，這主要得益於盈利和淨收入轉換的提高。

We exited FY '23 with a cash conversion cycle of negative 4 days, which exceeded our expectation for neutral we communicated at SAM. Strong working capital management in Q4 and reduced transformation costs in FY '23 drove an improvement in our conversion of non-GAAP net earnings to free cash flow to approximately 80% in FY '23. Continued progress in FY '24 and beyond leads us to expect to reach 90% by FY '26.

我們在 23 財年結束時的現金轉換週期為負 4 天，這超出了我們在 SAM 溝通的中性預期。第四季強勁的營運資本管理和 23 財年轉型成本的降低推動我們將非 GAAP 淨收益轉化為自由現金流的比例在 23 財年提高至約 80%。 24 財年及以後的持續進步使我們預計到 26 財年將達到 90%。

We reiterate our FY '24 free cash flow guidance from SAM of $1.9 billion to $2.1 billion. And we returned over $1 billion in capital to shareholders in FY '23, including $421 million in share repurchases. While certain price and volume parameters of our trading program limited us to below our $500 million target in FY '23, we reiterate our goal of returning 65% to 75% of free cash flow to shareholders between FY '24 and FY '26.

我們重申 SAM 24 財年自由現金流預期，即 19 億美元至 21 億美元。我們在 23 財年向股東返還了超過 10 億美元的資本，其中包括 4.21 億美元的股票回購。雖然我們交易計畫的某些價格和數量參數限制我們在23財年的收入低於5億美元的目標，但我們重申在24財年至26財年期間向股東返還65%至75%的自由現金流的目標。

Before we dive into our outlook, let me remind you that we will exclude H3C earnings and gain on sale from our non-GAAP results in FY '24, as we noted at SAM. We continue to expect the process to conclude and the receipt of the cash proceeds in the first half of calendar 2024.

在我們深入探討我們的展望之前，讓我提醒您，正如我們在 SAM 所指出的那樣，我們將從 24 財年的非 GAAP 業績中排除 H3C 收益和銷售收益。我們繼續預計流程將在 2024 年上半年完成，並收到現金收益。

For Q1, we expect revenues in the range of $6.9 billion to $7.3 billion. This incorporates historical seasonality in our overall business, including in the HPC & AI segment after its strong Q4. We expect AI revenue acceleration to drive sequential growth in HPC & AI and total HPE revenue in Q2 '24. We expect GAAP diluted net EPS between $0.24 and $0.32 and non-GAAP diluted net EPS between $0.42 and $0.50.

對於第一季度，我們預計營收在 69 億美元至 73 億美元之間。這體現了我們整體業務的歷史季節性，包括 HPC 和 AI 部門在第四季度表現強勁之後的季節性。我們預計 AI 收入加速將推動 HPC 和 AI 以及 2024 年第二季 HPE 總營收的連續成長。我們預計 GAAP 稀釋淨每股收益在 0.24 美元至 0.32 美元之間，非 GAAP 稀釋淨每股收益在 0.42 美元至 0.50 美元之間。

We're reiterating our prior year 2024 guidance of 2% to 4% revenue growth in constant currency. We expect OI&E, which as noted excludes H3C going forward, to be a headwind of approximately $300 million and our non-GAAP structural tax rate to be 15%. Our full year GAAP diluted net EPS guidance of $1.81 to $2.01 is $0.02 lower than our prior view as a bit of transformation costs slipped into FY '24.

我們重申我們先前對 2024 年的預期，即以固定匯率計算收入成長 2% 至 4%。我們預計 OI&E（如上所述，不包括 H3C）未來將帶來約 3 億美元的逆風，而我們的非 GAAP 結構性稅率為 15%。我們對全年 GAAP 攤薄淨每股收益的預期為 1.81 美元至 2.01 美元，比我們之前的預期低 0.02 美元，因為部分轉型成本在 24 財年有所下滑。

We are reiterating our full year non-GAAP diluted net EPS guidance of $1.82 to $2.02. We also reiterate free cash flow guidance of between $1.9 billion and $2.1 billion. We continue to believe our FY '24 performance will be weighted to the second half of the year as GPU supply improves. We are increasing our dividend by 8% in FY '24 and intend to return 65% to 75% of free cash flow to shareholders this year.

我們重申全年非公認會計準則攤薄淨每股收益預期，即 1.82 美元至 2.02 美元。我們也重申19億美元至21億美元之間的自由現金流預期。我們仍然相信，隨著 GPU 供應的改善，我們 24 財年的表現將主要集中在下半年。我們將在24財年將股利提高8%，並計劃今年向股東返還65%至75%的自由現金流。

So to conclude, while there continues to be unevenness in some areas of the IT market, our investment in growing areas of our portfolio is paying off. We are confident in our ability to continue to capture the AI explosion in demand. AI also opens broader customer discussions about the benefits of AI inferencing at the edge and the need to be hybrid by design. Our HPE GreenLake cloud platform, our AI native portfolio and our services expertise are perfectly aligned to the needs of our customers that we believe will turn into profitable growth for our shareholders.

總而言之，雖然 IT 市場某些領域仍然存在不平衡，但我們對投資組合中不斷增長的領域的投資正在獲得回報。我們有信心能夠持續滿足人工智慧的爆炸性成長需求。人工智慧也引發了更廣泛的客戶討論，關於邊緣人工智慧推理的好處以及設計混合的必要性。我們的 HPE GreenLake 雲端平台、我們的 AI 原生產品組合和服務專業知識與客戶的需求完美契合，我們相信這將為我們的股東帶來獲利成長。

Now let's open it up for questions.

現在讓我們開始提問。

(Operator Instructions) The first question is from Mike Ng with Goldman Sachs.

（操作員指示）第一個問題來自高盛的 Mike Ng。

I just have one on the APU orders of $3.6 billion. I was wondering if you could talk a little bit about the mix between compute, supercomputing and Cray XD and if you could offer any color on the type of customers that are making this order. Is this your typical enterprise customer? Is it more of a Tier 2 or AI CSP? Any thoughts there would be great.

我剛剛收到一份價值 36 億美元的 APU 訂單。我想知道您是否可以談論計算、超級計算和 Cray XD 之間的混合，以及您是否可以提供有關下此訂單的客戶類型的信息。這是您典型的企業客戶嗎？它更像是 Tier 2 還是 AI CSP？任何想法都很好。

Well, thanks, Mike. This is Antonio. So pretty much all the orders are in the HPC & AI segment, the vast majority. We saw now in Q4 some uptick in demand in what I call the traditional compute. But what you have to think about it is AI is a life cycle, training, tuning and inferencing. And the products we talk here, whether it's HPE Cray XD or EX, really are on the training side and the tuning side.

好吧，謝謝，麥克。這是安東尼奧。因此幾乎所有訂單都在 HPC 和 AI 領域，佔絕大多數。我們現在看到第四季度傳統計算領域的需求上升。但你必須考慮到，人工智慧是一個生命週期，訓練、調整和推理。我們在此討論的產品，無論是 HPE Cray XD 或 EX，其實都是屬於訓練方面和調整方面的。

And so when I think about the type of customers, I think about the model builders, right? So these are unique customers that in the past generally we have talked about it. Think about companies like Recursion Pharmaceuticals, Crusoe Energy, obviously, large language models like Aleph Alpha, Taiga Data or Northern Data, GE Research and the like. These are big model builders, and they need a large amount of computational power.

因此，當我考慮客戶類型時，我會考慮模型建構者，對嗎？因此，這些都是獨特的客戶，我們過去一般都談論過這個問題。想想像 Recursion Pharmaceuticals、Crusoe Energy 這樣的公司，顯然還有像 Aleph Alpha、Taiga Data 或 Northern Data、GE Research 這樣的大型語言模型。這些都是大型模型建構者，他們需要大量的運算能力。

Now what we start seeing is an uptick in the tuning side with enterprises because generally, they don't tend to build models. They tend to leverage foundation models in the open source or some of these companies provide. And then they tune those models with their data, but they want to do it in a private, secure and obviously sustainable way. And that's why we have made announcements with NVIDIA, and you can see further announcements later in the week.

現在我們開始看到企業在調整方面的上升趨勢，因為一般來說，他們不傾向於建立模型。他們傾向於利用開源或其中一些公司提供的基礎模型。然後他們利用數據調整這些模型，但他們希望以一種私密、安全且可持續的方式來進行。這就是我們與 NVIDIA 共同發佈公告的原因，您可以在本週稍後看到更多公告。

And then AI inferencing, I call it, for using a sport analogy, the singles and the doubles, right? So these are maybe a server with 8 GPUs or accelerator of sorts where they start deploying these models. They have been trained and tuned into production and think about where their real-time processing data happen, where business transformation takes place or maybe doing some sort of POC or pretraining experiments.

然後我稱之為人工智慧推理，使用體育類比，單打和雙打，對嗎？因此，這些可能是具有 8 個 GPU 或某種加速器的伺服器，他們在那裡開始部署這些模型。他們已經接受了培訓並適應了生產，並考慮他們的即時處理資料發生在哪裡，業務轉型發生在哪裡，或者可能進行某種 POC 或預訓練實驗。

And so now we start seeing that increasing. But the vast majority of compute is still CPU centric. And we saw some uptick in the GPUs. But the vast majority of all the APUs that we talked about are in the traditional high density for training and tuning, and that's where our HPE Cray set of platforms plays a big role. And obviously, they also find its way to supercomputing at large scale that we have talked about from [TRL Capital] and in Aurora or LUMI and [Norsk] and the like. Anything you want to add?

現在我們開始看到這種現像在增加。但絕大多數計算仍以 CPU 為中心。我們看到 GPU 有所上升。但我們談到的所有 APU 中的絕大多數都具有傳統的高密度，用於訓練和調整，而這正是我們的 HPE Cray 平台組發揮重要作用的地方。顯然，他們也找到了大規模超級計算的方法，正如我們在 [TRL Capital]、Aurora 或 LUMI 和 [Norsk] 等公司討論過的那樣。還有什麼要加嗎？

The only thing I would just add to that is the supercomputing piece, specific to your question, is less than 10% in the total APU orders in FY '23. The other interesting point to add on to the journey towards inferencing that Antonio mentioned is within compute, although against a very small base, we did see non-Tier 1 customer orders around GPU or APUs for compute increased about 100% in the quarter. So again, against a small base but an interesting point to see the focus start moving towards that realm.

我唯一想補充的是，針對您的問題，超級運算部分在 23 財年的總 APU 訂單中所佔比例不到 10%。 Antonio 提到的推理之旅的另一個有趣的點是在計算領域，儘管基數很小，但我們確實看到本季度非一級客戶圍繞 GPU 或 APU 的計算訂單增長了約 100%。因此，再次強調，儘管基數較小，但有趣的是，我們看到焦點開始轉向那個領域。

Yes. Thanks, Jeremy. And again, only 10% of the APUs were consumed in supercompute. The rest was all in AI.

是的。謝謝，傑里米。再次，只有 10% 的 APU 用於超級運算。其餘的都與人工智慧有關。

Thanks, Mike. Gary, can we have the next question, please?

謝謝，麥克。加里，我們可以問下一個問題嗎？

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

下一個問題來自美國銀行的 Wamsi Mohan。

Antonio, we've seen some networking companies talk about a slowdown and some inventory digestion. You're still delivering very strong growth in edge, high 30s. I know you called out a more front-end loaded performance for edge. But curious how you're thinking about the rate and pace of that business trajectory both on revenue and margin terms as you go through the course of the year?

安東尼奧，我們看到一些網路公司談論經濟放緩和一些庫存消化。您仍然在邊緣實現非常強勁的成長，高達 30%。我知道您呼籲 Edge 提供更多的前端載入效能。但我很好奇，您是如何看待這一年中業務軌蹟的收入和利潤率變化速度和步伐的？

Yes. Thanks, Wamsi. I'm going to talk about demand, and then Jeremy can talk about revenue and margin, which we were very clear at, right, what to expect. Obviously, we have driven a significant growth over the last 2 years. I think it was the 12th consecutive quarter of year-over-year revenue growth. That's very impressive. We added $2 billion of revenue in the last 2 years. And obviously, that came with an expanded set of gross margins because of our software and subscription-based models that we have been driving.

是的。謝謝，Wamsi。我將談談需求，然後傑里米可以談談收入和利潤，我們對此非常清楚，對吧，可以期待什麼。顯然，我們在過去兩年裡實現了顯著的成長。我認為這是連續第 12 個季度實現同比增長。這真是令人印象深刻。我們在過去兩年中增加了 20 億美元的收入。顯然，由於我們一直推動的軟體和基於訂閱的模式，毛利率也隨之擴大。

Traditionally, we think about that has been in the campus and branch where -- in switching or Wi-Fi. But more and more lately, obviously, it's all software-driven and as well as expansion into the new TAMs we discussed at the Security Analyst Meeting, including SD-WAN and security to create the SSE framework or the SASE framework. We talked about it. And going forward, we are also going to add the private 5G, which we discussed. And let's not forget, we have been also gaining momentum in the data center with our offers, which are an extension of our switching portfolio in the data center.

傳統上，我們認為這是在校園和分支機構中進行的——在交換或 Wi-Fi 中。但最近，顯然一切都是由軟體驅動的，並且擴展到我們在安全分析師會議上討論的新 TAM，包括 SD-WAN 和安全性，以創建 SSE 框架或 SASE 框架。我們討論過這件事。展望未來，我們也將加入我們討論過的私有 5G。請別忘記，我們的產品也是我們在資料中心領域獲得發展動力的因素，這些產品是我們在資料中心的交換產品組合的延伸。

So it definitely was a quarter-over-quarter slight decline in orders for demand for products in the campus and branch, but we saw strong demand in the software and the security space. And that's why at the Security Analyst Meeting with Phil, we talked about this multiple adjacency we have added to the platform, and they are all incorporated into the HPE Aruba Central platform, which is part of HPE GreenLake.

因此，校園和分行產品需求的訂單肯定比上一季略有下降，但我們看到軟體和安全領域的需求強勁。這就是為什麼在與 Phil 的安全分析師會議上，我們討論了我們已添加到平台中的多個鄰接，它們都納入了 HPE Aruba Central 平台，該平台是 HPE GreenLake 的一部分。

So again, we are committed to grow revenues next year on the higher pace that we created, again, on the $5.2 billion we just delivered, albeit you should not expect a 40% growth, obviously. And that's why I want to have Jeremy talk about what we are reaffirming here in terms of revenue and as well as margins.

因此，我們再次致力於明年的營收成長，速度比我們剛剛實現的 52 億美元更快，儘管顯然你不應該期望成長 40%。這就是為什麼我想讓傑里米談談我們在收入和利潤方面重申的內容。

Right. Yes. We -- as Antonio mentioned, we did, at SAM, mentioned a slight growth we're expecting year-over-year on a full year basis for edge. As I think about that, I'd almost break that down into 2 halves. The first half, as we mentioned, we do still benefit from a backlog position or an order book position that will go into the first half and help support that revenue performance. And then the second half will be more dependent on demand improvements and in the areas that Antonio mentioned and the investment areas that were helping drive that expectation.

正確的。是的。正如安東尼奧所提到的，我們在 SAM 確實提到過，我們預計全年邊緣運算的同比增長將略有增長。當我想到這一點時，我幾乎將其分成兩半。正如我們所提到的，上半年我們仍然受益於積壓訂單或訂單簿狀況，這些將進入上半年並有助於支持收入表現。然後下半年將更加依賴需求的改善以及安東尼奧提到的領域和有助於推動這一預期的投資領域。

From an operating margin perspective, I would say that I would expect the first half to still benefit from higher operating margins as, again, the order book consumption has been at higher price with lower cost. So that's helped drive our operating margin performance, which again this quarter at 29.5% is exceptional. But we would expect, over time, probably more in the second half, you'll start seeing that operating margin rate get more back towards the mid-20s that we had mentioned at SAM is our expectation for the long term in this business.

從營業利潤率的角度來看，我預計上半年仍將受益於更高的營業利潤率，因為訂單消費價格更高且成本更低。這有助於提高我們的營業利潤率表現，本季我們的營業利潤率再次達到 29.5%，表現十分出色。但我們預計，隨著時間的推移，可能在下半年，您將開始看到營業利潤率回到我們在 SAM 提到的 20 年代中期，這是我們對這項業務的長期預期。

Thanks very much, Wamsi. Gary, could we have the next question?

非常感謝，Wamsi。加里，我們可以問下一個問題嗎？

The next question is from Toni Sacconaghi with Bernstein.

下一個問題來自伯恩斯坦的托尼·薩科納吉。

Yes. If I look at your -- take the midpoint of your first quarter revenue guide and I run out normal seasonality, I get revenues down about 5% for the year. You've just stated that the edge business is going to be weaker in the second half, so below normal seasonality. So clearly, you're expecting a huge ramp in HPC & AI over the course of the year. And I'm wondering if you can dimension that. Or are you expecting greater-than-normal seasonality in the traditional server business? And why would that be? And then finally, can you just comment on total backlog for HPC & AI exiting the quarter? You said it was $3 billion exiting last quarter or greater than $3 billion. What is it today?

是的。如果我看一下你的——取第一季收入指南的中點，並且超出正常的季節性，我會發現今年的收入下降了約 5％。您剛才說過，邊緣業務在下半年將會變得更弱，因此會低於正常的季節性。因此顯然，您預計今年 HPC 和 AI 將大幅成長。我想知道您是否可以測量該尺寸。或者您預計傳統伺服器業務的季節性將比正常情況更為強烈？為什麼會這樣呢？最後，您能否評論一下本季結束前的 HPC 和 AI 總積壓訂單量？您說上個季度的退出金額是 30 億美元或超過 30 億美元。今天是什麼日子？

Toni, this is Jeremy. I'll take those 2 for you. So you're spot on. As we think about next year, we will have a seasonal drop in HPC & AI. That's largely as Q4 really benefited from a meaningful amount of Cray EX acceptances. And as you know, the time between order and acceptance can be a long period of time. And Q4 saw a larger number of acceptances, which helped drive the revenue performance. We'll see a bit of a dip back down in Q1 and then Q2 and the second half really benefiting from the acceleration in AI as well as some additional supercomputing business. And so you'll see a pretty significant ramp as we go from Q1 to Q2 and then sustaining that or ramping beyond that in the second half of the year. And HPC & AI will be a big part of the revenue growth story for FY '24.

東尼，這是傑瑞米。我會為你拿那兩個。所以你說得對。展望明年，HPC 和 AI 將出現季節性下降。這主要是因為第四季度確實受益於大量 Cray EX 的接受度。如您所知，從訂單到驗收之間的時間可能會很長。第四季的接受數量有所增加，這有助於推動收入表現。我們將看到第一季和第二季出現小幅下滑，而下半年將真正受益於人工智慧的加速發展以及一些額外的超級運算業務。因此，從第一季到第二季度，您會看到一個相當顯著的成長，並且在下半年維持這種成長勢頭，或進一步成長。 HPC 和 AI 將成為 24 財年營收成長的重要組成部分。

Your question on the order book total, it landed at just over $3.2 billion, which was slightly above where we landed in Q3. And that really came off the back though of Q4 revenue performance in HPC & AI up about $350 million on a quarter-over-quarter basis. And so what happened is you had a runoff of order book from revenue performance in Q4 and then had a rebuild of that order book coming largely from AI demand in Q4 that took it back up to its historical high level of just over $3.2 billion.

您對訂單總額的問題，訂單總額剛超過 32 億美元，略高於第三季的水平。但這實際上是由於 HPC 和 AI 第四季的營收季增了約 3.5 億美元。因此發生的情況是，由於第四季度的收入表現，訂單量出現下滑，然後由於第四季度對人工智慧的需求而重新進行了訂單，使其回升至歷史最高水平，即略高於 32 億美元。

Thanks very much, Toni. Gary?

非常感謝，托尼。加里？

The next question is from Samik Chatterjee with JPMorgan.

下一個問題來自摩根大通的 Samik Chatterjee。

I guess in your prepared remarks, you did make it a point to highlight the uneven demand backdrop that you're seeing. I was wondering if you can flesh that out a bit more in terms of what you're seeing between the different product groups. And particularly when I look at that in contrast to the strong AI demand you're seeing, would you really sort of then see some of the AI demand from enterprises cannibalizing their own spend towards the other product groups? Or is the uneven demand more of inventory correction?

我想，在您準備好的發言中，您確實特意強調了您所看到的不均衡的需求背景。我想知道您是否可以根據您對不同產品組之間的了解進一步詳細說明這一點。特別是當我將其與您所看到的強勁人工智慧需求進行比較時，您是否真的會看到企業對人工智慧的部分需求蠶食其對其他產品組的支出？還是需求不均衡更多的是庫存調整？

Yes, sure. I mean no question, we see an explosion in demand in AI. Jeremy just commented on that. And I will say of that order book that Jeremy just talked about in Q4, only very little was converted in AI. And to the point he made, the growth we had in Q4 was driven by the supercomputing acceptances. But we have a very, very large pipeline in front of us, which is very exciting. But ultimately, it's going to come down to time to revenue based on the GPU availability.

是的，當然。我的意思是，毫無疑問，我們看到了人工智慧需求的爆炸性增長。傑里米剛剛對此發表了評論。我要說的是，傑里米在第四季度談到的訂單中，只有很少一部分轉化為人工智慧。正如他所說，我們第四季的成長是由超級運算的接受度所推動的。但我們面前有一條非常非常大的管道，這非常令人興奮。但最終，這將取決於基於 GPU 可用性的收入時間。

But I will say that business is going to continue to be super strong. And clearly, when I speak to customers, which I do more than 50% of my time, there is a huge amount of interest in AI and how to accelerate the deployment of AI across the enterprise, understanding there are challenges, whether it's sustainability challenges, where there are data center capacity, power and cooling and others. And that's why HPE went bold on that front last June to basically make the announcement we're going to offer supercomputing as a public cloud instance, so customers can use it as a virtual private cloud. So that -- we feel very good about it.

但我要說的是，業務將繼續保持強勁成長。顯然，當我與客戶交談時（我的 50% 以上的時間都在與客戶交談），他們對人工智慧以及如何加速整個企業部署人工智慧有著濃厚的興趣，了解其中存在的挑戰，無論是可持續性挑戰，還是資料中心容量、電力和冷卻等方面的挑戰。這就是為什麼 HPE 去年 6 月大膽宣布將提供超級運算作為公有雲實例，以便客戶可以將其用作虛擬私有雲。所以——我們對此感覺非常好。

GreenLake continued to be very strong. Just to give context, we added $1 billion in TCV quarter-over-quarter. We added 2,000 customers, and ARR obviously is a function of the deferred revenue that we materialize over time. But what customers really love about our experience is that it's hybrid by design. They can consume anything from edge to cloud to HPE GreenLake. Whether paid CapEx or OpEx, it doesn't really matter in the end. But they really love the experience. And that's why we're building the AI components into the same platform. So those 2, very strong.

GreenLake 持續保持強勁勢頭。舉個例子，我們上一季的 TCV 增加了 10 億美元。我們增加了 2,000 名客戶，ARR 顯然是我們隨著時間推移實現的遞延收入的函數。但客戶真正喜歡我們的體驗是因為它的混合設計。他們可以使用從邊緣到雲再到 HPE GreenLake 的任何東西。無論支付的是資本支出 (CapEx) 還是營運支出 (OpEx)，最終都並不重要。但他們確實熱愛這種體驗。這就是我們將 AI 元件建置到同一平台的原因。所以這兩個人非常強大。

Edge obviously had tremendous momentum. I think we're going to have the typical adjustments. But that's why we spend a lot of energy and time on adding more capabilities to the edge platform, security, SD-WAN, private 5G, data center networking, which adds to the momentum, understanding there will be potentially some digestion in the campus and branch. But as Jeremy said, we have very well covered for the first half of 2024. So we have to see that.

Edge 顯然擁有巨大的發展動能。我認為我們將進行典型的調整。但這就是為什麼我們花費大量精力和時間為邊緣平台、安全性、SD-WAN、私人 5G、資料中心網路添加更多功能，這增加了發展勢頭，我們知道校園和分支機構可能會有一些消化。但正如傑里米所說，我們已經為 2024 年上半年做好了充分的準備。

And then compute, right, is a typical business that goes through this cyclicality, right? So last year, obviously, we had a huge amount of orders. We converted the order faster than people expected. And in the back half of this year, we saw sequential demand improvements in units and stable AUPs. And now we start seeing upticks in the mix with AI inferencing, which has these accelerators. But Q3 demand was higher than Q2, and Q4 was higher than Q3. So I think it's fair to say we are stabilized, and we are improving. I would not call it yet a recovery.

然後計算，對吧，是一個經歷這種週期性的典型業務，對吧？因此，去年我們顯然有大量訂單。我們轉換訂單的速度比人們預期的要快。今年下半年，我們看到單位需求持續改善且 AUP 穩定。現在我們開始看到具有這些加速器的人工智慧推理的混合上升。但第三季的需求高於第二季度，第四季的需求高於第三季。因此我認為可以公平地說我們已經穩定下來，而且正在進步。我暫時還不認為這是復甦。

And then on storage, I believe we're going to see some improvements over time because of AI demand, which require file and object. And we have a great portfolio with HPE Alletra, and we intend to capitalize on that. But for 3 consecutive quarters now, we have seen stability and improvement. And in Q4, we saw revenue improvement on a sequential basis. So customers are prioritizing the spend where it makes sense, but ultimately we have a portfolio that can meet their needs, wherever they are. And HPE GreenLake is the way we deliver all of this which ultimately for shareholders drive higher margins and higher recurring revenues and profit.

然後在儲存方面，我相信隨著時間的推移我們會看到一些改進，因為人工智慧的需求需要文件和物件。我們與 HPE Alletra 擁有出色的產品組合，我們打算充分利用這一點。但現在已經連續三個季度看到了穩定和改善。在第四季度，我們看到營收環比成長。因此，客戶會優先考慮合理的支出，但最終我們會擁有一個可以滿足他們需求的產品組合，無論他們身在何處。 HPE GreenLake 是我們實現所有這些目標的方式，最終為股東帶來更高的利潤率、更高的經常性收入和利潤。

Cannibalization.

同類相食。

Cannibalization, sorry, Jeff just reminded me. Cannibalization, we have no evidence of that yet. I think that will become clear when the traditional compute CPU-driven return to some normal levels. But remember, not every customer has deployed cloud across their enterprise. Still quite a bit of journey to go. And there are clear customers assessing where is the best place to deploy that, whether it's in the private cloud or whether it's repatriating on-prem because of the cost or because of data.

同類相食，對不起，傑夫剛才提醒了我。同類相食，我們還沒有這方面的證據。我認為，當傳統運算 CPU 驅動恢復到正常水準時，這一點就會變得清晰。但請記住，並非每個客戶都在其企業內部署了雲端。還有很長的路要走。顯然，客戶正在評估部署的最佳地點，是在私有雲中，還是由於成本或資料而將其遣返回本地。

I think AI is a huge driver of repatriation in my mind because if you have data distributed across multiple estates, it's very hard to really train and fine-tune the models when you have data everywhere. And our focus there is really providing them an automated data pipeline with our unified analytics platform. So fundamentally, early to say. But so far, in the traditional Compute business, we have not seen evidence of cannibalization at this point in time.

我認為人工智慧是遣返的巨大推動力，因為如果你的資料分佈在多個房地產，那麼在資料無所不在的情況下，真正訓練和微調模型會非常困難。我們的重點是透過統一的分析平台為他們提供自動化的數據管道。因此從根本上來說，現在說這個還為時過早。但到目前為止，在傳統的計算業務中，我們還沒有看到蠶食的證據。

Samik, thank you. Gary?

薩米克，謝謝你。加里？

The next question is from Simon Leopold with Raymond James.

下一個問題來自雷蒙德詹姆斯 (Raymond James) 的西蒙利奧波德 (Simon Leopold)。

I wanted to see if you could talk a bit about the trends you're seeing in compute for the non-accelerated platforms. And really, the thing I'm trying to tease out here is sort of this issue of the AIOps projects pulling budget or sucking oxygen out of the room versus organizations buying up compute platforms to prepare for AI inferencing and embracing AI as an inferencing element, not just training.

我想看看您是否可以談談您在非加速平台運算方面看到的趨勢。實際上，我在這裡試圖闡明的問題是 AIOps 專案削減預算或消耗房間氧氣的問題，而組織則購買計算平台來為 AI 推理做準備，並將 AI 作為推理元素，而不僅僅是訓練。

Yes. Thanks, Simon. Again, maybe I will elaborate a little more to the comments I made before. So we saw Q4 over Q3 and Q3 over Q2 improvement in demand in units. And a lot of that was CPU driven, although there is a small base of AI accelerated kind of APUs, if you will, that we saw an increase in Q4. But I will say the unit growth in Q4 was not driven by the APUs. It was driven by a combination of CPUs, the vast majority, and some APUs because the base is still very, very small.

是的。謝謝，西蒙。再次，也許我會對我之前所發表的評論進行更詳細的闡述。因此，我們看到第四季的單位需求較第三季有所改善，第三季的單位需求較第二季有所改善。其中很大一部分是由 CPU 驅動的，儘管有一小部分 AI 加速類型的 APU，但我們在第四季度看到了成長。但我要說的是，第四季的單位成長並不是由 APU 推動的。它是由 CPU（絕大多數）和一些 APU 的組合驅動的，因為基數仍然非常非常小。

So definitely, customers are preparing for that. Again, they are all assessing what is the best place to deploy this model. That's why I do believe the influencing side will accelerate over time, whether it's to do some pretraining or POCs or really deploying in production. And I think many customers also will accelerate deployments of tuning solutions on-prem because of the data aspect I talked before.

因此，客戶肯定正在為此做準備。再次，他們都在評估部署該模型的最佳地點。這就是為什麼我確實相信影響力會隨著時間的推移而加速，無論是進行一些預訓練或 POC 還是真正部署在生產中。而且我認為，由於我之前談到的數據方面，許多客戶也將加速在本地部署調整解決方案。

Now question is still digesting what about last time on the CPU side of the house. But again, we saw some improvements in demand sequentially in units. And then let's remind ourselves that we also, for us, in the industry, we are going through the transition of Sapphire Rapids and ultimately what we call the Gen11 platform. That became now, what, Jeremy? 25% of the mix which...

現在的問題仍然是消化上次 CPU 方面的情況。但我們再次看到，單位需求有所改善。然後讓我們提醒自己，對於我們這個產業來說，我們也正在經歷 Sapphire Rapids 的轉型，以及最終所謂的 Gen11 平台的轉型。那現在變成了什麼，傑瑞米？ 25% 的混合物...

53% of orders in Q4.

第四季的訂單量為 53%。

53% of the orders in Q4, 25% of the revenue mix. And so that's good for us because, obviously, it drives higher density. And obviously, we can attach more options to the same platform. And customers like the sustainability piece of that and the hybrid by design nature of that, which is actually well optimized. And Gen11, by the way, was conceived to accept any type of processing unit, whether it's a CPU or whether it's an APU, including ARM-based solutions or GPU-based solutions, whether it's Intel or AMD, on the x86.

第四季的訂單量為53%，營收佔比為25%。這對我們來說是好事，因為顯然它可以提高密度。顯然，我們可以在同一個平台上添加更多選項。客戶喜歡它的可持續性和混合設計特性，它實際上已經得到了很好的優化。順便說一下，Gen11 的設計初衷是接受任何類型的處理單元，無論是 CPU 還是 APU，包括基於 ARM 的解決方案或基於 GPU 的解決方案，無論是英特爾還是 AMD，還是 x86 上的。

So that gives us tremendous amount of flexibility. But ultimately, it's not just about the server. It's the software that comes with it. And this is where we spend a lot of time building the partnerships and relationships with NVIDIA. So now you can deploy a tuning or inferencing with the NVIDIA stack and our software as well, all part of the HPE GreenLake.

這給了我們極大的彈性。但最終，這不僅僅與伺服器有關。它是隨附的軟體。我們花了大量時間與 NVIDIA 建立合作關係。因此，現在您可以使用 NVIDIA 堆疊和我們的軟體來部署調整或推理，它們都是 HPE GreenLake 的一部分。

Thank you very much, Simon. Gary?

非常感謝，西蒙。加里？

The next question is from Tim Long with Barclays.

下一個問題來自巴克萊銀行的提姆朗。

Can you just touch on the Storage business a little bit? It's been kind of challenged like some of the other businesses on macro. Could you talk a little bit about the outlook for recovery there? And also, if you could just touch on the third-party business there that's kind of impacted gross margin, profitability. How does that look to be trending as we look out over the next year or 2?

您能稍微談談儲存業務嗎？就像宏觀領域的其他一些業務一樣，它也面臨挑戰。能稍微談談那裡的復甦前景嗎？另外，如果您可以談談第三方業務，這對毛利率和獲利能力有一定影響。展望未來一兩年，這種趨勢會是怎麼樣？

Sure. I can take that, particularly towards the latter part of that. I think Antonio already hit on some of the demand dynamics, again, where we've seen 3 quarters of flat to increasing demand, and so some positive trends from that perspective. I think from an operating margin perspective, certainly, we saw a reduction in Q4. That was driven off a combination of several things, including a higher third-party mix that you mentioned as well as the fact that we see -- we saw some incremental OpEx in this segment and that OpEx as a comparison to the revenue performance in the quarter also put pressure onto that operating margin.

當然。我可以接受這一點，特別是後半部。我認為安東尼奧已經觸及了一些需求動態，我們再次看到 3 個季度的需求持平到增加，因此從這個角度來看存在一些積極的趨勢。我認為從營業利潤率的角度來看，我們在第四季度確實看到了下降。這是由多種因素造成的，包括您提到的第三方產品組合更高，以及我們看到這一部門的營運支出有所增加，與本季度的收入表現相比，營運支出也給營業利潤率帶來了壓力。

However, we do expect a pretty quick recovery there. We -- as we look into Q1. In particular, revenue is not expected to accelerate meaningfully, but we think the mix will improve as far as towards our IP product. And the -- we should see some OpEx moderation and favorability as we go into the quarter coming out of Q4 and some of the investments that we made there. And so I expect to see that get back into a low double-digit kind of area. And then as we work through the quarter and that IP mix starts to improve more on demand acceleration, then we should start seeing us working back towards our mid-teen target that we identified at SAM for this segment.

然而，我們確實預計那裡將會很快復甦。我們—當我們回顧第一季時。特別是，預計收入不會顯著增加，但我們認為，就我們的 IP 產品而言，收入組合將會改善。隨著我們進入第四季度，我們應該會看到營運支出有所緩和且有利，以及我們在那裡進行的一些投資。因此我期望看到這一數字回到兩位數以下的區域。然後，隨著我們整個季度的工作進展，並且 IP 組合開始隨著需求加速而進一步改善，那麼我們應該開始看到我們努力回到我們在 SAM 為該部門確定的中等水平目標。

I will say also, if you look at our HPE Alletra product, it's the fastest ramp we ever had in the history of the company. This quarter -- this past quarter grew another 50%. But also, there is some short-term impact because a portion of that revenue gets deferred because of the subscription, the software on the platform. And so that was an intentional strategy because ultimately, the infrastructure is one piece of it, but the operating system and the cloud services that comes with it are actually a subscription to HPE GreenLake.

我還要說的是，如果你看看我們的 HPE Alletra 產品，你會發現它是公司歷史上成長最快的產品。本季——上個季度又成長了 50%。但同時，也存在一些短期影響，因為部分收入會因為訂閱（平台上的軟體）而被延後。這是一個有意為之的策略，因為最終，基礎設施是其中的一部分，而隨之而來的作業系統和雲端服務實際上是對 HPE GreenLake 的訂閱。

So while we're growing 50%, we are not materializing the full revenue because a portion of that gets deferred at least to over 3 years. And that's good because it ultimately comes at significantly higher margins for us. But our strategy is to dramatically improve the mix to IP. And you will see more announcement this week in the storage portfolio all geared to the AI opportunity with file and object, and that will accelerate some of the momentum we have in the storage IP portfolio.

因此，雖然我們的成長了 50%，但我們並沒有實現全部收入，因為其中一部分收入至少要推遲 3 年以上。這是件好事，因為它最終會為我們帶來更高的利潤。但我們的策略是大幅改善 IP 組合。本週您將在儲存產品組合中看到更多公告，這些公告均針對文件和物件的 AI 機會，這將加速我們在儲存 IP 產品組合中的一些發展勢頭。

The next question is from Sidney Ho with Deutsche Bank.

下一個問題來自德意志銀行的 Sidney Ho。

I want to ask about ARR. And it was flat quarter-over-quarter but still up very strongly, 39% year-over-year. Can you walk us through the dynamics why it didn't change in the quarter? You just talked about GreenLake being very strong multiple times. Are there some negatives, maybe some cancellations offsetting the growth? Or is that more a pause of the 2 very, very strong quarters? And lastly, was there much contribution from AI servers in the ARR number at this point?

我想問ARR。與上一季相比，該數字持平，但與去年同期相比仍增長強勁，達 39%。您能否向我們介紹本季沒有改變的原因？您剛剛多次談到GreenLake非常強大。是否存在一些負面因素，例如一些取消訂單的情況會抵消成長嗎？或者這更像是兩個非常非常強勁的季度之間的停頓？最後，此時 AI 伺服器對 ARR 數字的貢獻有多大？

Thank you. I'll take that. So just on the last point, no, there wasn't any meaningful AI impact. But we do expect that to be an accelerator, particularly in FY '24 as we go to Q2 and towards the second half. That will be a big part of our ARR story, and we expect that to be an accelerator for us in FY '24. On the quarter-over-quarter, this business, similar to what I mentioned on the supercomputing area, does have some time between order to revenue recognition, in this case, when ARR begins to be reported.

謝謝。我接受。所以關於最後一點，沒有任何有意義的人工智慧影響。但我們確實預計這將是一個加速器，特別是在24財年，當我們進入第二季和下半年時。這將是我們 ARR 故事的重要組成部分，我們預計這將成為我們 24 財年的加速器。與上一季相比，這項業務與我在超級計算領域提到的類似，從訂單到收入確認確實需要一些時間，在這種情況下，也就是開始報告 ARR 的時候。

And so I think the sequential story was more about early in the year, we had seen more as the backlog had been burning down and some of those deals that have been waiting in the pipeline turning into -- and converting that helped drive and accelerate the ARR through the first 3 quarters. We saw a little bit less of that in Q4. But I don't think that at all as an indication of a slowdown in this space. In fact, between the 35% and 45% kind of CAGR or annual growth, I expect us in FY '24 to see the higher end of that range.

因此，我認為連續的故事更多是關於今年年初，我們看到更多的是，隨著積壓訂單的減少，一些一直在等待的交易變成了——並且轉化幫助推動和加速了前三個季度的 ARR。在第四季度，我們看到的情況略有減少。但我根本不認為這是該領域發展放緩的跡象。事實上，在 35% 到 45% 的複合年增長率之間，我預計我們在 24 財年將看到該範圍的高端。

Sidney, there's some rounding in there that we can talk through. But thank you. Gary, this should be our last question, I think.

西德尼，這裡面有一些事情我們可以討論一下。但謝謝你。加里，我想這應該是我們的最後一個問題。

And our final question will be from Meta Marshall with Morgan Stanley.

我們的最後一個問題來自摩根士丹利的 Meta Marshall。

This is [Mary] on for Meta Marshall. I just had a question on demand trends. Can you speak to linearity within the quarter and whether pockets of weakness you saw during the quarter changed as the quarter went on?

我是 Meta Marshall 的 [Mary]。我只是想問需求趨勢。您能否談談本季的線性情況，以及您在本季看到的弱點是否隨著季度的推移而改變？

Yes. I think overall it was more back ended, I would say, in the quarter. We saw strengthening as we went through the weeks. As always said, we have 13 weeks in the quarter, and we saw stronger momentum as we built along the way. And remember what we said at SAM, right? So at SAM, we said in AI year-to-date to October 19, I think, was the SAM date, we had $3 billion in cumulative orders booked between supercomputing and AI specifically and we ended the year at $3.6 billion. So in the last 12 days, we booked $600 million in incremental AI orders. That tells you they're strong.

是的。我認為，總體而言，本季的情況更糟。隨著時間的推移，我們看到了力量的增強。正如常說的那樣，本季已有 13 週，我們在前進的過程中看到了更強勁的勢頭。還記得我們在 SAM 說過的話嗎？因此在 SAM，我們表示，在 AI 領域，截至 10 月 19 日（我想是 SAM 日期），我們在超級運算和 AI 之間的累積訂單量為 30 億美元，年底的訂單量為 36 億美元。因此在過去的 12 天裡，我們獲得了 6 億美元的增量 AI 訂單。這說明他們很強大。

It was true also for compute and storage, by the way. The last few weeks, call it, 3, 4, 5 weeks, were stronger than the beginning of the quarter. So I would not make much out of that. Sometimes customers take the time. We still actually live in elongated sales cycles. That's for sure. Customers taking more time to make those decisions and ultimately issue the POs. But what really is giving me the confidence is the strong pipeline we have ahead of us. That's obvious and clear in AI. It's significantly stronger than we ever imagined.

順便說一句，這對於計算和存儲也是如此。過去幾週，也就是第 3、4、5 週，表現都比本季初強勁。所以我不會從中得到太多的收穫。有時顧客會花時間。我們實際上仍然處於延長的銷售週期。這是肯定的。客戶需要花費更多時間來做出這些決定並最終發出採購訂單。但真正讓我充滿信心的是我們前方強大的後盾。這在人工智慧中是顯而易見且明確的。它比我們想像的要強大得多。

And the only challenge we have there, as Jeremy said, right, so it's time to revenue. We really recognize very little revenue in AI in Q4. That's why we expect the acceleration starting in Q2 and beyond as lead times improve and some of the supercomputing also gets accepted. But the reality, 2024 will be the year of AI revenue growth. And then in the edge, obviously, we have the momentum that we talked about in the subscription, the scale of our software and the incremental engines that we have. So overall, it was a typical quarter but stronger on the back versus the front.

正如傑里米所說，我們面臨的唯一挑戰是，現在是時候實現收入了。我們在第四季在人工智慧方面的收入確實很少。這就是為什麼我們預計隨著交貨時間的縮短和一些超級計算得到接受，從第二季度及以後將開始加速。但現實是，2024 年將成為 AI 收入成長的一年。然後在邊緣方面，顯然，我們有我們在訂閱、軟體規模和增量引擎方面談到的勢頭。因此總體而言，這是一個典型的季度，但後端表現強於前端。

Yes. I think we have time for one more.

是的。我想我們還有時間再做一次。

Let's do one more.

我們再做一次吧。

Thanks. Gary, please.

謝謝。請叫我加里。

Maybe just one more question.

也許還剩一個問題。

The final question will be from Aaron Rakers with Wells Fargo.

最後一個問題來自富國銀行的 Aaron Rakers。

This is Michael on behalf of Aaron. I just want to ask around AI software. Can you just help us appreciate or understand how your own AI software solutions that you guys talked about at SAM compared to NVIDIA's own AI software suite? I'm just trying to understand. Is your software complementary? Or is it more of a substitute? Just how to think about that overall.

這是代表 Aaron 的 Michael。我只是想問一下有關人力智慧軟體的問題。您能否幫助我們評價或理解你們在 SAM 上談論的你們自己的 AI 軟體解決方案與 NVIDIA 自己的 AI 軟體套件相比有何不同？我只是想了解一下。你們的軟體是互補的嗎？或者它更像是一種替代品？只是整體上如何思考這個問題。

Yes. No, great question. And I will say overall, there is a lot of complementary and there are some places overlap, obviously. But with Jensen and the team, we have a clear joint plan to win together in different segments of the market. But let me break it out because we talk about software in general terms, but let's start first at the infrastructure level.

是的。不，這個問題問得好。我想說，總的來說，雙方有很多互補之處，而且顯然有一些地方有重疊。但有了 Jensen 和他的團隊，我們有一個明確的聯合計劃，以在不同領域共同獲勝。但讓我將其分解開來，因為我們一般性地討論軟體，但讓我們先從基礎設施層面開始。

We have unique software that allows us to run these supercomputers and AI system, which are cloud AI native by nature at massive scale. Think about when you run a model, you need to start and complete the model training. And you have to have unique technologies for checkpoints and making sure that all the compute power is acting as unified system because unlike the public cloud or the cloud as we know it, you run multiple workloads on multiple nodes. In this case, you run one workload on multiple nodes. And that's parallel computing as we know it. And ultimately, you need the software to run this at scale.

我們擁有獨特的軟體，可以讓我們運行這些超級電腦和人工智慧系統，它們本質上是大規模的雲端人工智慧原生。想想當你運行一個模型時，你需要啟動並完成模型訓練。您必須擁有獨特的檢查點技術，並確保所有運算能力都作為統一系統運行，因為與公有雲或我們所知的雲端不同，您在多個節點上運行多個工作負載。在這種情況下，您在多個節點上執行一個工作負載。這就是我們所知的平行計算。最終，您需要軟體來大規模運行它。

The magic around that is that checkpoint. And then the second piece of that is our networking interconnect fabric, which allows us to really connect every accelerated unit to every accelerated unit in a cohesive approach. And that's our Slingshot contingent fabric as we know it. And then on top of that, we have our machine learning development environment. This is where developers and the like use our machine learning development services to prepare the models to automate the data pipeline. One of the biggest challenges customers have is to prepare the data, data is everywhere, but ultimately bring it in terms sort of one place so you can use the data to train the models.

那個檢查站就是其中的神奇之處。第二部分是我們的網路互連結構，它使我們能夠以一種有凝聚力的方式將每個加速單元真正連接到每個加速單元。這就是我們所知的 Slingshot 應急結構。除此之外，我們還擁有機器學習開發環境。在這裡，開發人員等使用我們的機器學習開發服務來準備模型以實現資料管道的自動化。客戶面臨的最大挑戰之一是準備數據，數據無所不在，但最終要把數據集中到一個地方，這樣就可以使用數據來訓練模型。

And then with NVIDIA, we use their AI enterprise software, including some of the foundation models that they provide in order to provide a complete solution. And obviously, we leverage their APUs, call it, GPUs, whether it's H100, L40 or L40S, A100s in the past. And going forward, the announcement we made at Supercomputing 2023 in Denver, we are leveraging the Grace Hopper H200. So it's a combination depending on the use case. And we feel pretty good about what we're doing. And stay tuned because Thursday, we're going to make further announcements about our partnership with NVIDIA. But it's our IP and their IP that makes us together unique and differentiated in the AI space.

然後與 NVIDIA 合作，我們使用他們的 AI 企業軟體，包括他們提供的一些基礎模型，以提供完整的解決方案。顯然，我們利用他們的 APU，也就是 GPU，無論是 H100、L40 或 L40S，過去的都是 A100。展望未來，我們在丹佛 2023 超級電腦大會上宣布，我們將利用 Grace Hopper H200。所以這是一個取決於用例的組合。我們對自己所做的事情感到非常滿意。請繼續關注，因為我們將在周四進一步宣布與 NVIDIA 的合作關係。但正是我們的 IP 和他們的 IP 使我們在 AI 領域獨一無二、差異化。

Okay. Well, thank you, everyone. Appreciate always the time. I know you're busy covering all the earnings. But I will say just to wrap in fiscal year 2023, clearly, we demonstrated our strategic investments and the extraordinary innovation across the growth areas of edge, hybrid cloud and AI and even compute for that matter are really resonating with customers and is helping us accelerate revenue growth and profit diversification. That's why you see the growth in gross margin and profit. And I believe we will continue to capitalize on this growing market opportunities. And I'm confident to continue to increase the returns to our shareholders. And that's why we are raising the dividend for 2024. So thank you for your time today. I wish you all a fulfilling end of the calendar year and a special holiday season. Talk to you soon.

好的。好吧，謝謝大家。時刻珍惜時間。我知道你正忙著彌補所有的收入。但我要說的是，在 2023 財年結束時，我們清楚地展示了我們的策略投資以及在邊緣、混合雲和人工智慧甚至運算等成長領域的非凡創新，這些都真正引起了客戶的共鳴，並幫助我們加速營收成長和利潤多元化。這就是你看到毛利率和利潤成長的原因。我相信我們將繼續利用這個不斷成長的市場機會。我有信心繼續增加股東的回報。這就是我們提高 2024 年股息的原因。我祝福大家新年快樂、節日快樂。很快再和你聊。

Ladies and gentlemen, this concludes our call for today. Thank you for attending the presentation. You may now disconnect.

女士們、先生們，今天的通話到此結束。感謝您出席本次演講。您現在可以斷開連線。