Taiwan's AI data center operator Zettabyte has announced that Pegatron will integrate Zware, its self-developed GPU software solution, to boost AI server performance and set a new standard for high-performance AI computing.
Apple Intelligence is targeting launch in the Chinese market by April 2025, but reports indicate that its collaboration with Baidu is facing difficulties.
Hewlett Packard Enterprise (HPE) surpassed revenue and operating margin expectations in its latest quarter, driven by strong AI-server demand, though margin pressures raise questions about sustained profitability.
Foxconn reported a November revenue of NT$672.585 billion (approx. US$20.76 billion), surpassing the total earnings of the previous year in the first 11 months of 2024. Despite a dip in Smart Consumer Electronics and Cloud & Networking Products, there was growth in Computing Products and Components & Others, driven by strong demand for AI products and new Apple releases. The company expects further improvement in the second half of the year, especially in the fourth quarter.
Sakura Internet is collaborating with NEC to develop a generative AI (GenAI) service platform, integrating NEC's "cotomi" AI technology into its offerings. To support this initiative, Sakura Internet plans to procure 10,000 Nvidia GPUs annually, highlighting the growing demand for GenAI solutions.
Amid surging electricity demands and growing potential in green energy, Foxconn is strengthening its market presence by leveraging AI to enhance operational efficiency and value creation.
Nvidia plans to collaborate with the Thai government to establish a world-class AI infrastructure, though questions remain about Thailand's ability to fully capitalize on its strengths in developing a significant presence in the AI data center ecosystem.
FineTek, a leading Taiwanese sensor manufacturer, is experiencing a sharp increase in orders from server ODM clients this December, driven by the mass production of Nvidia's GB200 liquid-cooled servers. The company anticipates continued strong momentum beyond the first quarter of 2025, with expectations of "astounding" order volumes in the second quarter.
Workstations are the latest hardware category to see an AI-related boom. As noted by James Huang, chairman of rugged industrial PC (IPC) specialist Getac Technology, the company has seen strong sales of its X600 Server, a fully rugged command server, due to enterprise demand for AI functions. In addition, Quanta executive VP Alan Tsai has recently noted untapped potential in workstations.
Amazon Web Services (AWS) has unveiled its latest innovation, the Trainium3 chip, featuring an advanced 3nm manufacturing process and scheduled for release in late 2025. The announcement event notably featured Apple's presence, highlighting the tech giant's significant reliance on AWS cloud services.
AWS has expanded beyond its existing Titan series with the introduction of the Nova family of AI models. The announcement came during the annual AWS conference, where Amazon CEO Andy Jassy made a special appearance to unveil various Nova versions including Micro, Lite, Pro, and Premier. While the first three versions are now generally available, Premier is scheduled for a later release.
The latest wave of export restrictions from the US on High Bandwidth Memory (HBM), particularly targeting advanced technologies such as HBM2 and above, is set to compel China's memory supply chain to expedite the development of core HBM technologies. As these regulations tighten, concerns arise within the memory supply chain regarding China's current investment in HBM, which remains at a nascent stage. Poor yield rates could further impact the circulation of secondary products like DDR5, leading to market price volatility.
Marvell reported strong fiscal third-quarter results with a 6.87% revenue increase, driven by record data center revenue, primarily from custom AI silicon. The company forecasts continued growth in the fourth quarter, with revenue expected to increase by 19% quarterly and 26% year-over-year. Analysts predict that AI sales will exceed expectations, potentially reaching US$1.6 billion.
Amazon Web Services (AWS) has announced three significant reforms in data center technology, including the streamlining of power distribution and structural design, the introduction of liquid cooling systems, and enhancements to energy efficiency. The company is also collaborating with Nvidia to advance the design of liquid-cooled server racks.
Megvii Technology, one of China's original "AI Elite Four" which specializes in computer vision and deep learning, announced its application for an initial public offering (IPO) on the Shanghai Stock Exchange Science and Technology Innovation Board (STAR) has been withdrawn.
Orders for brand AI servers are surging, prompting Taiwan-based electronic contract manufacturing firms to capitalize on the trend regardless of order variations.
Demand for AI servers will not peak in 2025, according to Quanta Computer executive VP and Quanta Cloud Technology general manager Mike Yang. In his opinion, the AI industry has seen an end-to-end progression encompassing all aspects from training and inference to end-user devices. Quanta has also seen AI server orders steadily increase.
The development of AI robots spans multiple sectors, from smart factories to smart homes and specialized applications. However, for semiconductor giants like Nvidia and Qualcomm, penetrating the industrial sector presents challenges reminiscent of the autonomous vehicle market.
Donald Trump has accelerated his cabinet appointments post-election, naming all 15 cabinet-level positions and over 20 key officials within 20 days—outpacing his first-term timeline.
Leading AI companies like OpenAI, Anthropic, and Google are encountering significant challenges in developing their next-generation AI models. Progress has fallen short of expectations, with models like OpenAI's "Orion" and Anthropic's "Claude 3.5 Opus" failing to meet internal benchmarks, resulting in delayed releases.
Both Tesla CEO Elon Musk and Nvidia CEO Jensen Huang consider humanoid robots to be the ultimate chapter in AI development. According to a Goldman Sachs research forecast for the first half of 2024, humanoid robot shipments are expected to exceed 250,000 units by 2030, with nearly all deployments in industrial sectors.
Elon Musk's artificial intelligence company xAI has made headlines with its urgent push for advanced AI hardware. Industry sources indicate that Musk directly approached Nvidia CEO Jensen Huang, offering a premium to expedite a US$1.08 billion order. Production of these AI server systems will be handled by Foxconn, Nvidia's key manufacturing partner.
Global smartphone shipments have slowed in recent years due to high inflation, market saturation, and extended replacement cycles. However, the surge in generative AI (GenAI) has injected new life into the stagnant market, with global smartphone demand projected to grow 5% year-over-year, reaching nearly 1.2 billion units in 2024.
South Korean AI chip startup DeepX is in the final stages of preparation for its first mass-produced chips. Through Samsung Electronics' foundry division, the company has achieved an 88% yield rate using advanced 5nm process technology, with expectations to surpass 90% in mass production—a promising indicator for this cutting-edge AI hardware initiative.
