DIGITIMES Research believes tight CoWoS capacity, which is needed for packaging HBM, will result in the supply of high-end AI servers falling short of demand by more than 35% in 2023, but high-end AI server shipments worldwide are still expected to grow fivefold from a year ago, according to the latest server report covering the AI server market.
To align with global space industry development trends and establish comprehensive solutions for Taiwan's space industry, Taiwan Accelerator Plus (TAcc+), a Taiwanese government-sponsored startup accelerator has formed cooperation with international accelerators and organizations to invite overseas SpaceTech startups to visit Taiwan for a whole month, starting August 15.
With tensions between Washington and Beijing taking center stage, concerns emerged regarding the status of supply chains and the competitive landscape among various Asian countries. To gain a more profound understanding of this situation, DIGITIMES conducted the "Asia Supply Chain: ASC 250 Rankings," with a specific focus on the Asian ICT industry chain and an analysis of the operational performance of nearly 5,000 companies.
Chipmaking and PCB production equipment suppliers are bullish on advanced chip demand as a result of a surge in AI servers and related products, according to industry sources.
Because of the US chip ban and the shortage of Nvidia's GPUs, many China-based cloud service providers (CSPs) have turned to domestically produced chips with higher energy consumption that require stronger heat dissipation solutions, for their AI needs. This is thus bringing new business opportunities to China's local cooling solution providers, according to sources in the cooling solution business.
At the ongoing Intel Innovation 2023 conference, Intel CEO Pat Gelsinger announced a number of the company's important AI progress initiatives, including the upcoming launch of its 14th Gen Meteor Lake CPUs, which will power tens of millions of PCs due to ship next year.
Chinese server vendor Inspur has announced its investment in Shanghai Yunsilicon Technology, a startup that develops and innovates cloud datacenter ASIC products.
Professor Alberto Sangiovanni-Vincentelli is a legend of the electronic design automation (EDA) industry, but his stories were seldom told. DIGITIMES Asia is privileged to have this exclusive interview to share his wisdom and insights for the semiconductor industry during uncertain times in Part 1 of the interview. In Part 2, he recounts the founding of the world's two largest EDA tool companies that he co-established, and shares some exciting tech innovation projects that he is currently engaging with:
IC design company Weltrend Semiconductor has seen operations recover quarter by quarter as it moves inventory at a healthy pace. The company sees promising futures for both its main product lines: USB-PD chips are expected to benefit from a steady increase in notebook shipments, while fan motor driver chips are aimed at the rising demand for AI servers.
The semiconductor industry's correction period has already exceeded one year due to weak demand for smartphones and notebook computers. Nevertheless, there is a light at the end of the tunnel – applications created by generative AI are propelling robust growth for AI servers, but what comes as a surprise is the strong shipment growth seen by motherboard manufacturers.
Ever since tensions between the US and China started to intensify, countries around the world have been adjusting their military spending. The Russia-Ukraine war further brought these actions from behind the scenes into the spotlight. Industrial PC (IPC) manufacturers stated that while military-industrial orders are stable and have high profit margins, the market is highly sensitive. Thus, they need to maintain a low profile with their approach.
Alberto Sangiovanni-Vincentelli, professor and Edgar L. and Harold H. Buttner Chair of Electrical Engineering and Computer Science Department at the University of California, Berkeley, sat down to an exclusive interview with DIGITIMES Asia at the sidelines of SEMICON Taiwan 2023 on September 8. As a legend who created two of the world's most prominent electronic design automation (EDA) companies and who has accumulated decades of knowledge and experience in the semiconductor industry, Professor Sangiovanni-Vincentelli shed light on how far generative AI can go as a tool to advance the semiconductor industry. He also shared his thoughts on the solution of geopolitical tensions surrounding the semiconductor supply chain. Here is part one of the interview:
Chain Reaction, an Israel-based startup founded in 2019, is leveraging Taiwan's cutting-edge semiconductor manufacturing capability to build a privacy processor chip using a privacy enhancing technology called Fully Homomorphic Encryption (FHE) which transforms the public cloud into a trusted environment. This empowers use cases and workloads from big pharma, finance, aerospace, defense, and government where data needs to be securely and efficiently processed. This enablement solves the challenge commonly referred as the holy grail of cloud computing.
In a recent interview with DIGITIMES Asia, Fujitsu Executive Director of Strategic Alliance Unit Taisuke Iwai finds digital annealing is the game changer in the field of quantum computing.
Will it take until the second half of 2024 for the general-purpose server market to recover? According to Inventec, the industry's leading server motherboard manufacturer, it is still too early to tell when the market will break out of its slump: although high expectations abound, the number of orders actually placed tells a different story. Most likely, the market will need to wait until the third quarter of 2024 for a rebound.
Elton Yang, senior VP at Quanta and president of Quanta Cloud Technology (QCT), signals an unrelenting momentum in the AI server market and dispels concerns over any imminent demand drop-off. Yang underscores that the only pressing concern right now is the abundance of orders and a concurrent scarcity of GPUs. However, he optimistically predicts that the supply-demand imbalance will gradually ease by 2024.
Following the release of Huawei's Mate60 Pro series of smartphones, it has been reported that its Harmony operating system (OS) will enter the PC sector in 2024.
Recently, Chinese smartphone manufacturers are aggressively incorporating large language models (LLMs) into their devices, and this trend deserves a closer look. Apple appears not in a hurry to join the battle any time soon, and therefore do not expect the iPhone 15 that is slated to be launched these days to have LLM inside.
Quantum computing, often considered the next frontier in computing technology, has garnered immense attention in recent years. While giants like Google and IBM have made significant strides in the field, another player, IonQ, is making waves with its unique ion trap approach to quantum computing.
ChatGPT not only drives the demand for AI servers but also underscores the urgency to adopt high-speed computing applications. However, high-speed computing cannot rely solely on servers with heterogeneous computing as the core. It also requires peripherals such as transmission and storage devices, thus driving the demand for optoelectronic signal conversion. Under the trend, Lintes Technology Co., with its expertise in optoelectronic signal conversion engine technology, has not only attracted investment from Intel but also proposed relevant solutions for AI server demands.
Server ODMs have begun large-scale shipments of AI servers equipped with Nvidia's H100 GPU in the third quarter; however, a shortfall in GPU supply may impact their deliveries, according to industry sources.
Blumind is pitching a farsighted approach of neural network processing using an all-analog architecture for AI applications to be more sustainable in the future. The company has completed the latest phase of R&D and is now looking to commercialize its high-performance, low-cost, and sustainable AI chips.
Following the meeting between India's Prime MInister Narendra Modia and Nvidia CEO Jensen Huang, Nvidia announced a partnership with Reliance Industries for buildning AI infrastructure and to train engineers at Tata Consultancy Services, as the AI chip giant is looking to capitalize on India's AI potential.
