Microsoft Corporation (MSFT.US) cutting data centers triggers panic, Goldman Sachs Group, Inc. reassures market: AI computing power demand still strong, $80 billion spending remains unchanged.

Recently, the news that the American technology giant Microsoft Corporation (MSFT.US) terminated two large data center lease agreements has sparked a global discussion in the stock market. This undoubtedly poured a bucket of cold water on the continuous global investment frenzy in artificial intelligence (AI) since 2023, leading investors to significantly reduce their bullish sentiment towards AI infrastructure construction, especially in the areas of AI chips, high-performance network hardware devices, and power equipment. However, Wall Street financial giant Goldman Sachs Group, Inc. swiftly released a research report to stabilize market sentiment after the news, emphasizing that Microsoft Corporation's capital expenditure of up to 80 billion dollars will not be cut to any degree, and reiterating a "buy" rating on Microsoft Corporation stock with a target price of 500 dollars. Despite recent rumors that Microsoft Corporation adjusted its AI data center leasing plans causing market fears of the AI frenzy subsiding, Goldman Sachs Group, Inc. believes that this essentially reflects Microsoft Corporation's cautious attitude towards AI investments. They prioritize high-return focused reasoning AI infrastructure projects, while shifting capital expenditure towards assets with shorter lifecycles, relatively lower costs, and cost-effectiveness (such as edge computing devices), which helps balance short-term costs with long-term benefits. Goldman Sachs Group, Inc. stated in the report that Microsoft Corporation has strong capabilities at all levels of cloud computing (applications, platforms, infrastructure), covering applications such as Copilot, GitHub Copilot, Microsoft 365 Copilot, and Azure A. This strengths at all levels of cloud computing undoubtedly enable it to fully benefit from the transfer of generative AI from the infrastructure layer to the platform and application layers, which will certainly be positive for Microsoft Corporation's long-term profitability. Goldman Sachs Group, Inc. also believes that if Microsoft Corporation continues to efficiently integrate its generative AI technology across product lines and generate strong business revenue, any delays or changes in data center lease agreements are unlikely to have a negative impact on Microsoft Corporation's AI-related business revenue. The Goldman Sachs Group, Inc. analysis team also believes that by optimizing capital expenditure structure and shifting towards high-profit margin AI inference layer business, Microsoft Corporation's future EPS growth is expected to accelerate again. Goldman Sachs Group, Inc. remains bullish on Microsoft Corporation, emphasizing that Microsoft Corporation is a big winner benefiting from the AI boom. With DeepSeek leading the "low-cost computational power new paradigm sweeping the world", the costs of AI training and application inference end are increasingly declining, while driving AI application software (especially generative AI software and AI agents) to penetrate various industries worldwide at an accelerated pace, fundamentally innovating efficiency of various business scenarios and significantly increasing revenue. Cloud computing and software giants Microsoft Corporation, Amazon.com, Inc., Oracle Corporation, Alibaba Group Holding Limited Sponsored ADR, and Tencent.AI will potentially see exponential growth in revenue and profits, becoming the biggest winners in the DeepSeek global explosion and the "DeepSeek low computational cost impact wave". If killer-level AI application software/AI agents start to emerge on a large scale from 2025, it will be a major positive for global cloud giants like Microsoft Corporation, Amazon.com, Inc. These AI software, whether in the early AI software developer ecosystem platform or the later massive cloud AI inference computational resources, rely on the powerful computational platform support provided by these cloud giants. These cloud giants focus on establishing B-end and C-end AI application software developer ecosystems related to generative AI, aiming to significantly reduce the technical barriers for non-IT professionals in various industries to develop AI applications. Unlike leaders in AI such as OpenAI, xAI, which provide AI chat Siasun Robot & Automation as well as AI agents, software companies like Microsoft Corporation, Amazon.com, Inc., ServiceNow, and Datadog have exclusive one-stop "AI application platforms" for their own software ecosystems. They will greatly benefit from this unprecedented global enterprise layout AI boom. They focus on establishing B-end and C-end AI application software developer ecosystems related to generative AI, aiming to significantly reduce the technical barriers for non-IT professionals in various industries to develop AI applications, and provide cloud AI training/inference computational resources based on NVIDIA Corporation AI GPU or self-developed AI chips to greatly simplify the one-stop deployment of AI large models based application software and provide AI inference computational resources to support AI workloads. Goldman Sachs Group, Inc. states that due to the data centers terminated in 2023 and 2024, there may be:The wave of heart construction or expansion is sweeping across the globe. Under the high base effect, the growth rate of capital expenditure related to artificial intelligence is inevitably slowing down. Generative AI is shifting from the infrastructure layer to SaaS platforms, as well as covering applications in various industries. As one of the few global super cloud computing service providers with widely commercialized applications, Microsoft Corporation is able to fully leverage this shift, which will undoubtedly be positive for Microsoft Corporation's long-term profitability.The analysis team of Goldman Sachs Group, Inc. added that Microsoft Corporation has strong market share in various areas of cloud computing, including applications, platforms, and underlying infrastructure. Therefore, Goldman Sachs Group, Inc. believes that Microsoft Corporation has the ability to capture some long-term and continuous growth trends, such as generative AI software, AI agents sweeping across various industries globally, public cloud consumption with SaaS, digital transformation, AI/ML, business intelligence/analytics, DevOps, etc. Furthermore, the Goldman Sachs Group, Inc. team emphasized in their research report that Microsoft Corporation, with its large user base (over 400 million business users), full-stack cloud computing services (IaaS/PaaS/SaaS), and generative AI tools (Copilot+ series), has a unique advantage in the AI application layer. Microsoft Corporation covers the complete AI technology stack from underlying computational infrastructure (Azure AI infrastructure) to upper-level AI applications (Copilot, GitHub Copilot). Combined with its enterprise customer base (e.g. RPO orders exceeding $300 billion), Microsoft Corporation has formed a "data-model-application" closed loop ecosystem that is difficult to replicate globally. Goldman Sachs Group, Inc. stated that as Microsoft Corporation's cloud computing business reaches an annual revenue scale of approximately $100 billion, Microsoft Corporation's operating leverage will continue to drive significant growth in earnings per share (EPS), with EPS expected to double by the fiscal year 2028. Goldman Sachs Group, Inc. reiterated its buy stock rating for Microsoft Corporation in the research report, with a 12-month target stock price of up to $500. In comparison, Microsoft Corporation's stock price closed at $397.900 on the U.S. stock market on Tuesday. Microsoft Corporation reaffirms its $80 billion spending plan to stabilize market confidence in its artificial intelligence business. According to recent reports from multiple media outlets, Microsoft Corporation has cancelled several lease agreements with multiple private data center operators, involving a total power capacity of several hundred megawatts. Among them, based on media tracking data, Microsoft Corporation has scrapped two large data centers in Wisconsin Kenosha and Georgia Atlanta. The former was due to Microsoft Corporation not participating in the Stargate Project, transferring Wisconsin Kenosha to the Stargate Project, while the latter was because Microsoft Corporation's data center team overestimated the AI computing power demand in the Atlanta area. In addition, Microsoft Corporation has also paused the conversion of negotiated and signed Statements of Qualifications (SOQs) into lease agreements. However, shortly after this, Microsoft Corporation quickly released a statement, stating that the cancellation or delay of data center leases would not have any impact on Microsoft Corporation's capital expenditure target of up to $80 billion for the current fiscal year, reaffirming the $80 billion capital expenditure target but also admitting that there may be strategic adjustments or slowdowns in infrastructure construction in certain areas. Facing market doubts, Microsoft Corporation responded quickly. A company spokesperson stated in an email on Monday, local time, that "our plan to invest over $80 billion in infrastructure for artificial intelligence-related projects this fiscal year is proceeding as planned, and we will continue to grow at record speed to meet the AI computing needs of our customers." Although reaffirming its investment commitment, Microsoft Corporation also acknowledges that strategic adjustments may be made. The company spokesperson stated in the declaration, "Thanks to the significant investments we have made so far, we are well prepared to meet current and growing customer demands. In just one year, we have added capacity that surpasses any previous year in history. While we may strategically adjust the pace or scale of infrastructure in certain regions, we will continue to maintain strong growth in all regions. This allows us to invest and allocate resources to areas of future growth." Some analysts have pointed out that while Microsoft Corporation reaffirming its $80 billion spending plan demonstrates the company's increasingly forward-looking view of AI cloud computing needs, the emergence of DeepSeek leading a new "low-cost + high-performance" paradigm for AI training with large models will undoubtedly lead to a significant decrease in future AI training costs and AI computing needs. AI inference computing needs are expected to increase significantly, which also explains why Microsoft Corporation is cutting data center leases - after all, with the decline in AI training computing needs, one of Microsoft Corporation's largest customers, OpenAI, may no longer need such a large training computing resource. Gavin Baker, Managing Partner and Chief Investment Officer of Atreides Management Company, recently pointed out in a post that OpenAI's first-mover advantage is fading, with Alphabet Inc. Class C's Gemini, xAI's Grok-3, and D.The latest models of Eepseek have reached a level of technological sophistication similar to GPT-4. Gavin Baker states that in the future, only 2-3 massive data centers will be needed for pre-training, and the allocation of computational resources for pre-training and inference will be 5/95.In his latest article, Baker emphasizes that focusing on AI reasoning end is an extremely computationally intensive area of artificial intelligence large models, with unparalleled powerful artificial intelligence computing infrastructure, models can efficiently complete reasoning tasks. However, unlike the previous situation where pre-training and inference stage computing resource allocation are roughly equal, it is expected that in the near future, the field of AI computing infrastructure will shift to "pre-training accounting for only 5%, while inference computing resources account for 95%", highlighting the importance of outstanding inference infrastructure.