Microsoft President Brad Smith announced the investment of USD 3.3 billion to construct an artificial intelligence data center in Wisconsin, aiming to make the state a core driver of the innovation economy. Notably, the site of the facility was originally intended for a LCD panel plant promised by Foxconn six years ago.

According to Microsoft’s press release, the AI data center in Wisconsin is expected to create 2,300 union construction opportunities by 2025 and will provide long-term employment opportunities over the next several years.

Microsoft’s press release highlights that this investment will be utilized for constructing cloud computing and artificial intelligence infrastructure, establishing the first AI co-innovation lab in the United States focused on the manufacturing industry, and promoting AI training programs with the goal of enabling over 100,000 Wisconsin residents to acquire necessary AI skills.

The press release also notes that Microsoft will collaborate with Gateway Technical College to establish a Data Center Academy, aiming to train more than 1,000 students within five years, equipping them to enter roles in data centers or information technology departments.

Microsoft’s new facility in Racine County, Wisconsin, was originally intended to be the site of a LCD panel plant planned by Foxconn, a subsidiary of Hon Hai Precision Industry Co., Ltd. (Foxconn Group), according to a report by CNA.

In June 2018, then-chairman of Foxconn, Terry Gou, and then-US President Donald Trump attended the groundbreaking ceremony for the panel plant. Foxocnn announced an investment of USD 10 billion, and Trump described the project as the “8th wonder of the world.”

Read more

• [News] Microsoft Reportedly Stockpiling AI Chips to Boost Hardware Capability
• [News] Breaking Away from NVIDIA Dependency, Microsoft Reportedly Developing In-House AI Server High-Speed Network Card

(Photo credit: iStock)

With the skyrocketing demand for AI, cloud service providers (CSPs) are hastening the development of in-house chips. Apple, making a surprising move, is actively developing a data center-grade chip codenamed “Project ACDC,” signaling its foray into the realm of AI accelerators for servers.

As per a report from global media The Wall Street Journal, Apple is developing an AI accelerator chip for data center servers under the project name “Project ACDC.” Sources familiar with the matter revealed that Apple is closely collaborating with TSMC, but the timing of the new chip’s release remains uncertain.

Industry sources cited by the same report from Commercial Times disclosed that Apple’s AI accelerator chip will be developed using TSMC’s 3-nanometer process. Servers equipped with this chip are expected to debut next year, further enhancing the performance of its data centers and future cloud-based AI tools.

Industry sources cited in Commercial Times‘ report reveal that cloud service providers (CSPs) frequently choose TSMC’s 5 and 7-nanometer processes for their in-house chip development, capitalizing on TSMC’s mature advanced processes to enhance profit margins. Additionally, the same report also highlights that major industry players including Microsoft, AWS, Google, Meta, and Apple rely on TSMC’s advanced processes and packaging, which significantly contributes to the company’s performance.

Apple has consistently been an early adopter of TSMC’s most advanced processes, relying on their stability and technological leadership. Apple’s adoption of the 3-nanometer process and CoWoS advanced packaging next year is deemed the most reasonable solution, which will also help boost TSMC’s 3-nanometer production capacity utilization.

Generative AI models are rapidly evolving, enabling businesses and developers to address complex problems and discover new opportunities. However, large-scale models with billions or even trillions of parameters pose more stringent requirements for training, tuning, and inference.

Per Commercial Times citing industry sources, it has noted that Apple’s entry into the in-house chip arena comes as no surprise, given that giants like Google and Microsoft have long been deploying in-house chips and have successively launched iterative products.

In April, Google unveiled its next-generation AI accelerator, TPU v5p, aimed at accelerating cloud-based tasks and enhancing the efficiency of online services such as search, YouTube, Gmail, Google Maps, and Google Play Store. It also aims to improve execution efficiency by integrating cloud computing with Android devices, thereby enhancing user experience.

At the end of last year, AWS introduced two in-house chips, Graviton4 and Trainium2, to strengthen energy efficiency and computational performance to meet various innovative applications of generative AI.

Microsoft also introduced the Maia chip, designed for processing OpenAI models, Bing, GitHub Copilot, ChatGPT, and other AI services.

Meta, on the other hand, completed its second-generation in-house chip, MTIA, designed for tasks related to AI recommendation systems, such as content ranking and recommendations on Facebook and Instagram.

Read more

• [News] Breaking Away from NVIDIA Dependency, Microsoft Reportedly Developing In-House AI Server High-Speed Network Card
• [Insights] AI Market: A Battleground for Tech Giants as Six Major Companies Develop AI Chips

(Photo credit: Apple)

Four major cloud service providers (CSPs) including Google, Microsoft, Amazon, and Meta, sequentially released their first-quarter financial performance for the year 2024 (January 2024 to March 2024) at the end of April.

Each company has achieved double-digit growth of the revenue, with increased capital expenditures continuing to emphasize AI as their main development focus. The market’s current focus remains on whether AI investment projects can successfully translate into revenue from the previous quarter to date.

TrendForce’s Insights:

1. Strong Financial Performance of Top Four CSPs Driven by AI and Cloud Businesses

Alphabet, the parent company of Google, reported stellar financial results for the first quarter of 2024. Bolstered by growth in search engine, YouTube, and cloud services, revenue surpassed USD 80 billion, marking a 57% increase in profit. The company also announced its first-ever dividend payout, further boosting its stock price as all metrics exceeded market expectations, pushing its market capitalization past USD 2 trillion for the first time.For Google, the current development strategy revolves around its in-house LLM Gemini layout, aimed at strengthening its cloud services, search interaction interfaces, and dedicated hardware development.

Microsoft’s financial performance is equally impressive. This quarter, its revenue reached USD 61.9 billion, marking a year-on-year increase of 17%. Among its business segments, the Intelligent Cloud sector saw the highest growth, with a 21% increase in revenue, totaling $26.7 billion. Notably, the Azure division experienced a remarkable 31% growth, with Microsoft attributing 7% of this growth to AI demand.

In other words, the impact of AI on its performance is even more pronounced than in the previous quarter, prompting Microsoft to focus its future strategies more on the anticipated benefits from Copilot, both in software and hardware.

This quarter, Amazon achieved a remarkable revenue milestone, surpassing USD 140 billion, representing a year-on-year increase of 17%, surpassing market expectations. Furthermore, its profit reached USD 10.4 billion, far exceeding the USD 3.2 billion profit recorded in the same period in 2023.

The double-digit growth in advertising business and AWS (Amazon Web Services) drove this performance, with the latter being particularly highlighted for its AI-related opportunities. AWS achieved a record-high operating profit margin of 37.6% this quarter, with annual revenue expected to exceed $100 billion, and short-term plans to invest USD 150 billion in expanding data centers.

On the other hand, Meta reported revenue of USD 36.46 billion this quarter, marking a significant year-on-year growth of 27%, the largest growth rate since 2021. Profit also doubled compared to the same period in 2023, reaching USD 12.37 billion.

Meta’s current strategy focuses on allocating resources to areas such as smart glasses and mixed reality (MR) in the short and medium term. The company continues to leverage AI to enhance the user value of the virtual world.

2. Increased Capital Expenditure to Develop AI is a Common Consensus, Yet Profitability Remains Under Market Scrutiny

Observing the financial reports of major cloud players, the increase in capital expenditure to solidify their commitment to AI development can be seen as a continuation of last quarter’s focus.

In the first quarter of 2024, Microsoft’s capital expenditure surged by nearly 80% compared to the same period in 2023, reaching USD 14 billion. Google expects its quarterly expenditure to remain above USD 12 billion. Similarly, Meta has raised its capital expenditure guidance for 2024 to the range of USD 35 to USD 40 billion.

Amazon, considering its USD 14 billion expenditure in the first quarter as the minimum for the year, anticipates a significant increase in capital expenditure over the next year, exceeding the USD 48.4 billion spent in 2023. However, how these increased investments in AI will translate into profitability remains a subject of market scrutiny.

While the major cloud players remain steadfast in their focus on AI, market expectations may have shifted. For instance, despite impressive financial reports last quarter, both Google and Microsoft saw declines in their stock prices, unlike the significant increases seen this time. This could partly be interpreted as an expectation of short- to medium-term AI investment returns from products and services like Gemini and Copilot.

In contrast, Meta, whose financial performance is similarly impressive to other cloud giants, experienced a post-earnings stock drop of over 15%. This may be attributed partly to its conservative financial outlook and partly to the less-than-ideal investment returns from its focused areas of virtual wearable devices and AI value-added services.

Due to Meta’s relatively limited user base compared to the other three CSPs in terms of commercial end-user applications, its AI development efforts, such as the practical Llama 3 and the value-added Meta AI virtual assistant for its products, have not yielded significant benefits. While Llama 3 is free and open-source, and Meta AI has limited shipment, they evidently do not justify the development costs.

Therefore, Meta still needs to expand its ecosystem to facilitate the promotion of its AI services, aiming to create a business model that can translate technology into tangible revenue streams.

For example, Meta recently opened up the operating system Horizon OS of its VR device Quest to brands like Lenovo and Asus, allowing them to produce their own branded VR/MR devices. The primary goal is to attract developers to enrich the content database and thereby promote industry development.

Read more

• [News] NVIDIA’s H100 Chip Heat Declines, CoWoS Capacity Eases
• [News] Microsoft Reportedly Stockpiling AI Chips to Boost Hardware Capability

According to sources cited by the American news outlet Business Insider, Microsoft plans to double its inventory of GPUs to 1.8 million, primarily sourced from NVIDIA. Having more chips on hand will enable Microsoft to launch AI products that are more efficient, faster, and more cost-effective.

The source does not detail specific future applications for these chips, but acquiring a large quantity of chips means that Microsoft can extensively deploy them across its own products, including cloud services and consumer electronics.

The sources cited by the same report further revealed that Microsoft plans to invest USD 100 billion in GPUs and data centers by 2027 to strengthen its existing infrastructure.

Microsoft’s significant stockpiling of AI chips underscores the company’s efforts to maintain a competitive edge in the AI field, where having robust computing power is crucial for innovation.

On the other hand, NVIDIA recently stated that the AI computer they are collaborating on with Microsoft will operate on Microsoft’s Azure cloud platform and will utilize tens of thousands of NVIDIA GPUs, including their H100 and A100 chips.

NVIDIA declined to disclose the contract value of this collaboration. However, industry sources cited by the report estimate that the price of each A100 chip ranges from USD 10,000 to 12,000, while the price of the H100 is significantly higher than this range.

Additionally, Microsoft is also in the process of designing the next generation of the chip. Not only is Microsoft striving to reduce its reliance on NVIDIA, but other companies including OpenAI, Tesla, Google, Amazon, and Meta are also investing in developing their own AI accelerator chips. These companies are expected to compete with NVIDIA’s flagship H100 AI accelerator chips.

Read more

• [News] TSMC Q1 2024 Earnings Call Highlighted Two Major Revisions
• [News] Taiwan Hit by 7.2 Magnitude Earthquake: TSMC and Other Semiconductor, Panel Supply Chain Updates

(Photo credit: NVIDIA)

Microsoft is reportedly developing a customized network card for AI servers, as per sources cited by global media The Information. This card is expected to enhance the performance of its in-house AI chip Azure Maia 100 while reducing dependency on NVIDIA as the primary supplier of high-performance network cards.

Leading this product initiative at Microsoft is Pradeep Sindhu, co-founder of Juniper Networks. Microsoft acquired Sindhu’s data center technology startup, Fungible, last year. Sindhu has since joined Microsoft and is leading the team in developing this network card.

According to the Information, this network card is similar to NVIDIA’s ConnectX-7 interface card, which supports a maximum bandwidth of 400 Gb Ethernet and is sold alongside NVIDIA GPUs.

Developing high-speed networking equipment tailored specifically for AI workloads may take over a year. If successful, it could reduce the time required for OpenAI to train models on Microsoft AI servers and lower the costs associated with the training process.

In November last year, Microsoft unveiled the Azure Maia 100 for data centers, manufactured using TSMC’s 5-nanometer process. The Azure Maia 100, introduced at the conference, is an AI accelerator chip designed for tasks such as running OpenAI models, ChatGPT, Bing, GitHub Copilot, and other AI workloads.

Microsoft is also in the process of designing the next generation of the chip. Not only is Microsoft striving to reduce its reliance on NVIDIA, but other companies including OpenAI, Tesla, Google, Amazon, and Meta are also investing in developing their own AI accelerator chips. These companies are expected to compete with NVIDIA’s flagship H100 AI accelerator chips.

Read more

• [Insights] Microsoft Unveils In-House AI Chip, Poised for Competitive Edge with a Powerful Ecosystem
• [Insights] AI Market: A Battleground for Tech Giants as Six Major Companies Develop AI Chips

(Photo credit: Microsoft)