In addressing high computing costs for Generative AI (GenAI), there is a market gap in on-device AI model fine-tuning computation between cloud AI servers and AI PC terminals.
Japan-based semiconductor company Renesas released its financial results for the first quarter of 2024 on April 25. The company experienced a limited profit decline and expected the weaker performance to continue in the second quarter.
As the US-China technology war enters its seventh year, with the process of domestic substitution policy of China's semiconductor industry continuing to deepen, China's local Electronic Design Automation (EDA) software products have made considerable progress in recent years and accelerated their development to bridge the 15 years gap with their American and European rivals.
Nvidia is arguably the biggest "beneficiary" of Artificial Intelligence (AI) so far, but the recent 10% plunge in Nvidia's stock price and the evaporation of its market capitalization of US$211 billion have triggered concerns in the South Korean semiconductor industry.
The governments of Europe, the US, Japan, and China see quantum computers as a long-term strategic investment project. With the support of stable and long-term funding, the related start-up investments grew at an annual rate of about 20% between 2021-2023, according to CB Insights.
With Nvidia's recently announced Blackwell GPU architecture poised to become the mainstream for the next generation of AI servers, industry watchers expect Blackwell-equipped AI servers to begin shipping at the end of 2024 alongside the launch of the next-gen GPUs, while mass shipments are expected to begin in 2025.
Nvidia announced that Japan's new quantum supercomputer — designed to advance the nation's quantum computing initiative — will be powered by Nvidia platforms for accelerated and quantum computing.
Japan's Ministry of International Trade and Industry (MITI) unveils a JPY 72.5 billion (US$480 million) subsidy to back Generative AI (GenAI) development by five local firms, including telecom leader KDDI.
Samsung Electronics is reportedly expanding its artificial intelligence (AI) chip design research organization in Silicon Valley to challenge Nvidia, the current leader in the AI semiconductor market, by developing AI chips based on the RISC-V architecture.
The consumer electronics market has remained sluggish since 2023, with no evident signs of recovery even after the first quarter of 2024, whether for smartphones, notebooks, or consumer networking devices. Market observers attribute this prolonged downturn to persistently high inflation rates in Europe and the US.
Affected by the overall global economic situation in 2023, Taiwanese industrial PC (IPC) manufacturers saw high inventory levels among their downstream customers with DIGITIMES Research estimating that the total revenue of Taiwanese IPC manufacturers reached NT$283.66 billion (US$8.66 billion) for annual decrease of 1.3% in 2023, a relatively weak performance, according to DIGITIMES Research's latest AIoT report covering Taiwanese IPC makers' performance in recent years.
Lite-On Technology expects its power supply shipments for AI servers to rise 20-30% sequentially in the second quarter, said the company's president, Anson Chiu. All of its 3kW and 5kW AI server power supplies are manufactured at its Kaohsiung factory in southern Taiwan.
Allied Circuit (ACCL) is optimistic about its prospects for the second quarter of 2024, pointing to a rise in server market order visibility from 4-6 to 8 weeks. Customers' pulling of products has also increased dramatically in volume, according to the server PCB specialist.
Following the recent launch of Nvidia's GB200 superchip comes news of major changes to its first tier of AI server system manufacturers, as Foxconn and Wiwynn, which did not manufacture H100-based systems, now join the ranks of Supermicro, Quanta, Wistron, and ZT Systems in assembling the next generation of Nvidia's AI servers.
As Artificial Intelligence (AI) applications continue to materialize, there is an increasing demand for rugged computing products. This has enabled numerous Industrial PC (IPC) makers, such as Winmate, Neousys Technology, Ubiqconn Technology, and Vecow to capitalize on the rising trend of edge computing and mobility.
Microsoft's substantial investment of US$1.5 billion has secured it a board seat in the United Arab Emirates-based AI firm G42, albeit with a trade-off.
In recent years, Taiwan-based IPC (Industrial PC) manufacturers have accelerated their production diversification, driven by geopolitical factors and local manufacturing incentive policies.
According to PTI and Analytics India, Hewlett Packard Enterprise (HPE) has begun large-scale deployment of its made-in-India servers within a year of announcing plans to manufacture them locally, the company announced on Tuesday.
As OpenAI continues to be at the forefront of Generative AI developments, Apple has been criticized for lagging behind the pack in iPhone innovation and has been on the verge of losing its grasp of the market in China. To this end, Apple needs a Chinese-language AI model, of which Baidu's ERNIE Bot is one of more than 40 authorized by the Chinese government.