International Business Machines Corp (IBM) 2024 Q2 法說會逐字稿

Welcome and thank you for standing by. (Operator Instructions) Today's conference is being recorded. If you have any objections, you may disconnect at this time.

歡迎並感謝您的支持。 （操作員指示）今天的會議正在錄音。如果您有任何異議，您可以此時斷開連接。

Now I will turn the meeting over to Olympia McNerney, IBM's Global Head of Investor Relations. Olympia, you may begin.

現在我將會議交給 IBM 全球投資者關係主管 Olympia McNerney。奧林匹亞，你可以開始了。

Thank you. I'd like to welcome you to IBM second-quarter 2024 earnings presentation. I'm Olympia McNerney, and I'm here today with Arvind Krishna, IBM's Chairman and Chief Executive Officer; and Jim Kavanaugh, IBM's Senior Vice President and Chief Financial Officer.

謝謝。歡迎您參加 IBM 2024 年第二季財報發表會。我是 Olympia McNerney，今天和 IBM 董事長兼執行長 Arvind Krishna 一起來到這裡。吉姆‧卡瓦諾 (Jim Kavanaugh)，IBM 資深副總裁兼財務長。

We'll post today's prepared remarks on the IBM investor website within a couple of hours, and a replay will be available by this time tomorrow. To provide additional information to our investors, our presentation includes certain non-GAAP measures. For example, all of our references to revenue and signings growth are at constant currency. We provided reconciliation charts for these and other non-GAAP financial measures at the end of the presentation, which is posted to our investor website.

我們將在幾個小時內將今天準備好的評論發佈在 IBM 投資者網站上，明天此時將提供重播。為了向我們的投資者提供更多信息，我們的簡報包括某些非公認會計原則措施。例如，我們所有提到的收入和簽約成長都是以固定匯率計算的。我們在演示結束時提供了這些和其他非公認會計準則財務指標的調節圖表，該圖表已發佈到我們的投資者網站上。

Finally, some comments made in this presentation may be considered forward looking under the Private Securities Litigation Reform Act of 1995. These statements involve factors that could cause our actual results to differ materially. Additional information about these factors is included in the Company's SEC filings.

最後，根據 1995 年《私人證券訴訟改革法案》，本簡報中的一些評論可能被視為前瞻性的。有關這些因素的更多資​​訊包含在公司向 SEC 提交的文件中。

So with that, I'll turn the call over to Arvind.

因此，我會將電話轉給 Arvind。

Thank you for joining us today to discuss IBM second-quarter earnings. We delivered a strong quarter, exceeding our expectations, driven by solid revenue growth, profitability, and cash flow generation. We had strong performance in software and infrastructure above our model as investment and innovation is yielding organic growth while consulting remained below model. Our results underscore the continued success of our hybrid cloud and AI strategy and the strength of our diversified business.

感謝您今天加入我們討論 IBM 第二季財報。在穩健的營收成長、獲利能力和現金流產生的推動下，我們實現了強勁的季度業績，超越了我們的預期。我們在軟體和基礎設施方面的表現強勁，高於我們的模型，因為投資和創新正在產生有機成長，而諮詢仍低於模型。我們的業績凸顯了我們的混合雲和人工智慧策略的持續成功以及我們多元化業務的實力。

Let me start with a few comments on the macroeconomic environment. Technology spending remains robust as it continues to serve as a key competitive advantage allowing businesses to scale, drive efficiencies and fuel growth. As we stated last quarter, factors such as interest rates and inflation impacted timing of decision making and discretionary spend in consulting. Overall, we remain confident in the positive macro-outlook for technology spending but acknowledge this impact.

首先我對宏觀經濟環境談幾點看法。技術支出仍然強勁，因為它繼續作為關鍵的競爭優勢，幫助企業擴大規模、提高效率並推動成長。正如我們上季所述，利率和通貨膨脹等因素影響了決策的時機和諮詢方面的可自由支配支出。總體而言，我們對技術支出的正面宏觀前景仍然充滿信心，但承認這種影響。

It has been a year since we introduced watsonx and our generative AI strategy to the market. We have infused AI across the business. From the tools clients use to manage and optimize their hybrid cloud environments, to our platform products across .ai, .data and .gov, to infrastructure and consulting, you can find AI innovation in all of our segments. For example, in software, our broad suite of automation products like Apptio and watsonx Orchestrate are leveraging AI and we expect to do the same with HashiCorp, once the acquisition is complete. Red Hat is bringing AI to OpenShift AI and RHEL AI.

自從我們向市場推出 watsonx 和我們的生成式 AI 策略以來已經一年了。我們已將人工智慧融入整個業務中。從客戶用來管理和優化其混合雲環境的工具，到我們跨.ai、.data 和.gov 的平台產品，再到基礎設施和諮詢，您可以在我們的所有細分市場中找到人工智慧創新。例如，在軟體方面，我們廣泛的自動化產品套件（如 Apptio 和 watsonx Orchestrate）正在利用人工智慧，我們希望在收購完成後對 HashiCorp 採取同樣的做法。紅帽正在將人工智慧引入 OpenShift AI 和 RHEL AI。

In transaction processing, we’re seeing early momentum in watsonx Code Assistant for Z. In infrastructure, IBM Z is equipped with real time AI inferencing capabilities. And in consulting, our experts are helping clients design and implement AI strategies.

在事務處理方面，我們看到了 watsonx Code Assistant for Z 的早期動能。在諮詢方面，我們的專家正在幫助客戶設計和實施人工智慧策略。

Our enterprise AI strategy is resonating as we evolve to meet client needs. Let me start by discussing IBM models. Choosing the right AI model is crucial for success in scaling AI. While large general-purpose models are great for starting on AI use cases, clients are finding that smaller models are essential for cost effective AI strategies. Smaller models are also much easier to customize and tune.

隨著我們不斷發展以滿足客戶需求，我們的企業人工智慧策略正在引起共鳴。讓我先討論 IBM 模型。選擇正確的人工智慧模型對於擴展人工智慧的成功至關重要。雖然大型通用模型非常適合開始人工智慧用例，但客戶發現較小的模型對於具有成本效益的人工智慧策略至關重要。較小的模型也更容易自訂和調整。

IBM's Granite models, ranging from 3 billion to 34 billion parameters and trained on 116 programming languages, consistently achieve top performance for a variety of coding tasks. To put cost in perspective, these fit-for-purpose models can be approximately 90% less expensive than large models. Hybrid cloud remains a top priority for clients as flexibility of deployment of AI models across multiple environments and data sovereignty remain a key focus.

IBM 的 Granite 模型擁有 30 億到 340 億個參數，並接受 116 種程式語言的訓練，始終能夠在各種編碼任務中實現最佳效能。從成本角度來看，這些專用型號比大型型號便宜約 90%。混合雲仍然是客戶的首要任務，因為跨多個環境部署人工智慧模型的靈活性和資料主權仍然是重點關注點。

We believe in the power of open innovation, and recently announced at IBM Think, that we open-sourced IBM’s Granite family of models, now available under Apache 2.0 licenses on both Hugging Face and GitHub. We see parallels to Linux becoming dominant in the enterprise server space; thanks to the speed and innovation offered by open source. We are confident that the same dynamic will play out with AI as we benefit from developer mindshare and community innovation.

我們相信開放創新的力量，最近在 IBM Think 上宣布，我們開源了 IBM 的 Granite 系列模型，現在可在 Hugging Face 和 GitHub 上根據 Apache 2.0 授權使用。我們看到 Linux 在企業伺服器領域佔據主導地位的相似之處；感謝開源提供的速度與創新。我們相信，隨著我們從開發者思想共享和社群創新中受益，人工智慧也將發揮同樣的作用。

We also recently launched InstructLab, a tool for more rapid model tuning through synthetic data generation, allowing our clients to more efficiently customize models using their own data and expertise. The last 12 months of AI pilots has made it clear that sustained value from AI requires truly leveraging enterprise data.

我們最近也推出了 InstructLab，這是一種透過合成資料產生更快速地調整模型的工具，使我們的客戶能夠利用自己的數據和專業知識更有效地客製化模型。過去 12 個月的人工智慧試點已經明確表明，人工智慧的持續價值需要真正利用企業數據。

In summary, our AI strategy is a comprehensive platform play. RHEL AI and OpenShift AI are the foundation of our enterprise AI platform. They combine open-source IBM Granite LLMs and InstructLab model alignment tools with full-stack optimization, enterprise grade security and support, and model indemnification.

綜上所述，我們的AI策略是全面的平台玩法。 RHEL AI 和 OpenShift AI 是我們企業 AI 平台的基礎。他們將開源 IBM Granite LLM 和 InstructLab 模型對齊工具與全端優化、企業級安全性和支援以及模型賠償相結合。

On top of that, we have an enterprise AI middleware platform with watsonx and an embed strategy with our AI assistants infused throughout our software portfolio and those of our ecosystem partners. In addition, our consulting services are critical in helping clients build their AI strategies from the ground up.

最重要的是，我們擁有一個具有 watsonx 的企業 AI 中間件平台，以及一個嵌入策略，其中我們的 AI 助理已融入我們的軟體組合和生態系統合作夥伴的軟體組合中。此外，我們的諮詢服務對於幫助客戶從頭開始製定人工智慧策略至關重要。

We also continue to see our infrastructure segment play a larger role as clients leverage their hardware investments in their AI strategies. Our book of business related to generative AI now stands at greater than $2 billion inception to date. The mix is roughly one-quarter software and three-quarters consulting signings.

隨著客戶在人工智慧策略中利用硬體投資，我們也持續看到我們的基礎設施部門發揮更大的作用。迄今為止，我們與生成式 AI 相關的業務規模已超過 20 億美元。其中約四分之一是軟體簽約，四分之三是顧問簽約。

We believe these strong results highlight our momentum and traction with clients. Our early leadership positions us for long-term success and this transformational technology, which is still in the initial stages of adoption.

我們相信這些強勁的業績凸顯了我們的動力和對客戶的吸引力。我們早期的領導地位使我們能夠取得長期成功，而這種變革性技術仍處於採用的初始階段。

As clients build out AI strategies, the IT landscape is becoming increasingly complex. Labor demographic shifts further emphasize the importance of optimizing IT spend and automating business processes. We continue to innovate and invest and have created a leading automation portfolio to capture this opportunity which you can see in our results. This includes Apptio for cost management, capabilities for observability and resource management, and with the announced acquisition of HashiCorp, the automation of cloud infrastructure.

隨著客戶制定人工智慧策略，IT 環境變得越來越複雜。勞動力人口結構的變化進一步強調了優化 IT 支出和自動化業務流程的重要性。我們不斷創新和投資，並創建了領先的自動化產品組合來抓住這個機會，您可以在我們的成果中看到這一點。其中包括用於成本管理的 Apptio、可觀察性和資源管理功能，以及宣布收購 HashiCorp 後的雲端基礎設施自動化。

The powerful combination of Red Hat's Ansible and Terraform will simplify provisioning and configuration of applications across hybrid cloud environments. The latest addition to this portfolio is IBM Concert, also announced at Think, a GenAI powered tool which helps clients get end-to-end visibility across business applications.

紅帽 Ansible 和 Terraform 的強大組合將簡化混合雲環境中應用程式的配置和配置。該產品組合的最新成員是 IBM Concert，它也是在 Think 上發布的，這是一款 GenAI 支援的工具，可協助客戶獲得跨業務應用程式的端到端可見性。

We also recently completed the acquisition of the StreamSets and webMethods assets from Software AG. This acquisition brings together leading capabilities in integration, API management, and data ingestion.

我們最近也完成了從 Software AG 收購 StreamSets 和 webMethods 資產。此次收購匯集了整合、API 管理和資料攝取方面的領先能力。

Let me now spend a minute on the continued strength we are seeing in infrastructure. IBM Z, our mainframe solution, is an integral part of our clients' hybrid cloud environments, driving their most secure and mission-critical workloads. Our latest cycle, z16, is uniquely tailored to offer clients security, scalability, and resiliency, which help clients address both cybersecurity threats and complex regulatory requirements.

現在讓我花一點時間談談我們在基礎設施方面看到的持續強勁勢頭。 IBM Z 是我們的大型主機解決方案，是客戶混合雲環境不可或缺的一部分，可驅動他們最安全和關鍵任務的工作負載。我們最新的周期 z16 是專門為客戶提供安全性、可擴展性和彈性而客製化的，可協助客戶應對網路安全威脅和複雜的監管要求。

z16's Telum processor is a unique differentiator, driving real time, in-line AI inferencing at unprecedented speed and scale for applications like real time fraud detection. Our storage offerings are also benefitting from generative AI as clients address data readiness and need high-speed access to massive volumes of unstructured data.

z16 的 Telum 處理器是一個獨特的優勢，能夠以前所未有的速度和規模驅動即時、內聯 AI 推理，適用於即時詐欺偵測等應用。隨著客戶解決數據準備問題並需要高速存取大量非結構化數據，我們的儲存產品也受益於生成式人工智慧。

We continue to invest in innovation and make great progress in emerging technology like quantum computing. This quarter, we expanded Qiskit, IBM’s quantum computing software, into a comprehensive stack aimed at optimizing performance on utility-scale quantum hardware. These updates aim to enhance the stability, efficiency, and usability of Qiskit, supporting advanced quantum algorithm development and fostering broader adoption across various industries.

我們持續投資創新，並在量子運算等新興技術方面取得巨大進展。本季度，我們將 IBM 的量子運算軟體 Qiskit 擴展為一個綜合堆疊，旨在優化公用事業規模量子硬體的效能。這些更新旨在提高 Qiskit 的穩定性、效率和可用性，支援先進的量子演算法開發並促進各行業更廣泛的採用。

This strong momentum and innovation across the portfolio manifests itself in client adoption. In virtually all industries and geographies, clients leverage IBM solutions to help them transform their operations and create better experiences for end-users. Names like Virgin Money, Crédit Mutuel, and Panasonic all turned to IBM in the quarter.

整個產品組合的強勁勢頭和創新體現在客戶採用上。幾乎在所有行業和地區，客戶都利用 IBM 解決方案來幫助他們轉變營運方式並為最終用戶創造更好的體驗。 Virgin Money、Crédit Mutuel 和 Panasonic 等公司在本季都轉向了 IBM。

We also continue to strengthen our ecosystem. At our Think event, we announced a series of new AI partnerships with industry leaders like Adobe, AWS, Microsoft, Meta, Mistral, Salesforce, and SAP. In May, IBM and Palo Alto Networks announced a partnership to deliver AI-powered security solutions using watsonx. As part of this, Palo Alto is acquiring IBM's QRadar SaaS assets, and we are partnering to offer seamless migration for QRadar customers to XSIAM. IBM will also train over 1,000 security consultants on Palo Alto Network products to drive a significant book of business with them.

我們也繼續加強我們的生態系統。在 Think 活動中，我們宣布與 Adob​​e、AWS、Microsoft、Meta、Mistral、Salesforce 和 SAP 等行業領導者建立一系列新的 AI 合作夥伴關係。 5 月，IBM 和 Palo Alto Networks 宣佈建立合作夥伴關係，利用 watsonx 提供由人工智慧驅動的安全解決方案。作為該計劃的一部分，Palo Alto 正在收購 IBM 的 QRadar SaaS 資產，我們正在合作為 QRadar 客戶提供到 XSIAM 的無縫遷移。 IBM 也將針對 Palo Alto Network 產品對 1,000 多名安全顧問進行培訓，以推動與他們的重要業務往來。

In summary, we are excited to continue delivering strong results. Given our first-half performance, we are raising our expectations for free cash flow to greater than $12 billion for the year.

總之，我們很高興能夠繼續取得強勁的業績。鑑於我們上半年的業績，我們將今年自由現金流的預期提高到超過 120 億美元。

I will now hand over to Jim to walk you through the details of the quarter. Jim, over to you

現在我將請吉姆向您介紹本季的詳細資訊。吉姆，交給你了

Thanks, Arvind. In the second quarter, we delivered $15.8 billion in revenue, $2.8 billion of operating pre-tax income, and $2.43 operating diluted earnings per share. Our 4% revenue growth at constant currency combined with greater than 200 basis points of operating pre-tax margin expansion drove 17% operating pre-tax income growth and 11% operating diluted earnings per share growth, highlighting our strong execution.

謝謝，阿爾溫德。第二季度，我們實現了 158 億美元的收入、28 億美元的稅前營業收入和 2.43 美元的稀釋後每股營業收益。以固定匯率計算，我們4% 的收入增長加上超過200 個基點的稅前營業利潤率擴張，推動了17% 的稅前營業收入增長和11% 的營業攤薄每股收益增長，凸顯了我們強大的執行力。

And through the first half, we generated $4.5 billion of free cash flow. Our free cash flow generation is the strongest first half level we have reported in many years. We are pleased with these results, exceeding our expectations for revenue, profitability, free cash flow. and earnings per share.

上半年，我們創造了 45 億美元的自由現金流。我們的自由現金流產生是我們多年來報告的最強勁的上半年水準。我們對這些結果感到滿意，超越了我們對收入、獲利能力和自由現金流的預期。和每股收益。

Revenue growth was led by software and infrastructure. It is clear that our investments in innovation are yielding results and driving strong organic growth across these segments. Software grew by 8%, with solid growth across hybrid platform and solutions and transaction processing, and strong transactional performance.

收入成長由軟體和基礎設施帶動。顯然，我們對創新的投資正在產生成果，並推動這些領域的強勁有機成長。軟體成長了 8%，混合平台、解決方案和事務處理均實現穩健成長，事務效能強勁。

Infrastructure had great performance, up 3%, delivering growth across IBM Z and distributed infrastructure. Consulting was up 2% and continued to be impacted by a pullback in discretionary spending.

基礎架構表現出色，成長了 3%，推動了 IBM Z 和分散式基礎架構的成長。諮詢業成長 2%，並持續受到可自由支配支出縮減的影響。

Looking at our profit metrics, we expanded operating gross margin by 190 basis points and operating pre-tax margin by 220 basis points over last year, inclusive of about a 30-basis point currency headwind to pre-tax margin. Margin expansion was driven by our operating leverage, product mix, and ongoing productivity initiatives.

從我們的利潤指標來看，我們的營業毛利率比去年提高了 190 個基點，營業稅前利潤率提高了 220 個基點，其中稅前利潤率面臨約 30 個基點的貨幣逆風。利潤率的成長是由我們的營運槓桿、產品組合和持續的生產力計劃所推動的。

Driving productivity is core to our operating and financial model. This includes enabling a higher value workforce through automation and AI, streamlining our supply chain, aligning our teams by workflow, and reducing our real estate footprint.

提高生產力是我們營運和財務模式的核心。這包括透過自動化和人工智慧打造更高價值的勞動力、簡化我們的供應鏈、透過工作流程調整我們的團隊以及減少我們的房地產足跡。

These actions allowed for continued investments in innovation, with R&D up 9% in the first half. This includes investments in both AI and hybrid cloud, as well as infrastructure ahead of our next Z program in 2025, which we expect to accelerate our organic growth profile over time.

這些行動使得對創新的持續投資得以實現，上半年研發成長了 9%。這包括對人工智慧和混合雲的投資，以及 2025 年下一個 Z 計畫之前的基礎設施投資，我們預計該計畫將隨著時間的推移加速我們的有機成長。

Our results this quarter reflect broad-based growth and the strength in the fundamentals of our business, with revenue up about $300 million, operating pre-tax income up about $400 million, adjusted EBITDA up more than $350 million, and free cash flow up about $500 million. For the first half, we generated $4.5 billion of free cash flow, up $1.1 billion year over year.

我們本季的業績反映了廣泛的成長和業務基本面的實力，營收成長約 3 億美元，稅前營業收入成長約 4 億美元，調整後 EBITDA 成長超過 3.5 億美元，自由現金流成長約5億美元。上半年，我們產生了 45 億美元的自由現金流，比去年同期增加 11 億美元。

The largest driver of this first half growth comes from adjusted EBITDA, up about $550 million year over year, and timing of capex. We are a few points ahead of our two-year average attainment levels through the first half.

上半年成長的最大驅動力來自調整後的 EBITDA（年成長約 5.5 億美元）以及資本支出的時機。上半年我們比兩年平均領先幾個百分點。

In terms of cash uses, we returned $3.1 billion to shareholders in the first half in the form of dividends. From a balance sheet perspective, we have a very strong liquidity position with cash of $16 billion, up $2.5 billion since year end 2023. Our debt balance at the end of the second quarter was flat with yearend 2023 at $56.5 billion, including $11.1 billion from our financing business.

在現金使用方面，上半年我們以股利的形式向股東返還了31億美元。從資產負債表的角度來看，我們擁有非常強勁的流動性頭寸，現金為160 億美元，自2023 年底以來增加了25 億美元。美元，其中包括來自2023 年底的111 億美元。

Putting this all together, our business fundamentals remain solid with continued revenue growth, margin expansion, cash generation, and a strong balance sheet with financial flexibility to support our business.

總而言之，我們的業務基本面仍然穩固，收入持續成長，利潤率擴大，現金產生，以及強大的資產負債表和財務靈活性來支持我們的業務。

Turning to the segments, software revenue growth accelerated to 8% this quarter. Both hybrid platform and solutions and transaction processing grew, as clients leverage the capabilities of our AI and hybrid cloud platforms. This performance reflects the investments we’ve been making in software, both organically, which drove more than 6 points of the growth, as well as acquisitions.

轉向細分市場，本季軟體收入增速加快至 8%。隨著客戶利用我們的人工智慧和混合雲平台的功能，混合平台和解決方案以及交易處理都在成長。這項業績反映了我們在軟體方面的有機投資（推動了超過 6 個百分點的成長）以及收購。

As mentioned in January, the software revenue growth drivers for the year include Red Hat growth, the combination of innovation, recurring revenue, and transaction processing, as well as acquisitions. Let me spend a minute on each of these elements.

正如一月份提到的，今年軟體收入成長的驅動因素包括紅帽的成長、創新、經常性收入、交易處理以及收購的結合。讓我花一點時間來討論這些要素。

In Red Hat, annual bookings growth accelerated to over 20% this quarter. Within that performance, OpenShift annual bookings were up over 40% and RHEL and Ansible growth was double digit. This strength reflects the demand for our hybrid cloud solutions, including app modernization, management automation, generative AI, and virtualization. In a subscription-based business, the majority of revenue is under contract for the next two quarters. Think of it as cRPO for the next six months.

在紅帽，本季年度預訂量成長加速至 20% 以上。在這項業績中，OpenShift 年度預訂量成長了 40% 以上，RHEL 和 Ansible 成長了兩位數。這種優勢反映了對我們混合雲端解決方案的需求，包括應用程式現代化、管理自動化、生成式人工智慧和虛擬化。在基於訂閱的業務中，大部分收入來自未來兩季的合約。將其視為未來六個月的 cRPO。

This metric is growing in the mid-teens and accelerated more than 5 points versus the first half of the year. We continue to bring new innovation to our portfolio and it’s contributing nicely to our software performance. Our new innovation includes generative AI offerings like watsonx, our AI middleware, watsonx assistants, the recently announced IBM Concert, and others, which contributed about $0.5 billion to our AI book of business inception to date.

該指標正以十幾歲左右的速度增長，與今年上半年相比加速了 5 個百分點以上。我們不斷為我們的產品組合帶來新的創新，這對我們的軟體效能做出了很好的貢獻。我們的新創新包括生成式 AI 產品，如 watsonx、我們的 AI 中間件、watsonx 助理、最近宣布的 IBM Concert 等，這些產品迄今為止為我們的 AI 業務貢獻了約 5 億美元。

And we delivered good growth across our recurring revenue base, which is about 80% of annual software revenue. This is evident in hybrid platform and solutions where our ARR is now $14.1 billion, and up 9% since last year.

我們的經常性收入基礎實現了良好的成長，約佔年度軟體收入的 80%。這在混合平台和解決方案中表現得尤為明顯，我們的 ARR 目前為 141 億美元，自去年以來增長了 9%。

Transaction processing delivered 13% revenue growth. This performance demonstrates the innovation and value of our mission critical hardware stack across IBM Z, Power, and Storage. The combination of growing demand for capacity, good client renewals, and strong large deal performance fueled our results. And, notably, our new generative AI portfolio innovation, watsonx Code Assistant for Z, is resonating well with clients. Together, these dynamics contributed to both recurring and transactional software revenue growth again this quarter.

交易處理帶來了 13% 的收入成長。這項效能展示了我們跨 IBM Z、Power 和 Storage 的關鍵任務硬體堆疊的創新和價值。不斷增長的產能需求、良好的客戶續約以及強勁的大宗交易業績共同推動了我們的業績。值得注意的是，我們新的生成式人工智慧產品組合創新 watsonx Code Assistant for Z 引起了客戶的良好共鳴。這些動態共同推動了本季經常性和交易軟體收入的成長。

Revenue performance this quarter also benefitted from our focused M&A strategy, including synergies realized across the portfolio. This included the recent Apptio acquisition. Less than 12 months since closing, we have accelerated annual bookings and are seeing an uptick in ARR growth, already in the mid-teens. The synergy between Apptio's FinOps offerings and our broader automation portfolio helps clients manage, optimize, and automate technology spending decisions.

本季的營收表現也受惠於我們專注的併購策略，包括整個投資組合所實現的綜效。這包括最近收購的 Apptio。自交易結束後不到 12 個月，我們加快了年度預訂速度，並看到 ARR 成長加快，已達到 15 左右。 Apptio 的 FinOps 產品與我們更廣泛的自動化產品組合之間的協同作用可協助客戶管理、最佳化和自動化技術支出決策。

Earlier this month, we completed the acquisition of StreamSets and webMethods from Software AG and expect the HashiCorp acquisition to close by year end. Looking at software profit, gross profit margin expanded, and segment profit was up over 350 basis points year to year, with the latter reflecting operating leverage driven by our revenue scale and mix this quarter.

本月早些時候，我們完成了對 Software AG 的 StreamSets 和 webMethods 的收購，並預計對 HashiCorp 的收購將在年底前完成。從軟體利潤來看，毛利率擴大，部門利潤年增超過350個基點，後者反映了本季我們的營收規模和組合驅動的營運槓桿。

Our consulting revenue was up 2%, consistent with last quarter and largely reflecting organic growth. In April, we discussed that we were seeing solid demand for our large transformational offerings as clients continue to prioritize driving productivity with AI and analytics. At the same time, we saw a pull back on discretionary projects as clients prioritized their spending. The second-quarter buying behavior played out much in the same way.

我們的諮詢收入成長了 2%，與上個季度一致，很大程度上反映了有機成長。四月份，我們討論過，隨著客戶繼續優先考慮透過人工智慧和分析來提高生產力，我們看到對我們的大型轉型產品的強勁需求。同時，隨著客戶優先考慮支出，我們發現可自由支配項目減少。第二季的購買行為也大致如此。

Signings for the quarter were $5.7 billion, driven by solid demand for large engagements across finance and supply chain transformation, cloud modernization, and application development. This contributed to backlog growth of 5% year over year and our trailing 12-month book-to-bill remaining over 1.15. Meanwhile, continued discretionary spending constraints impacted our small engagement performance and backlog realization in the quarter.

由於金融和供應鏈轉型、雲端現代化和應用程式開發方面的大型活動的強勁需求，該季度的簽約額達到 57 億美元。這使得積壓訂單年增 5%，並且我們過去 12 個月的訂單出貨比保持在 1.15 以上。同時，持續的可自由支配支出限制影響了我們本季的小額參與績效和積壓實現。

As Arvind mentioned, our book of business in generative AI inception to date is greater than $2 billion, and about three-quarters of it represents consulting signings, with strong quarter-over-quarter momentum. Our extensive industry and domain expertise has placed us in an early leadership role, which is crucial at the onset of a technology shift.

正如 Arvind 所提到的，迄今為止，我們在生成式 AI 領域的業務規模已超過 20 億美元，其中約四分之三是顧問簽約，季度環比勢頭強勁。我們廣泛的行業和領域專業知識使我們處於早期的領導角色，這在技術轉變開始時至關重要。

IBM has both technology and consulting, which is a unique and powerful combination to help clients navigate this technology transition. Similar to previous technology shifts, such as the advent of the Internet, globalization, and cloud computing, generative AI is driving the next wave of growth. In a human capital-based business, signings represent clients re-prioritizing spend on this technology transition while there is some potential for lift as the total addressable market expands.

IBM 擁有技術和諮詢服務，這是一個獨特且強大的組合，可以幫助客戶駕馭這項技術轉型。與網路、全球化和雲端運算等先前的技術變革類似，生成式人工智慧正在推動下一波成長。在以人力資本為基礎的企業中，簽約代表客戶重新確定技術轉型支出的優先順序，同時隨著總目標市場的擴大，還有一定的提升潛力。

We are delivering value in two ways. First, partnering with our clients to design and scale AI solutions, whether that be leveraging AI capabilities of IBM, our partners, or a combination. Second, we are developing new ways of working, driving productivity and improving delivery, all with our Consulting Advantage platform.

我們透過兩種方式提供價值。首先，與我們的客戶合作設計和擴展 AI 解決方案，無論是利用 IBM、我們的合作夥伴還是兩者的結合的 AI 功能。其次，我們正在開發新的工作方式，提高生產力並改善交付，所有這些都透過我們的諮詢優勢平台實現。

In summary, GenAI is acting as a catalyst for companies to grow revenues, cut costs, and change the ways they work, creating a significant opportunity for us. We are seeing this already as IBM is the strategic partner of choice for clients using this technology including WPP, Elevance Health, and the UK’s Department of Work and Pensions.

總之，GenAI 正在充當公司增加收入、削減成本和改變工作方式的催化劑，為我們創造了重要的機會。我們已經看到了這一點，因為 IBM 是使用該技術的客戶的首選策略合作夥伴，這些客戶包括 WPP、Elevance Health 以及英國工作和退休金部門。

Turning to our lines of business, business transformation revenue grew 6% led by finance and supply chain transformations. Data transformation also contributed to growth. In technology consulting, revenue was up 1%. Growth was driven by application modernization services. Application operations revenue declined reflecting weakness in on-prem custom application management, partially offset by strength in cloud-based application management offerings.

轉向我們的業務線，在財務和供應鏈轉型的帶動下，業務轉型收入成長了 6%。數據轉型也促進了成長。技術諮詢業務的收入成長了 1%。成長是由應用現代化服務推動的。應用程式營運收入下降，反映出本地自訂應用程式管理的疲軟，但部分被基於雲端的應用程式管理產品的強勁所抵消。

Looking at consulting profit, we expanded gross profit margin by 40 basis points, driven by productivity and pricing actions we have taken. Segment profit margin was modestly down, reflecting continued labor inflation and currency.

就諮詢利潤而言，在我們採取的生產力和定價行動的推動下，我們的毛利率提高了 40 個基點。部門利潤率小幅下降，反映出持續的勞動力通膨和貨幣上漲。

Moving to infrastructure, revenue was up 3%. We’re capitalizing on the strong and broad-based demand for our hardware platforms, especially IBM Z.

轉向基礎設施領域，收入成長了 3%。我們正在利用對我們的硬體平台（尤其是 IBM Z）的強勁而廣泛的需求。

Within hybrid infrastructure, IBM Z revenue was up 8% this quarter. We’re now more than two years into the z16 cycle and the revenue performance continues to outperform prior cycles. Our clients are facing increasing demands for workloads given rapid business expansion, the complex regulatory environment, and increasing cybersecurity threats and attacks. IBM Z addresses these needs with the combination of cloud native development for hybrid cloud, embedded AI at scale, quantum-safe security, energy efficiency, and strong reliability and scalability.

在混合基礎設施中，IBM Z 本季營收成長了 8%。我們現在進入 z16 週期已經兩年多了，收入表現繼續優於之前的周期。由於業務的快速擴張、複雜的監管環境以及日益增加的網路安全威脅和攻擊，我們的客戶面臨日益增長的工作負載需求。 IBM Z 透過結合混合雲的雲端原生開發、大規模嵌入式 AI、量子安全、能源效率以及強大的可靠性和可擴展性來滿足這些需求。

Increasing workloads translates to more Z capacity or MIPS, which are up about 3 fold over the last few cycles. IBM Z remains an enduring platform for mission critical workloads, driving both hardware and related software, storage, and services adoption.

工作負載的增加意味著更多的 Z 容量或 MIPS，在過去幾個週期中增加了約 3 倍。 IBM Z 仍然是任務關鍵型工作負載的持久平台，推動硬體和相關軟體、儲存和服務的採用。

In distributed infrastructure, revenue grew 5% driven by strength in both Power and Storage. Power growth was fueled by demand for data intensive workloads on Power 10 led by SAP Hana. Storage delivered growth again this quarter, including growth in high-end storage tied to the z16 cycle and solutions tailored to protect, manage, and access data for scaling generative AI.

在分散式基礎設施領域，由於電力和儲存實力的推動，收入增加了 5%。 SAP Hana 主導的 Power 10 上的資料密集型工作負載需求推動了電力成長。本季儲存再次實現成長，包括與 z16 週期相關的高階儲存的成長以及為保護、管理和存取資料以擴展生成式 AI 而客製化的解決方案。

Looking at infrastructure profit, we delivered solid gross profit margin expansion and segment profit accelerated quarter to quarter to the high-teens. Segment profit margin was down 230 basis points in the quarter, reflecting key investments we’re making in the business across areas like AI, hybrid cloud. and quantum, and almost a point of impact due to currency

就基礎設施利潤而言，我們實現了穩健的毛利率擴張，部門利潤逐季加速至高雙位數。本季的部門利潤率下降了 230 個基點，反映出我們在人工智慧、混合雲等領域的業務進行了關鍵投資。和量子，幾乎是由於貨幣的影響點

Now, let me bring it back to the IBM level to wrap up. We feel good about our performance in the first half with revenue growth reflecting the investments we’ve been making, both organically, as well as acquisitions. Our focus on execution and the strength in the fundamentals of our business resulted in strong performance in the quarter across revenue, margin expansion, and growth in profitability and earnings.

現在，讓我回到 IBM 層面來總結。我們對上半年的業績感到滿意，收入成長反映了我們一直在進行的有機投資和收購。我們對執行力的關注以及業務基本面的實力導致本季在營收、利潤率擴張以及獲利能力和收益成長方面表現強勁。

Looking to the full year 2024, we are holding our view of revenue. We see full-year constant currency revenue growth in line with our mid-single-digit model, still prudently at the low end. For free cash flow, given the strength in our performance in the first half, we feel confident in raising our expectations to greater than $12 billion, driven primarily by growth in adjusted EBITDA. This also includes a modest contribution resulting from the Palo Alto QRadar transaction, largely offset by related structural actions to address stranded costs. We continue to expect the QRadar transaction to close by the end of the third quarter.

展望 2024 年全年，我們維持對收入的看法。我們認為全年恆定貨幣收入成長符合我們的中個位數模型，但仍謹慎地處於低端。對於自由現金流，鑑於我們上半年的強勁業績，我們有信心將預期提高到 120 億美元以上，這主要是由調整後 EBITDA 的成長推動的。這也包括帕洛阿爾托 QRadar 交易產生的適度貢獻，但大部分被解決擱淺成本的相關結構性行動所抵消。我們仍然預計 QRadar 交易將在第三季末完成。

On the segments, in software, we had a solid first half performance, up more than 7%. This performance reflects strength in our recurring revenue base and early traction in GenAI. With this performance, we are raising our view of growth in software to high-single digits for the year. And given ongoing productivity initiatives and operating leverage, we now expect software segment profit margin to expand by over a point.

在軟體領域，我們上半年表現強勁，成長了 7% 以上。這項業績反映了我們經常性收入基礎的實力以及 GenAI 的早期吸引力。憑藉這一業績，我們將今年軟體成長的預期提高到高個位數。鑑於持續的生產力計劃和營運槓桿，我們現在預計軟體部門的利潤率將擴大一個百分點以上。

In consulting, given the continued pressure we have seen on spending related to discretionary projects, we now expect low single digit growth for the year and segment profit margin to expand by about half a point. And given the strength in infrastructure in the first half, we now expect it to be about neutral for the year with segment profit margin in the mid- to high-teens.

在諮詢方面，鑑於我們看到與可自由支配項目相關的支出持續面臨壓力，我們現在預計今年的低個位數成長和部門利潤率將擴大約半個百分點。鑑於上半年基礎建設的強勁勢頭，我們現在預計今年的基礎建設將保持中性，部門利潤率將達到中雙位數至高雙位數。

With these segment dynamics, we are raising our expectations of operating pre-tax margin expansion to over half a point year to year. And we are maintaining our view of operating tax rate in the mid-teens range, consistent with last year. On currency, given the strengthening of the dollar, we now expect a 100 to 200 basis point impact to revenue growth for the year.

鑑於這些細分市場的動態，我們將稅前營業利潤率較去年同期擴張的預期提高到半個百分點以上。我們維持對營業稅率在十幾歲範圍內的看法，與去年一致。貨幣方面，鑑於美元走強，我們目前預計今年營收成長將受到 100 至 200 個基點的影響。

For the third quarter, we see revenue growth consistent with the full year. For profit, we expect our net income skew through the third quarter to remain a couple points ahead of the prior year, driven by the strength of our business. And again, we expect the gain from the Palo Alto QRadar transaction will be offset by related structural actions to address stranded costs.

對於第三季度，我們認為營收成長與全年一致。就利潤而言，在我們業務實力的推動下，我們預計第三季的淨利潤偏差將比去年同期高出幾個百分點。同樣，我們預期 Palo Alto QRadar 交易的收益將被解決擱淺成本的相關結構性行動所抵銷。

In closing, we are pleased with our performance this quarter and for the first half, driving confidence in our updated expectations. We are positioned to grow revenue, expand operating profit, and grow free cash flow for the year.

最後，我們對本季和上半年的表現感到滿意，這增強了我們對最新預期的信心。我們的目標是增加今年的收入、擴大營業利潤並增加自由現金流。

Arvind and I are now happy to take your questions. Olympia, let’s get started.

阿溫德和我現在很高興回答您的問題。奧林匹亞，讓我們開始吧。

Thank you, Jim. Before we begin the Q&A, I'd like to mention a couple of items. First, supplemental information is provided at the end of the presentation. And then second, as always, I'd ask you to refrain from multi-part questions. Operator, let's please open it up for questions.

謝謝你，吉姆。在開始問答之前，我想提幾件事。首先，在演示結束時提供補充資訊。其次，一如既往，我會要求你不要提出由多部分組成的問題。接線員，請打開它提問。

(Operator Instructions) Wamsi Mohan, Bank of America.

（操作員指示）Wamsi Mohan，美國銀行。

Thank you so much. Your long-term model on transaction processing is low-single digit, and you’ve just posted a very strong quarter with 13% growth in the quarter. How should we think about that trajectory of that in 2024 and maybe in 2025? I know, Jim, you noted a few different things including solid client renewals and some strong large deal performance. Was there anything very episodic or unusually large within that mix as well? Thank you so much.

太感謝了。你們的交易處理長期模型是低個位數的，而且你們剛剛發布了一個非常強勁的季度，該季度增長了 13%。我們該如何看待 2024 年甚至 2025 年的發展軌跡？我知道，吉姆，您注意到了一些不同的事情，包括穩定的客戶續約和一些強勁的大交易業績。在這個組合中是否也有任何非常偶然或異常大的事情？太感謝了。

Thanks, Wamsi. I appreciate the question overall. Very important. If you take a step back, we continue to be very pleased with our transaction processing performance overall, if you dial back to when we laid out our mid-term model, we said we converted this to a growth factor, low-single digit, overall. And if you look at the last couple of years, we've been averaging mid-single digit or better, overall. We've shifted this now to a growth contributor. And why is that important?

謝謝，瓦姆西。總的來說，我很欣賞這個問題。很重要。如果你退後一步，我們仍然對我們的整體交易處理性能感到非常滿意，如果你回撥到我們制定中期模型時，我們說我們將其轉換為增長因子，低個位數，全面的。如果你看看過去幾年，我們的平均水平一直在個位數左右或更好，總體而言。我們現在已將其轉變為成長貢獻者。為什麼這很重要？

One, high source of profit and cash to fund investment flexibility. And two, it provides a very solid incumbency base for the IBM or multiplier effect. But if you take a look at it, we are capitalizing on the strength that we've seen over the last three programs of our mainframe cycle. It's really instantiated the enduring value of that platform.

一、利潤來源高，現金資金投資靈活。第二，它為 IBM 或乘數效應提供了非常堅實的現有基礎。但如果你看一下，我們正在利用我們在大型機週期的最後三個專案中看到的優勢。它確實體現了該平台的持久價值。

Our MIPS over the last few programs are up 3x from an install perspective and over 80% of our clients are growing MIPS on the mainframe. I think that was a very different picture when you dial back five, seven years to go already. So we've taken that portfolio. We've invested now significantly, which I'll come to around watsonx Code Assistant for Z, but we've taken that from a down mid-single-digit portfolio to now capitalizing on the stack economics of our mainframe execution and moved that to a low-single digit.

從安裝角度來看，我們過去幾個程式的 MIPS 提高了 3 倍，並且超過 80% 的客戶正在大型主機上增加 MIPS。我認為，當你回撥五、七年後，情況會非常不同。所以我們已經採取了該投資組合。我們現在已經進行了大量投資，我將圍繞 Z 的 watsonx Code Assistant 進行討論，但我們已經將其從中等個位數的投資組合轉變為現在利用大型主機執行的堆疊經濟性並將其轉移到低個位數。

Now for the year, as you heard, we are taking up our guidance, just given the strength of first half, to mid-single digit. When you get into 2025, we'll talk about our guidance going forward, but we feel very confident that we can continue growing this, and that's why we're investing and bringing out new capabilities, like watsonx Code Assistant for Z, which is resonating extremely well.

現在，正如您所聽到的，鑑於上半年的強勁勢頭，我們正在將今年的指引調整至中個位數。當你進入 2025 年時，我們將討論我們的未來指導，但我們非常有信心能夠繼續發展這一點，這就是我們投資並推出新功能的原因，例如 watsonx Code Assistant for Z，它是共鳴非常好。

Operator, let's take the next question.

接線員，我們來回答下一個問題。

Toni Sacconaghi, Bernstein.

東尼薩科納吉、伯恩斯坦。

Yes. Thank you for taking the question. I'm wondering maybe you can discuss how do you think about AI signings and whether you believe they're really incremental or just a shift in client spending? And part of the reason I ask the question is if I -- it looks like your AI book of business was up about $1 billion sequentially. You're saying three quarters of that is consulting, such $700 million-plus in consulting signings in the quarter. If I take that out, your book to bill from the rest of your business is actually down. And despite the strong signings, you're lowering your consulting expectations for the year.

是的。感謝您提出問題。我想知道也許你可以討論一下你如何看待人工智慧簽約，以及你是否認為它們真的是增量或只是客戶支出的轉變？我問這個問題的部分原因是——看起來你們的人工智慧業務帳簿連續成長了約 10 億美元。你是說其中四分之三是諮詢，本季諮詢簽約金額超過 7 億美元。如果我把它去掉，你的其他業務的帳本收入實際上就會下降。儘管簽約人數眾多，但您還是降低了今年的諮詢預期。

So I am just wondering, do you think AI investments in consulting are a shift in spending or do you think are accretive or do you actually think they could even be cannibalistic to consulting spend, and more broadly, IT spend?

所以我只是想知道，您認為諮詢領域的人工智慧投資是支出的轉變，還是您認為是增值的，或者您實際上認為它們甚至可能會蠶食諮詢支出，更廣泛地說，是 IT 支出？

So Tony, let me start and then Jim will add more color on this topic. First, it's a great question, and you laid out some of the dynamics that were going on in there. If we just step back and just look at our comments on the macroeconomic environment, we stated that there is discretionary spend pressure in consulting.

東尼，讓我開始，然後吉姆將為這個主題添加更多色彩。首先，這是一個很好的問題，你列出了其中發生的一些動態。如果我們退後一步，看看我們對宏觀經濟環境的評論，我們會指出諮詢領域存在可自由支配的支出壓力。

When you do have that pressure or there is a demand for AI, I would look to the eye and say, probably, the bulk of that demand, not all, but the bulk is indeed a shift from other areas of consulting. We don't actually believe it's cannibalistic to the point you're pointing out.

當你確實有這種壓力或對人工智慧有需求時，我會看著眼睛說，可能是大部分需求，而不是全部，但大部分確實是來自其他諮詢領域的轉變。我們其實並不認為它是同類相食到你所指出的程度。

Now as time goes on and as people move from early experimentation and proving out the value to wanting to scale and really get the full benefits of generative AI, we do actually believe, at that point, even for consulting, these will turn into accretive and additive. But we are still some time away from when that will happen. That is just to give you some color and acknowledging that the bulk but not all is a shift, Jim?

現在，隨著時間的推移，隨著人們從早期實驗和證明價值轉向想要擴展並真正獲得生成式人工智慧的全部好處，我們確實相信，到那時，即使是諮詢，這些也將變成增值和添加劑。但距離實現這一目標還需要一段時間。這只是為了給你一些色彩，並承認大部分但不是全部都是轉變，吉姆？

Yeah. Thanks, Tony, for the question. Just building on what Arvind said. I mean, first of all, we're very pleased with the early momentum that we've gotten with our book of business around G&A, both on the technology side with our watsonx platform and now with our open innovation strategy around RHEL AI, OpenShift, Granite models, InstructLab, et cetera.

是的。謝謝托尼提出這個問題。只是以阿爾溫德所說的為基礎。我的意思是，首先，我們對圍繞 G&A 的商業手冊所獲得的早期勢頭感到非常滿意，無論是在技術方面的 watsonx 平台，還是現在圍繞 RHEL AI、OpenShift 的開放創新戰略、花崗岩模型、InstructLab 等等。

But let's just deep dive a little deeper into your question about consulting. Because I think when you look at consulting, first of all, why is it so important right now in the early part of a cycle? It's important because it's got to establish IBM consulting as a strategic provider of choice for enterprises as they're going through what we'd like to call Digital Transformation 2.0 with GenAI. Everyone is looking for who is going to be their strategic provider and partner.

但讓我們更深入地探討一下您關於諮商的問題。因為我認為當你考慮諮商時，首先，為什麼它在週期的早期階段如此重要？這很重要，因為必須將 IBM 諮詢打造成企業的首選策略供應商，因為它們正在經歷我們所謂的 GenAI 數位轉型 2.0。每個人都在尋找誰將成為他們的策略提供者和合作夥伴。

And I think a $1.5 billion -- over $1.5 billion book of business in the first 12 months, which by the way, is in excess of the ramp we saw play out with hybrid cloud in Red Hat, we're off to a pretty good start. Now to Arvind's point, in every technology shift, very different dynamics between the human capital base business and a product IP business. Human capital base business, we do see and we expected clients will shift and reprioritize spending. They're doing that now as they're driving large enterprise transformation projects, which is what our portfolio has been able to capture. And that's why you see nice acceleration and growth in our backlog up healthy at 5%. But to Arvind's point, we do think once you get through the early cycle, this is an incremental expansion of TAM that drives a long tail growth factor over time that has multiplied our opportunities for us.

我認為前 12 個月的業務賬面價值為 15 億美元——超過 15 億美元，順便說一句，這超出了我們在紅帽混合雲中看到的增長，我們的前景非常好開始。現在，阿文德指出，在每一次技術變革中，人力資本基礎業務和產品智慧財產權業務之間的動態截然不同。我們確實看到並且預計客戶將改變人力資本基礎業務並重新調整支出的優先順序。他們現在正在這樣做，因為他們正在推動大型企業轉型項目，而這正是我們的投資組合所能捕捉到的。這就是為什麼您會看到我們的積壓訂單數量以 5% 的速度健康成長。但就 Arvind 的觀點而言，我們確實認為，一旦經歷了早期週期，這就是 TAM 的增量擴張，隨著時間的推移，它會推動長尾成長因素，從而使我們的機會倍增。

So when you look at our consulting book of business, let's dive into the subsegments. You see business transformation services, which a lot of the GenAI plays out too early right now, that is how do you transform the way you operate, HR, finance, supply chain. We doubled and accelerated our growth quarter to quarter.

因此，當您查看我們的業務諮詢手冊時，讓我們深入了解各個細分市場。你會看到業務轉型服務，很多 GenAI 現在推出得太早了，這就是你如何改變你的營運方式、人力資源、財務、供應鏈。我們的成長速度逐季翻倍並加速。

What you're seeing is a reprioritization in dynamic spending decisions by clients because our AO where we have a lot of short-term discretionary staff augmentation work, there's a lot of trade-offs between those two. So it's important for us strategically what our client base, but I think you see how it plays out.

您看到的是客戶對動態支出決策的優先順序進行了重新排序，因為我們的行政辦公室有很多短期可自由支配的人員擴充工作，這兩者之間有很多權衡。因此，從策略上講，我們的客戶群對我們來說很重要，但我想您已經看到了它的結果。

Now just to wrap up, the full picture, software, I think, is fundamentally different. Our software book of business now has $1 billion through the first 12 months. I think inception to date right now, we're about two-thirds subscription SaaS, one-third perpetual. I think that's contributing nicely about a point of growth, and by the way, that's one of the two components of why we took our software up for the year. So I think that's predominantly all lift.

現在總結一下，我認為整個情況，軟體是根本不同的。目前，我們的軟體業務在前 12 個月就達到了 10 億美元。我認為從成立至今，我們大約有三分之二是訂閱型 SaaS，三分之一是永久型。我認為這對成長點做出了很好的貢獻，順便說一句，這是我們今年升級軟體的兩個組成部分之一。所以我認為這主要是升力。

Operator, let's take the next question.

接線員，我們來回答下一個問題。

Amit Daryanani, Evercore ISI.

阿米特·達裡亞納尼 (Amit Daryanani)，Evercore ISI。

Thanks for taking my question. I guess, my question is really on the consulting side. And when I think about this business growing low-single digits for '24, if I take out some of the M&A contribution, also some of the revenue from the AI book of bill -- book of business that you have at $1.5 billion, is it fair to think that maybe the non-AI consulting fees actually gets worse in H2 versus H1 for you? If you just talk about the puts and takes on the back half consulting expectations versus front half, that would be really helpful.

感謝您提出我的問題。我想，我的問題確實是在諮詢方面。當我想到這個業務在 24 年以低個位數增長時，如果我去掉一些併購貢獻，還有來自人工智能賬簿的一些收入——你擁有 15 億美元的業務賬簿，公平地認為，也許對你來說，下半年的非人工智慧諮詢費實際上比上半年更糟？如果你只談論後半部分諮詢預期與前半部分的看跌期權和承付金額，那將非常有幫助。

And then I'm curious have you talked to your customers, what is your sense on the duration of this weakness in consulting? And when do you think it starts to come back? Thank you.

然後我很好奇您是否與您的客戶交談過，您對諮詢業這種弱點的持續時間有何看法？您認為它什麼時候開始恢復？謝謝。

Hi. So Amit, let me just start and maybe address the second part of your question first and then -- I actually do not believe there's any secular macro trend around weakness. I think that this is a temporal based on a number of factors we have. The geopolitical uncertainty has gone longer than most people expected, and that weighs into people's heads for what that might happen and specifically the war in Europe as well as the war in the Middle East. Second, inflation has gone longer than people expected, which has the unfortunate consequence of higher interest rates and that begins to bear on people.

你好。所以阿米特，讓我開始，也許先解決你問題的第二部分，然後——我實際上不相信疲軟存在任何長期宏觀趨勢。我認為這是一個基於我們擁有的許多因素的時間。地緣政治不確定性持續的時間比大多數人預期的要長，這讓人們對可能發生的事情，特別是歐洲戰爭和中東戰爭產生了影響。其次，通貨膨脹持續的時間比人們預期的要長，這帶來了利率上升的不幸後果，並開始對人們產生影響。

If I look at those two all altogether and that then the moment you have higher interest rates and inflation, you have wage inflation, which does impact the bottom line of our clients, you put all of that into perspective. And is this going to go on for another six months? Likely. Is it going to go on for another year? I'm not so sure. But we got to get to the second half to be able to go there.

如果我把這兩者放在一起看，那麼當利率和通貨膨脹更高時，工資就會上漲，這確實會影響我們客戶的底線，你要把所有這些都放在正確的位置上。這種情況還會持續六個月嗎？有可能。難道還要再持續一年嗎？我不確定。但我們必須到下半場才能到達那裡。

So that is why we are optimistic about the medium- and long-term vector on consulting. And as Jim answered in a prior question, we do see that this is going to become a tailwind over time, at least for us.

這就是為什麼我們對諮詢業的中長期方向持樂觀態度。正如吉姆在之前的問題中回答的那樣，我們確實看到隨著時間的推移，這將成為一股順風，至少對我們來說是這樣。

Now in the short term for the next six months, we do think it holds up a little bit. In terms of answering the specifics and decomposing some of the numbers that you laid out in the first part of your question, I'm going to turn that over to Jim.

現在，在未來六個月的短期內，我們確實認為它有一點支撐。在回答具體問題並分解您在問題第一部分列出的一些數字方面，我將把它交給吉姆。

Yes. Thanks, Arvind, and thanks, Amit, for the question, overall. Let's put this in perspective, right? You go back 90 days ago; how did we see the year playing out with consulting? We said -- at that point in time, we had backlog grown nicely mid-single digit, albeit we did talk about durations going up because large-scale transformations were really where the spend was moving to. But we had a solid book-to-bill trailing 12 months over 1.15. We had GenAI momentum that can continue throughout the year early in the cycle. We had strategic partnerships, Red Hat growth profile, and we had future acquisitions as we're going to continue to be opportunistic around our M&A criteria and the synergistic value of how consulting plays to our portfolio.

是的。總的來說，謝謝阿溫德，也謝謝阿米特的問題。讓我們客觀地看待這個問題，對嗎？你回到90天前；我們如何看待這一年的諮詢業務？我們說，在那個時間點，我們的積壓工作成長得很好，達到了個位數中位數，儘管我們確實談到了持續時間的增加，因為大規模轉型確實是支出的轉移方向。但我們的訂單出貨比連續 12 個月都超過 1.15。我們的 GenAI 勢頭可以在周期初期持續一整年。我們擁有策略夥伴關係、紅帽成長概況，我們還有未來的收購，因為我們將繼續圍繞我們的併購標準以及諮詢如何對我們的投資組合發揮協同價值。

If you look right now, 90 days later, as we look to the second half, many of those are still playing out. You've got GenAI, which arguably were above our own expectations, right now, doubling by the way in consulting. Our GenAI book of business quarter to quarter, strategic partnerships, especially hyperscalers, Red Hat still growing nicely.

如果你現在看，90 天后，當我們展望下半年時，其中許多仍在進行中。 GenAI 可以說超出了我們自己的預期，目前，諮詢業務翻了一番。我們的 GenAI 業務季度報告、策略合作夥伴關係，尤其是超大規模企業、紅帽仍然成長良好。

What you're seeing at the end of the day, those are large-scale transformations, lower yield. That's why Arvind and I are saying these are longer0term growth vectors and tails that will play out into '25, '26, and beyond as we get that strategic provider of choice.

最終你會看到，這些都是大規模的轉變，產量較低。這就是為什麼阿溫德和我說這些是長期成長向量和尾巴，當我們選擇策略提供者時，它們將持續到 25 年、26 年及以後。

But in the interim, what you're seeing is that spending re-prioritization around short-term discretionary that I think everyone in the industry is talking about. We're all dealing with this. The key is we have to win that strategic provider of choice in GenAI and I would argue we're off to a great start. You look at competitor numbers overall. We got a $1.5 billion -- over $1.5 billion book of business, doubling quarter to quarter right now. I think we're in pretty good shape. That's what we're focused on because that will provide future revenue multiplier effect as we move forward.

但在此期間，你會看到圍繞短期可自由支配的支出重新確定優先順序，我認為業內每個人都在談論這一點。我們都在處理這個問題。關鍵是我們必須贏得 GenAI 的策略供應商選擇，我認為我們已經有了一個好的開始。您會查看競爭對手的總數。我們的業務帳面價值為 15 億美元——超過 15 億美元，目前每季都翻了一番。我認為我們的狀態非常好。這就是我們關注的重點，因為隨著我們的前進，這將提供未來的收入乘數效應。

Operator, let's take the next place.

接線員，我們去下一個地方吧。

James Schneider, Goldman Sachs.

詹姆斯·施奈德，高盛。

Thanks for taking my question. Maybe if I could just ask on a different topic for a second. Maybe talk about the environment you see right now for M&A and your intention to continue to drive through acquisitions. And do you believe you have sufficient skill in open source and DevOps software, in particular? And can you maybe comment on the attractiveness of multiples in the public market today relative to the private market?

感謝您提出我的問題。也許我可以問一個不同的話題。也許可以談談您現在看到的併購環境以及您繼續推動收購的意圖。您是否認為自己在開源和 DevOps 軟體方面擁有足夠的技能？您能否評論一下當今公開市場相對於私人市場的本益比的吸引力？

Jim, a great question, and thank you for asking this. Look, on overall M&A, I just want to begin with that our strategy has not changed. We are disciplined and we are focused. By focused, I mean, we stick to the areas that we are investing in, hybrid cloud and artificial intelligence. And by discipline, I mean, it has to be not just aligned to our strategy, but we expect synergy from the acquisition, especially the multiples are higher, as you pointed out. And it has to be accretive to free cash flow. If it's a larger, definitely, within two years at the outer end of the range.

吉姆，這是一個很好的問題，謝謝你提出這個問題。就整體併購而言，我只想先指出我們的策略沒有改變。我們遵守紀律，並且專注。我所說的專注，是指我們堅持投資的領域，混合雲端和人工智慧。我的意思是，它不僅必須與我們的策略保持一致，而且我們期望透過收購產生協同效應，尤其是倍數更高，正如您所指出的。它必須能夠增加自由現金流。如果它更大，那麼肯定會在兩年內處於該範圍的外端。

So having said that, [if I look at it] right now, we have HashiCorp out there. So we've got to get through that. We expect that to happen in the second half of this year. We just finished StreamSets and webMethods, and we've done a couple of smaller ones in the in the consulting space and in other technology tuck-ins.

話雖如此，[如果我看一下]現在，我們就有了 HashiCorp。所以我們必須解決這個問題。我們預計這將在今年下半年發生。我們剛完成了 StreamSets 和 webMethods，並且我們也在諮詢領域和其他技術領域完成了幾個較小的專案。

What do we see going into the space? Our valuation is rich. They're reasonably rich. They're not outrageous, I would say, like they had become in parts of late 2020 and 2021. So I would say that they are more reasonable than then, but they are richer than they were about 18 months ago.

我們在太空中看到了什麼？我們的估值很豐富。他們相當富有。我想說，他們並不像 2020 年底和 2021 年的部分地區那樣令人髮指。

There are different dynamics in both the public and the private markets. Public markets are quite variable. I mean, as we can see, some of the multiples and if you look at multiples to revenue, which is not a great metric, let me just acknowledge that, but it is one that's out there. If you look at 6, 7, 8, maybe 9 or 10 times, we can see our way there for a large deal as long as we have sufficient synergy.

公共和私人市場都有不同的動態。公共市場變化很大。我的意思是，正如我們所看到的，一些倍數，如果你看看收入的倍數，這不是一個很好的指標，讓我承認這一點，但它是存在的。如果你看6次、7次、8次，也許9次或10次，只要我們有足夠的協同作用，我們就可以找到解決大問題的方法。

Now for very small deals, that's not even a fair multiple. Very small deals are all about technology and people. In the private markets, we were very pleased with what we got done on StreamSets and webMethods. I would call that a private market deal, not a public market deal. And there, I think it all depends upon what's the property. What is its growth profile? What is the attractiveness of it to the seller versus supplier, in this case us? All of that play to those multiples.

現在對於非常小的交易來說，這甚至不是一個公平的倍數。非常小的交易都與技術和人員有關。在私募市場，我們對 StreamSets 和 webMethods 所做的工作感到非常滿意。我將其稱為私人市場交易，而不是公開市場交易。在那裡，我認為這完全取決於財產是什麼。其成長概況如何？它對賣方與供應商（在本例中為我們）有何吸引力？所有這些都對這些倍數起作用。

I do expect that on the private side, valuations will be slightly less, but then the risk of going public or some other exit is also taken away. And in some sense, you get a discount for taking that risk off the table. People who are venture backed, that's different. They are looking at IPO versus strategic exit and those are different multiples.

我確實預計，在私人方面，估值會略低，但上市或其他退出的風險也會被消除。從某種意義上說，您可以透過消除這種風險來獲得折扣。得到創投支持的人則不同。他們正在考慮 IPO 與戰略退出，它們的倍數不同。

But putting all of that together, we remain in the market. M&A is an important part of our growth methodology. We maintain a strong balance sheet for that purpose, and we have kind of been clear of that. All that said, this year we got a big one coming. So we want to wait and get that done because part of the discipline is also making sure that we digest them at the right rate and pace and put them into our global go-to-market distribution engine.

但把所有這些放在一起，我們仍然留在市場上。併購是我們成長方法的重要組成部分。為此，我們保持了強勁的資產負債表，而且我們已經明確了這一點。話雖如此，今年我們迎來了一件大事。因此，我們希望等待並完成這項工作，因為紀律的一部分還確保我們以正確的速度和節奏消化它們，並將它們放入我們的全球上市分銷引擎中。

Operator, let's take the next question.

接線員，我們來回答下一個問題。

Ben Reitzes, Melius Research.

本‧雷茨 (Ben Reitzes)，Melius 研究中心。

Yeah. Hey, thank you. Appreciate it.

是的。嘿，謝謝你。欣賞它。

Jim, I wanted to -- and Arvind, I wanted to see if -- it sounds like the margin progress is sustainable for the year. So while I appreciate that you guide to free cash flow and you've raised it a little bit, do you anticipate us being able to flow through the $0.25 upside on the EPS line? And can -- does that mean earnings as sustainable in the back half? And then I was just wondering if you have any more info on HashiCorp, in terms of the revenue contribution? Street was looking for about [750] in revenue next year. And on the dilution, there's -- it should be a loss of around $0.30 in interest income. So just wondering if you have any further views on the net effect to 2025 on that deal. Thanks so much, guys

吉姆，我想——阿文德，我想看看——聽起來今年的利潤成長是否可持續。因此，雖然我很欣賞您對自由現金流的指導，並且您已經將其提高了一點，但您是否預計我們能夠透過每股收益線的 0.25 美元上漲？這是否意味著下半年的獲利可持續？然後我想知道您是否有更多有關 HashiCorp 的收入貢獻資訊？ Street 預計明年的收入約為 [750]。在稀釋方面，利息收入應該損失約 0.30 美元。所以想知道您對該交易到 2025 年的淨影響是否有任何進一步的看法。非常感謝，夥伴們

Hey, Ben. Thank you. Appreciate it. Very good question, overall. Let's take a step back on your first part of the question around free cash flow. Yes, we're very pleased with the start of the year. Free cash flow of $4.5 billion, up $1.1 billion year to year, 4 points above historical attainment, so our largest first-half free cash flow generation as far back as I can go and count.

嘿，本。謝謝。欣賞它。總體而言，非常好的問題。讓我們退一步討論關於自由現金流的問題的第一部分。是的，我們對今年的開始感到非常滿意。自由現金流為 45 億美元，年增 11 億美元，比歷史水準高出 4 個百分點，這是我能統計到的上半年自由現金流產生量最大的數字。

So we're off to a pretty good start. And that gives us the confidence overall of how we're positioned in second half. But the second half and why we took the guidance up is entirely driven by the strength of the fundamentals of our business and flowing through the adjusted EBITDA overachievement. So read that. Although we don't guide on EPS, the strong overachievement of the $0.25 of EPS, we're flowing that through to adjusted EBITDA and that flows through to our guide take-up on free cash flow. The rest of the free cash flow dynamics we've been talking about all year long around, yes, we got benefits of change in retirement plans and cash tax that's going to be a headwind and other balance sheet items, none of that changes.

所以我們有了一個很好的開始。這讓我們對下半年的定位充滿信心。但下半年以及我們提高指引的原因完全是由我們業務基本面的實力以及調整後的 EBITDA 超額業績所驅動。所以讀一下吧。儘管我們不以 EPS 為指導，但每股收益 0.25 美元的強勁超額表現，我們將其轉化為調整後的 EBITDA，並轉化為我們對自由現金流的指導。我們全年都在談論的自由現金流動態的其餘部分，是的，我們得到了退休計劃和現金稅變化的好處，這將成為逆風和其他資產負債表項目，但這些都沒有變化。

One thing I will bring up and we said in the prepared remarks, but just so there's absolute clarity, we do expect to close the Palo Alto transaction here in the third quarter around certain assets of our QRadar business that will obviously generate a gain. We're excited about the new strategic relationship between our two great companies overall, but we will take structural actions to offset that gain to address stranded costs and all by the way to your second part of your question to accelerate our productivity initiatives in 2025.

我將提出一件事，我們在準備好的發言中說過，但為了絕對明確，我們確實預計在第三季度圍繞我們 QRadar 業務的某些資產完成帕洛阿爾托交易，這顯然會產生收益。總體而言，我們對兩家偉大公司之間的新策略關係感到興奮，但我們將採取結構性行動來抵消這一收益，以解決擱淺成本問題，並順便回答您問題的第二部分，以加快我們在2025 年的生產力計劃。

So you get the HashiCorp, first of all, the strategic transaction stands on its own. Arvind went through our M&A criteria. I think there's a very compelling strategic fit around end-to-end leadership, [hybrid car] platform. There's a lot of synergistic value both on product technology and go to market, but there's a very attractive financial profile that we talked about 90 days ago, higher revenue growth profile, adjusted EBITDA accretive in 12 months, free cash flow accretive, to Arvind's point, by two years. And we do see potential significant near-term costs in operating synergies that lead to about 30% to 40% free cash flow margin business over a handful of years.

所以你得到了 HashiCorp，首先，戰略交易是獨立的。 Arvind 詳細了解了我們的併購標準。我認為圍繞端到端領導力、[混合動力車]平台存在著非常引人注目的策略契合。產品技術和上市方面都有很多協同價值，但就Arvind 的觀點而言，我們在90 天前討論過的財務狀況非常有吸引力，收入增長狀況更高，調整後EBITDA 在12 個月內增加，自由現金流增加，到兩年。我們確實看到了營運綜效中潛在的巨大短期成本，這會導致幾年內業務自由現金流利潤率達到約 30% 至 40%。

Now when you look at dilution, we understand dilution. I mean, M&A has been an integral part of our financial model for decades. So underneath that, we understand the purchase growth of those transactions, the synergies of those transactions, the balance sheet capital structure implications of those. And with all that said, our model is to grow mid-single digit revenue and grow operating leverage, so we grow free cash flow quicker than revenue. We don't see that changing in 2025.

現在當你看到稀釋時，我們就理解了稀釋。我的意思是，幾十年來，併購一直是我們財務模式不可或缺的一部分。因此，在此之下，我們了解這些交易的購買成長、這些交易的綜效以及這些交易對資產負債表資本結構的影響。話雖如此，我們的模式是實現中個位數收入成長並提高營運槓桿，因此我們自由現金流的成長速度快於收入。我們認為這種情況在 2025 年不會改變。

We see growth profiles around revenue, around operating leverage, and around free cash flow overall. And that speaks to the diversity or diversification, I should say, of our business model around productivity. We entered the year, raised it to $3 billion. We're getting out ahead of that again. And you see that play out in our margins through the first half, up 180 basis points on pretax. So we've got many levers to deal with this overall. We know how to handle it.

我們看到圍繞著收入、營運槓桿和整體自由現金流的成長概況。我應該說，這說明了我們圍繞生產力的商業模式的多樣性或多樣化。我們進入了這一年，將其提高到 30 億美元。我們又要提前出去了。你可以看到我們上半年的利潤率得到了體現，稅前成長了 180 個基點。因此，我們有很多手段來解決這個問題。我們知道如何處理。

Operator, let's take the next question.

接線員，我們來回答下一個問題。

Erik Woodring, Morgan Stanley.

艾瑞克‧伍德林，摩根士丹利。

Hey, guys. Thanks so much for taking my question. Arvind or Jim, if you could just dig into the Red Hat business a bit more. Over the last few quarters, you've talked about some very healthy bookings growth numbers ranging anywhere between, call it, 15% and 20%. But we did see growth, obviously decelerate by about a point this quarter despite expectations that it would be flat to maybe increasing for the rest of the year.

大家好。非常感謝您提出我的問題。 Arvind 或 Jim，如果您能更深入地了解紅帽業務的話。在過去的幾個季度中，您談到了一些非常健康的預訂成長數字，範圍在 15% 到 20% 之間。但我們確實看到了成長，儘管預計今年剩餘時間成長將持平甚至可能成長，但本季明顯放緩約一個百分點。

So can you just kind of double-click on exactly what you're seeing with the Red Hat business today? What's kind of the offset to the strong bookings numbers? And how should we think about Red Hat growth now in constant currency for 2024? Thanks so much.

那麼，您能雙擊一下您今天所看到的紅帽業務嗎？強勁的預訂量有何抵消作用？我們該如何看待 2024 年以固定匯率計算的紅帽成長？非常感謝。

Great question, Erik. So let's just look at the Red Hat business in terms of how the dynamics function between our clients and ourselves. So clients come and create demand. We fulfill that. That shows up as bookings not as revenue because the Red Hat business model is a pure consumption business model.

好問題，埃里克。因此，讓我們從我們的客戶和我們自己之間的動態運作的角度來看看紅帽業務。所以客戶來創造需求。我們實現了這一點。這顯示為預訂量而不是收入，因為紅帽商業模式是純粹的消費商業模式。

Clients pay for what they're consuming. And so the bookings then play out. Now those bookings are a signal of further demand, and typically, they are anywhere from one to three years' worth of revenue that the client is pre-committing to.

客戶為他們所消費的東西付費。這樣預訂就完成了。現在，這些預訂是進一步需求的信號，通常，它們相當於客戶預先承諾的一到三年的收入。

So when we enter a year, about half the revenue, we can look at the bookings over the previous year and say that that gives us. The other half has to come over the quarter. Now that we have a year, not longer but a year of the double-digit demand that you're talking about, if I remember, right, it was 14%, 17%, 14%, 20% in terms of the demand, now that that full year is there, that points to that for the portion that we can see. And as we get into a quarter, it climbs up on that 50% to 60% to 70% to 80%.

因此，當我們進入一年時，大約有一半的收入，我們可以查看上一年的預訂量並說這給了我們。另一半必須在本季完成。現在我們有一年，不是更長，而是你所說的兩位數需求的一年，如果我沒記錯的話，就需求而言，是 14%、17%、14%、20%，現在全年已經到來，這顯示我們可以看到的部分。當我們進入一個季度時，它會攀升到 50% 到 60% 到 70% 到 80%。

And Jim mentioned in his prepared remarks, what he called, cRPO or the revenue performance obligations, we see those sitting around mid-10s for the second half of the year, to answer your questions. Now if that's about 80% and that will translate into low-double digit is what we can look at and feel quite comfortable on. By the way, we see these early signs of the demand continue into this quarter, and likely the half, which means that we expect to continue now in the low-double digits, going forward. So I hope that gives you a sense.

Jim 在準備好的演講中提到，他所說的 cRPO 或收入績效義務，我們看到那些在下半年坐在 10 左右的人來回答你的問題。現在，如果這個數字約為 80%，並且將轉化為低兩位數，那麼我們就可以看到並感到非常滿意。順便說一句，我們看到這些需求的早期跡象持續到本季度，甚至可能是下半季度，這意味著我們預計未來將繼續保持低兩位數。所以我希望這能讓你有所了解。

But I'm also excited by the underlying product capabilities. We see OpenShift, which is extremely important. It plays into containerization. It plays into virtualization. It's an important element of our clients exercise hybrid. It has been growing and the demand there grew again at about 40% this past quarter. But we also saw acceleration in Linux and Ansible, where both of those demand vectors grow into the low-double digits. That, given the size of the Linux business, is very good news for us going forward.

但我也對潛在的產品功能感到興奮。我們看到OpenShift，這是極其重要的。它參與了容器化。它發揮了虛擬化的作用。這是我們客戶混合鍛鍊的一個重要元素。它一直在成長，上個季度需求再次成長約 40%。但我們也看到了 Linux 和 Ansible 的加速發展，這兩個需求向量都成長到低兩位數。考慮到 Linux 業務的規模，這對我們未來的發展來說是一個非常好的消息。

So I hope that gives you some color on those pieces. And a vector that we have not talked about that will play, but probably into '25 and '26, we are very excited by our two open source AI projects inside the Red Hat business, RHEL AI as well as OpenShift AI. And as people begin to deploy at scale, but not only on public cloud, but also on-premise leveraging their hybrid environments, we expect that both of those will also contribute to the Red Hat business, but that will take more time.

所以我希望這能給你一些關於這些作品的色彩。我們還沒有談到的一個向量將會發揮作用，但可能會在 25 和 26 年開始發揮作用，我們對紅帽業務內的兩個開源 AI 專案 RHEL AI 和 OpenShift AI 感到非常興奮。隨著人們開始大規模部署，不僅在公有雲上，而且在本地利用其混合環境，我們預計這兩者也將為紅帽業務做出貢獻，但這將需要更多時間。

Operator, let's take one last question.

接線員，我們來回答最後一個問題。

Matt Swanson, RBC Capital Markets.

馬特‧斯旺森，加拿大皇家銀行資本市場部。

Thank you. Yeah, Arvind, if we could pick up right where you left off there, can you just give us a little more color on the decision to open up the Granite models and the code base? And then really, what you're seeing in the market that makes you feel like taking maybe a more developer-focused approach to those, I think, as you put it, fit-for-purpose models is the right long-term strategy?

謝謝。是的，Arvind，如果我們可以從您上次停下的地方繼續，您能否就開放 Granite 模型和程式碼庫的決定向我們提供更多資訊？然後，實際上，您在市場上看到的情況讓您覺得可能採取一種更加以開發人員為中心的方法來解決這些問題，我認為，正如您所說，適合用途的模型是正確的長期策略？

So Matt, thank you for asking that question. And there was actually a question on developers before also. So I'm sorry, we didn't get to it fully. We'll get to it now in this question. Look, the question comes down to there was a thesis out there about 1.5 year ago that maybe one, maybe two extremely large models are going to run away with the bulk of the market share. We always felt that was both technically and economically infeasible. And I'll describe why.

馬特，謝謝你提出這個問題。之前其實也有一個關於開發人員的問題。所以很抱歉，我們沒有完全了解它。我們現在就這個問題來討論它。看，問題歸結為大約 1.5 年前有一篇論文認為，也許一個、也許兩個非常大的型號將奪走大部分市場份額。我們總覺得這在技術上和經濟上都是不可行的。我將描述原因。

If you run an extremely large model on public cloud, the model by its nature is going to be expensive because a very large model needs a lot more compute, a lot more network, a lot more storage, a lot more memory, and we can see some of those dynamics play out.

如果您在公有雲上運行一個非常大的模型，那麼該模型本質上會很昂貴，因為一個非常大的模型需要更多的計算、更多的網路、更多的儲存、更多的內存，我們可以看看其中一些動態的發揮。

If you can drop the model size, you can drop all of that by 90%. I would actually tell you 99% reduction in the compute and memory and network costs. But let's call it 90%, just for the sake of argument. So if you are running -- like one of our clients was describing to me, they run a couple of billion transactions through their internal systems each day.

如果您可以縮小模型尺寸，則可以將所有尺寸縮小 90%。實際上，我會告訴您計算、記憶體和網路成本降低了 99%。但為了方便討論，我們稱之為 90%。因此，如果您正在運行 - 就像我們的一位客戶向我描述的那樣，他們每天透過內部系統運行數十億筆交易。

If they have to go service those out to a large public cloud, the bill per day would have come back to be a couple of hundred million dollars. You multiply that by 250, that's kind of an infeasible cost. If you can drop it by 90%, you're now bringing it down to $10 million to $20 million a day. If you can actually run it using some of our Red Hat technologies on-premise, you can drop it by another 50%. You're not talking $5 million to $10 million. For what it can do, that is a very attractive proposition.

如果他們必須將這些服務轉移到大型公有雲上，那麼每天的費用將達到數億美元。你把它乘以 250，這是不可行的成本。如果你能降低 90%，那麼你現在每天的花費就可以降低到 1000 萬到 2000 萬美元。如果您實際上可以在本地使用我們的一些紅帽技術來運行它，則可以再降低 50%。你說的不是 500 萬到 1000 萬美元。就其功能而言，這是一個非常有吸引力的提議。

So now getting back to the models, if you have no idea what you're going to do, if you have no idea of what you might be looking for, you go to a very large model because it contains all the possible elements. If you have a sense of what you need to do -- I need to summarize emails, you need English language model, if we're sitting here in the United States. If you're going to go change your Java or C++ or Python program as to be more productive, you don't need a model that can write poetry and draw images. You need a model that understands programming languages.

現在回到模型，如果您不知道要做什麼，如果您不知道要尋找什麼，那麼您會使用一個非常大的模型，因為它包含所有可能的元素。如果你知道你需要做什麼——我需要總結電子郵件，你需要英語語言模型，如果我們坐在美國。如果您打算更改 Java、C++ 或 Python 程式以提高工作效率，則不需要可以寫詩和畫圖的模型。您需要一個理解程式語言的模型。

So we are very, very proud of what our team has done. We can produce models that can do these things. So these are two distinct models. One for programming. One for business language. They're one-tenth or less than that of the size of the extremely large models. But you can look on the leaderboard, they perform quite as well as the largest models. So that is what our strategy is.

因此，我們對我們團隊所做的事情感到非常非常自豪。我們可以生產可以做這些事情的模型。所以這是兩個不同的模型。一種用於編程。一種用於商務語言。它們的尺寸只有超大型模型的十分之一甚至更小。但你可以看看排行榜，它們的表現與最大的模型相當。這就是我們的策略。

However, if our clients want other models, we are also happy to work with other models, and we have had that perspective. So why open source, since that is part of your question? Why open source is because often, we find that clients want to increase the models efficacy by adding their own unique language. People might want to write e-mails in a certain way. They might want a program in a certain way. They like comments in a certain style. I call that refining the model.

然而，如果我們的客戶想要其他模型，我們也很樂意與其他模型合作，我們也有這樣的觀點。既然這是你問題的一部分，那為什麼要開源呢？為什麼開源是因為我們經常發現客戶希望透過添加自己獨特的語言來提高模型的效率。人們可能想以某種方式編寫電子郵件。他們可能想要某種特定方式的程式。他們喜歡某種風格的評論。我稱之為精煉模型。

We have a technique called InstructLab, but then clients get concerned. Wait, if I add my data, I don't want to give that away and back into a more public format, can I keep that to myself? So open sourcing our models under the Apache license gives our clients the freedom that what they add onto our underlying open model, they can keep to themselves. Now to the developer point, putting all of that machinery into Red Hat Linux now gives us an avenue to open it up to developers that they can go experiment and play.

我們有一種稱為 InstructLab 的技術，但客戶會擔心。等等，如果我添加我的數據，我不想將其洩露並返回到更公開的格式，我可以將其保留給自己嗎？因此，在 Apache 許可下開源我們的模型為我們的客戶提供了自由，他們可以將自己添加到我們底層開放模型中的內容保留給自己。現在到了開發人員的角度，將所有這些機制放入 Red Hat Linux 中，現在為我們提供了一條向開發人員開放的途徑，讓他們可以進行實驗和使用。

By the way, I'll turn around and tell you that for a developer who is not running production, who is just playing with things, like all people do it. On a MacBook, you can begin to play around with models that are in the low tens of billions of parameters. That's a massive market that opens up. They get the freedom and flexibility that they don't have to give it back to us unless they want to.

順便說一句，我會轉過身來告訴你，對於一個不運行生產的開發人員來說，他們只是在玩東西，就像所有人一樣。在 MacBook 上，您可以開始嘗試具有數百億參數的模型。這是一個正在開放的龐大市場。他們獲得了自由和靈活性，除非他們願意，否則他們不必將其還給我們。

I am not actually concerned about this gives away the IP. As we have found through, whether it's Red Hat Linux or whether other people have found through Mongo or other people have found through [HODU], enterprises do look for -- and the last few days have certainly shown us, people look for patching, people look for security, people look for backward compatibility. There has a lot of enterprise reasons why people will still do business with us for the open-source nature of what you ask. I'm so glad you did.

我實際上並不擔心這會洩露 IP。正如我們所發現的那樣，無論是 Red Hat Linux，還是其他人透過 Mongo 發現的，或者其他人透過 [HODU] 發現的，企業確實在尋找——過去幾天確實向我們表明，人們在尋找補丁，人們尋求安全性，人們尋求向後相容性。由於您所要求的開源性質，人們仍然會與我們開展業務有很多企業原因。我很高興你這麼做了。

That allows us to expand that market into the millions of developers who do run Linux on their home machines or their corporate machines or their laptops, and they can go experiment, add their innovation, and either give it back to the community or actually reserve it for their enterprise. So that's how we tap into the whole developer ecosystem.

這使我們能夠將該市場擴展到數百萬在家用電腦、公司電腦或筆記型電腦上運行 Linux 的開發人員，他們可以進行實驗，添加他們的創新，然後將其返還給社區或實際上保留它為了他們的企業。這就是我們進入整個開發者生態系統的方式。

So let me now wrap up the call. In the second quarter of 2024, we executed on our strategy to deliver revenue growth and cash generation. We saw strong performance across our portfolio. We're excited about our early traction in generative AI. We look forward to sharing our progress with you as we move through the rest of the year. Thank you, all.

現在讓我結束這通通話。 2024 年第二季度，我們執行了實現收入成長和現金產生的策略。我們的投資組合表現強勁。我們對產生人工智慧的早期發展感到興奮。我們期待在今年剩下的時間與您分享我們的進展。謝謝你們。

Thank you, Arvind. Operator, let me turn it back to close out the call.

謝謝你，阿爾溫德。接線員，讓我將其轉回以結束通話。

Thank you for participating on today's call. The conference has now ended. You may disconnect at this time.

感謝您參加今天的電話會議。會議現已結束。此時您可以斷開連線。