The latest quarterly reports from the big four cloud service providers (CSPs) have been released in succession. According to a report from Commercial Times, not only has there been significant revenue growth, but capital expenditures for these CSPs have also surged compared to the same period last year, underscoring the ongoing momentum in AI investments.

Industry scources cited by Commercial Times estimate that capital expenditures by CSPs will surpass USD 240 billion by 2025, reflecting an annual increase of over 10%.

The report indicated that the increase in capital expenditures by CSPs is expected to boost demand for Taiwanese companies in the supply chain during the fourth quarter of this year and into next year, benefiting companies such as Quanta, Wistron, Wiwynn, and Inventec.

According to the report, Microsoft’s capital expenditures for the first quarter of fiscal year 2025 (the third quarter of 2024) reached USD 20 billion, higher than USD 19 billion of the previous quarter, reflecting a 78% increase year-on-year. Microsoft noted that the demand for AI now exceeds available production capacity, and they plan to continue increasing investment, expanding data center construction, and promoting AI services.

The report indicated that the market estimates Microsoft’s total expenditures for fiscal year 2025 will reach USD 80 billion, an increase of over USD 30 billion compared to the previous year.

Google’s capital expenditures in the third quarter reached USD 13.1 billion, an annual increase of 62%, which means that total capital expenditures in 2024 will reach USD 51.4 billion, an annual increase of 59%, and capital expenditures will continue to increase next year, according to the report.

Amazon’s capital expenditures for the third quarter reached USD 22.62 billion, reflecting an 81% year-on-year increase. This year, Amazon’s total capital expenditures have reached USD 51.9 billion, and full-year investments are projected to be as high as USD 75 billion. Furthermore, capital expenditures for next year are expected to be even higher, as the report indicated.

According to the report, as for Meta, capital expenditures in the third quarter were USD 9.2 billion, an annual increase of 36%. Moreover, Meta adjusted their capital expenditure forecast for fiscal 2024 to an upward revision of USD 40 billion. The report indicated that its capital expenditures will continue to grow in 2025.

The report highlighted that AI business opportunities will continue to benefit Taiwan’s major server ODMs. Companies such as Quanta, Wistron, Wiwynn, Inventec, and Foxconn all reported strong results in the third quarter and are optimistic about the fourth quarter and the year ahead.

According to the report, Quanta’s third-quarter revenue reached a record high, driven by strong demand for AI server orders. Quanta Chairman Barry Lam also expressed an optimistic outlook on the future of AI, noting that as large-scale CSPs develop generative AI applications, the scale of AI data centers is continually expanding, leading to a substantial increase in orders.

After demonstrating strong growth momentum in the first half of the year, Wistron has benefited from urgent orders in the second half. Additionally, some B200 series products utilizing the next-generation Blackwell platform are scheduled to be shipped after the fourth quarter. The report indicated that Wistron is quite optimistic about its performance for this quarter and next year.

Inventec plans to ship servers to customers primarily from US-based CSPs in the second half of the year. The report highlighted that orders from Google have increased as the company expands its purchase of AI servers based on its own TPU architecture, in addition to acquiring general-purpose servers for new platforms.

Read more

• High-End AI Server Demand from North America’s Top Four CSPs Expected to Exceed 60% in 2024, Says TrendForce
• [Insights] Big Four CSPs Continue to Shine in Q1 2024 Financial Reports, AI Returns Garnering Attention

(Photo credit: Microsoft)

According to a report from The Register, DPU developer Xockets recently filed a lawsuit, accusing AI chip giant NVIDIA, Microsoft, and intellectual property risk management company RPX of colluding to avoid paying Xockets the fees it is owed, violating federal antitrust laws, and intentionally infringing on its patents.

The report states that in addition to seeking monetary compensation, Xockets is also requesting an injunction. If granted, this injunction would prevent NVIDIA from selling its upcoming Blackwell architecture GPUs.

Per Reuter’s report, Xockets, founded in 2012, claims that its invention, the Data Processing Unit (DPU), plays a critical role in some of NVIDIA’s and Microsoft’s systems. The company states that its technology helps offload and accelerate tasks that would otherwise place a heavy burden on server processors.

Reportedly, Xockets founder Parin Dalal began filing a series of DPU technology patents in 2012. These patents describe architectures used for the linear downloading, acceleration, and isolation of data-intensive computational operations from server processors.

Xockets claims that its DPU-related patents cover various applications including cloud computing, machine learning, security, network overlay, stream data processing, and cloud computing architectures. Xockets alleges that Microsoft and Mellanox, which was acquired by NVIDIA in 2020, which was acquired by NVIDIA in 2020, have infringed on these patents.

In a recent statement, Xockets claimed that NVIDIA has utilized DPU technology patented by Xockets, allowing NVIDIA to monopolize the AI server market using its GPUs. Meanwhile, Microsoft has allegedly monopolized the AI platform market using NVIDIA GPUs.

Xockets further claimed that it has made effort to engage in sincere negotiations with NVIDIA and Microsoft, but these attempts have been rejected.

Xockets’ lawsuit reveals that it actually demonstrated the relevant technology to Microsoft in 2016, and the technology was subsequently adopted by Mellanox within the same year for cloud computing downloads used by Redmond and other clients.

Additionally, NVIDIA’s ConnectX smartNIC, BlueField DPU, and NVLink switch, which are crucial for extending AI training and inference deployments across large GPU clusters, are said to infringe on Xockets’ patents.

Regarding this matter, NVIDIA has declined to comment, while Xockets’ spokesperson has also not provided any additional explanation.

The report highlights that Microsoft and NVIDIA may not be Xockets’ only targets but are at least the most profitable ones. Other companies, such as Broadcom, Intel, AMD, Marvell, Napatech, and Amazon, are also actively developing products similar to NVIDIA’s ConnectX, BlueField, and NVLink.

Regarding the lawsuit, the judge overseeing the case has approved a preliminary injunction hearing to be held on September 19.

Read more

• [News] NVIDIA’s Blackwell Overcomes Delays, as GB200 Reportedly Sets for December Mass Production
• [News] Foxconn Remains Unfazed Despite NVIDIA’s GB200 Delay, with Customers Aim to Secure Their Orders

(Photo credit: Xockets)

According to a report from The Wall Street Journal citing sources on August 13th, it’s revealed that Chinese internet companies and telecom operators have been testing Huawei’s latest processor, the “Ascend 910C,” in recent weeks. Reportedly, Huawei has informed potential customers that this new chip is comparable to NVIDIA’s H100 GPU, which cannot be directly sold in China.

Huawei’s ability to continue advancing its chip technology is a sign of its efforts to counter U.S. sanctions. However, the report also indicated that Huawei is already experiencing production delays with its current chips. The company faces additional U.S. restrictions, limiting its access to parts for production equipment and the latest memory used in AI hardware.

The sources cited by the same report point out that, TikTok’s parent company ByteDance, search giant Baidu, and state-owned telecom operator China Mobile are in preliminary talks with Huawei to secure the Ascend 910C chip. These negotiations suggest that Huawei could secure orders for more than 70,000 chips, valued at approximately USD 2 billion.

Reportedly, Huawei aims to begin shipping the Ascend 910C in October, but the final delivery schedule might differ from the initial plan and could be subject to adjustments.

Under U.S. sanctions, customers in China are forced to purchase the H20 from NVIDIA, which is a “downgraded” version of the AI chip designed specifically for the Chinese market.

Per a previous report from South China Morning Post, it’s expected that Chinese tech giants may be considering a shift towards local AI products, which could pose a challenge to NVIDIA. Currently, China accounts for 17% of NVIDIA’s revenue in the 2024 fiscal year, making the competition in the Chinese market increasingly fierce for NVIDIA.

Compared to NVIDIA’s customers in China, NVIDIA’s U.S. customers, such as OpenAI, Amazon, and Google, will soon have access to NVIDIA’s latest Blackwell architecture chips, including new products like the GB200, which NVIDIA claims offer significantly improved performance compared to existing products.

Meanwhile,  Wall Street Journal also has cited sources, pointed out that NVIDIA is working on another China-oriented chip called B20, but the design might have trouble getting U.S. approval for China export if the regulations are further tighten.

Read more

• [News] NVIDIA Reportedly Developing Tailored Version of the Blackwell Series for Chinese Market
• [News] U.S. Reportedly Enhancing Chip Export Restrictions to China, while NVIDIA’s H20 Might Get Banned

(Photo credit: NVIDIA)

According to Reuters, engineers at Amazon’s chip lab in Austin, Texas, recently tested highly confidential new servers. Per the Economic Times, the director of engineering at Amazon’s Annapurna Labs under AWS Rami Sinno revealed that these new servers feature Amazon’s AI chips, which can compete with NVIDIA’s chips.

It’s reported that Amazon is developing processors to reduce reliance on the costly NVIDIA chips, which will power some of Amazon’s AWS AI cloud services.

Amazon expects to use its self-developed chips to enable customers to perform complex calculations and process large amounts of data at a lower cost. The company’s competitors, Microsoft and Alphabet, are also pursuing similar efforts.

However, Amazon is a late starter in AI chip field, but a industrial leader in non-AI processing chip, whose main non-AI processing chip, Graviton, has been in development for nearly a decade and is now in its fourth generation. The other two AI chips, Trainium and Inferentia, are newer designs.

David Brown, AWS’s Vice President of Compute and Networking, stated that in some cases, the performance of these chips can be 40% to 50% higher compared to NVIDIA’s, and their cost is supposed to be about half of the same models of NVIDIA’s chips.

AWS accounts for nearly 20% of Amazon’s total revenue. The company’s revenue from January to March surged by 17% from the same period last year, reaching USD 25 billion. AWS controls about one-third of the cloud computing market, with Microsoft’s Azure comprising about 25%.

Amazon stated that it deployed 250,000 Graviton chips and 80,000 custom AI chips to handle the surge in platform activity during the recent Prime Day.

Read more

• [News] TSMC Reportedly Sees Strong 4nm Demand, with NVIDIA’s Order Up by 25%
• [News] Amazon Unveils New AWS-Designed Chips, Boosting Orders for TSMC and ALCHIP

(Photo credit: Amazon)

Taiwan’s semiconductor giant, TSMC, faces overwhelming demand for its 3nm technology, with major clients like Apple and NVIDIA fully allocate its production capacity.

According to a report from Commercial Times, orders are expected to be filled through 2026. Reportedly, TSMC is planning to raise its 3nm prices by over 5%, and advanced packaging prices are anticipated to increase by approximately 10% to 20% next year.

The members of TSMC’s 3nm family include N3, N3E, N3P, as well as N3X and N3A. As the existing N3 technology continues to be upgraded, N3E, which began mass production in the fourth quarter of last year, targets applications such as AI accelerators, high-end smartphones, and data centers.

N3P is scheduled for mass production in the second half of this year and is expected to become mainstream for applications in mobile devices, consumer products, base stations, and networking through 2026. N3X and N3A are customized for high-performance computing and automotive clients.

Source: TSMC

Per the industry sources cited by the same report, TSMC’s Zhunan advanced packaging plant (AP6), operational for a year now, has become Taiwan’s largest CoWoS base with the equipment moved into its AP6C plant. In the third quarter, CoWoS monthly production capacity is expected to double from 17,000 to 33,000 wafers.

Industry sources cited by the report further suggests that while AI accelerators do not use the most cutting-edge manufacturing processes, they rely heavily on advanced packaging technology. The ability of global semiconductor companies to secure more advanced packaging capacity from TSMC will determine their market penetration and control.

TSMC’s advanced packaging capacity is scarce, with primary customer NVIDIA having the highest demand, occupying about half of the capacity, followed closely by AMD. Broadcom, Amazon, and Marvell have also expressed strong interest in using advanced packaging processes. With gross margins close to 80%, NVIDIA is said to agree to price increases to secure more advanced packaging capacity, thereby distancing itself from competitors.

Previously, NVIDIA CEO Jensen Huang emphasized that TSMC is not just manufacturing wafers but also handling numerous supply chain issues. He also agreed that the current pricing is too low and would support TSMC’s price increase actions.

The industry sources cited by Commercial Times have indicated that TSMC plans to add CoWoS-related equipment by the third quarter and has requested equipment manufacturers to dispatch more engineers to fully staff its Longtan AP3, Zhunan AP6, and Central Taiwan Science Park AP5 plants.

In addition to Zhunan’s AP6C, the Central Taiwan Science Park plant, which originally only handled the latter stages of oS, will also gradually transition to CoW processes. Meanwhile, the Chiayi site is in the land preparation stage and is expected to progress faster than Tongluo.

Reportedly, industry sources further reveal that the prices for advanced process nodes such as 3nm and 5nm will also be adjusted. Particularly, strong demand for 3nm orders in the second half of the year is expected to drive utilization rates to near full capacity, extending through 2025. The 5nm process is experiencing similar demand dynamics, driven by AI needs.

Read more

• [News] Major Clients Reportedly Fully Allocate TSMC’s Production Capacity Until 2026, 3nm Process in High Demand
• [News] NVIDIA CEO Jensen Huang Supports C.C. Wei’s Theory and Backs for TSMC to Increase Price

(Photo credit: TSMC)