➀ NVIDIA CEO Jensen Huang states China's AI hardware capabilities are nearly on par with the U.S., citing Huawei's CloudMatrix 384 system as a key competitor to NVIDIA's GB200 NVL72.

➁ Huawei's CloudMatrix 384 delivers 300 PFLOPs BF16 compute with optical interconnects, but consumes 2.3x more power per FLOP than NVIDIA's solution.

➂ Despite Huawei's progress, NVIDIA's 2025 AI compute shipments (25,000-35,000 GB200 racks) will vastly outpace China's domestic deployments.

➀ VSORA secures $46 million from investors including Otium, Omnes Capital, and the European Innovation Council Fund to advance AI chip development;

➁ Funding will accelerate production of Jotunn8 (J8) inference chip, targeting silicon completion in 2025;

➂ Partnerships with semiconductor industry leaders ensure access to advanced technologies and manufacturing capabilities.

➀ TSMC adapts to diverse client needs by offering segment-specific leading-edge nodes (e.g., N3P, A16) and advanced packaging solutions for AI/HPC and consumer devices;

➁ The foundry's roadmap focuses on three directions: transistor scaling, backside power delivery optimization, and multi-die integration, with A16 (2026) and A14 (2028) targeting data centers;

➂ Despite slowing transistor density gains (7-10% per node), TSMC maintains 30% power efficiency improvements per generation, leveraging innovations like Super Power Rail and 3D packaging.

➀ AMD is facing challenges in catching up with NCCL and needs exclusive access to a persistent cluster of at least 1,024 MI300 class GPUs.

➁ AMD's RCCL library is a fork of Nvidia's NCCL and requires significant engineering hours to sync with Nvidia's major refactor.

➂ AMD is planning to rewrite RCCL from scratch to stop being a fork of NCCL.

➃ NVIDIA's NCCL continues to advance with new features and performance improvements.

➄ AMD has made progress in software infrastructure but is falling behind in ML libraries.

➅ AMD lacks support for features like disaggregated prefill and NVMe KV Cache Tiering.

➆ Recommendations are made to both AMD and NVIDIA for improving their competitive positions.

➀ The article emphasizes optimizing the RTL-to-GDSII flow to address increasing chip complexity and design challenges;

➁ Key strategies include improving RTL quality, leveraging AI-driven automation, streamlined synthesis, and early congestion analysis;

➂ Benefits include reduced turnaround time, enhanced power-performance-area metrics, and scalability for advanced semiconductor development.

➀ UCIe drives modern chip design with high-bandwidth data transfer, thermal efficiency, and reduced leakage;

➁ Integrates bioinspired advancements, system co-optimization, and 3D packaging to transform AI, HPC, and consumer electronics;

➂ Advances optical I/O to push technological boundaries in semiconductor innovation.

➀ A recent report has revealed that Meta's AI chatbot poses danger to children as it's allegedly having sexual conversations with minors.

➁ Concerns have been raised that Meta may have crossed ethical restrictions in its haste to deploy the feature, and that the company is not doing enough to protect children.

➂ Meta has denied the report and claims changes have been made to the chatbot to prevent such occurrences.

➀ Mike Lindell's legal team has been criticized for submitting an AI-generated document filled with erroneous legal citations in a defamation lawsuit;

➁ The document, submitted as part of a lawsuit against Lindell by a former Dominion Voting Systems employee, contained nearly 30 inaccuracies;

➂ The judge has demanded explanations and potential disciplinary action against the law firm and the attorneys involved.

➀ Emerging technologies like VR, AI, and IoT are transforming lab mouse research, enabling precise behavioral studies and ethical advancements;

➁ Northwestern University developed iMRSIV, a compact VR headset for mice, enhancing neuroscience research efficiency;

➂ AI-powered pain detection and IoT-enabled transport systems aim to reduce animal stress and improve research reliability.

➀ The article highlights the emerging humanoid robotics market, noting fewer than 20 companies globally developing such technology, including three UK-based firms focused on domestic and industrial applications;

➀ Ed explores leveraging UK government support to position the nation as a 'humanoid robot superpower,' citing bullish market forecasts ($36B by 2035 from Goldman Sachs, $7T by 2050 from Citigroup);

➂ The narrative humorously touches on potential use cases, from companion robots to government-backed robotic staff, while showcasing Ed's practical adoption of a robot receptionist in his office.

➀ Microsoft is relaunching Recall and adding an improved Windows Search feature for Copilot+ PCs.

➁ The Recall feature takes snapshots of your active screen and helps you find previously seen content.

➂ Microsoft has made efforts to address privacy and security concerns with Recall, emphasizing that collected information does not leave the user's computer.

➀ DigiTech ASEAN Thailand & AI Connect 2025 will take place from November 19th to 21st, 2025 in Bangkok, Thailand, following a successful edition in 2024.

➁ The event aims to connect global tech and digital markets, focusing on six key sectors including AI, IoT, cybersecurity, and smart solutions.

➂ Thailand's digital economy is projected to grow significantly, with strong government support and increasing demand for digital tools and solutions.

➀ TSMC unveiled its next-generation logic process technology, A14, at the North America Technology Symposium;

➁ A14 represents a significant advancement over TSMC's N2 process;

➂ It is designed to enhance AI transformation with faster computing and improved power efficiency.

➀ Intel reported Q1 revenue of $12.7 billion, flat y-o-y with a gross margin of 36.9% down 4.1 points y-o-y.

➁ The company expects Q2 revenue between $11.2-12.4 billion.

➂ CEO Lip-Bu Tan emphasized the need for better execution and operational efficiency to regain market share and drive sustainable growth.

➀ Data centres, driven by AI and cloud computing, are energy-intensive and face an environmental challenge due to their high power consumption.

➁ Current solutions like renewable energy and advanced cooling systems are insufficient to meet the growing demand for data processing and storage.

➂ Nuclear power is proposed as a potential solution for powering data centres, offering sustainable energy with benefits such as high energy density and low emissions, though challenges like safety concerns and high costs remain.

The ETH Zurich researchers have developed a method that makes AI answers more reliable over time. Their algorithm is highly selective in choosing data. Additionally, up to 40 times smaller AI models can achieve the same output performance as the best large AI models.

ChatGPT and similar tools often amaze us with the accuracy of their answers, but also often lead to doubt. One of the big challenges of powerful AI response machines is that they serve us with perfect answers and obvious nonsense with the same ease. One of the major challenges is how the underlying large language models (LLMs) of AI deal with uncertainty. It has been very difficult until now to judge whether LLMs focused on text processing and generation generate their answers on a solid foundation of data or whether they are on uncertain ground.

Researchers from the Institute for Machine Learning at the Department of Computer Science at ETH Zurich have now developed a method to specifically reduce the uncertainty of AI. 'Our algorithm can specifically enrich the general language model of AI with additional data from the relevant thematic area of the question. In combination with the specific question, we can then specifically retrieve those relationships from the depths of the model and from the enrichment data that are likely to generate a correct answer,' explains Jonas Hübotter from the Learning & Adaptive Systems Group, who developed the new method as part of his PhD studies.

➀ MIT engineers have developed a method to create thin electronic 'skins' that could lead to devices like sensors and imaging systems without cooling systems, potentially useful for night-vision glasses and autonomous driving in foggy conditions.

➁ The method involves growing semiconductor materials using remote epitaxy, allowing the film to peel off and be reused, which may extend to other types of electronics.

➂ Researchers are exploring the application of this technology in full night-vision systems and other areas such as environmental and biological sensing, as well as space signal detection.

➀ Some AI data centers in China are reportedly holding large stocks of China-specific 48GB Nvidia RTX 4090D GPUs and reselling them as new cards.

➁ Companies are dismantling and refurbishing the GPUs to generate profits.

➂ The report states that an AI data center requires a utilization rate of more than 70% to 75% to turn a profit, but the activation rates remain below 20%.

➀ Dr. Atul Lavand from Thermo Fisher Scientific discusses the application of molecular spectroscopy in the semiconductor and electronics industry, particularly in PCB manufacturing and failure analysis.

➁ He highlights the importance of techniques like FTIR and Raman spectroscopy in identifying materials and contaminants in PCBs, aiding in process optimization and defect prevention.

➂ The interview also touches upon the role of artificial intelligence in enhancing spectral interpretation and future product innovations, along with the company's approach to supporting startups and SMEs in India.