#semiconductors

A new platform allows users to submit and read community reviews for electronic components, aiming to provide peer feedback on parts like semiconductors and modules.

Researchers have developed a technique to create 3D chips by rolling thin sheets of silicon, potentially shrinking circuits and improving performance. This approach stacks multiple layers of circuitry in a compact rolled form, offering a new way to increase transistor density beyond traditional planar scaling.

Huawei is shifting its chip strategy toward boosting processing speed rather than shrinking transistor size, aiming to bypass US export restrictions that limit its access to advanced semiconductor manufacturing technology.

The European Union is seeking emergency powers to take control of semiconductor supplies during a crisis, aiming to reduce reliance on Asian chip producers and bolster the bloc's strategic autonomy in the face of potential supply disruptions.

Micron Technology's market capitalization surpassed $1 trillion for the first time, driven by surging demand for its AI memory chips. The milestone reflects the broader AI boom fueling growth for semiconductor companies producing high-bandwidth memory essential for AI data centers.

French AI startup Mistral is exploring the possibility of designing its own chips for AI data centers, a move that could reduce its reliance on external suppliers like Nvidia and give it more control over performance and costs.

Nvidia plans to invest $150 billion annually in Taiwan to build AI infrastructure, as part of a broader $500 billion U.S. AI investment push. The spending will focus on advanced chip manufacturing and data centers, deepening ties with Taiwanese suppliers like TSMC.

Researchers have developed memory-preserving transistors that can operate below the Boltzmann limit, a fundamental energy barrier for conventional electronics. The new design uses a ferroelectric layer to retain state information, allowing switches to function with significantly less energy. This breakthrough could lead to far more energy-efficient computing and memory devices.

This video explores the highly complex global supply chain behind semiconductor manufacturing, detailing how raw materials are sourced, processed, and assembled across multiple countries before becoming finished chips used in modern electronics.

Mistral AI CEO Arthur Mensch told CNBC the company is exploring designing its own chips for AI data centers, potentially joining a trend of tech firms developing in-house hardware to reduce reliance on external suppliers and optimize performance.

The DARE (Digital Autonomy with RISC-V in Europe) project aims to strengthen European digital sovereignty by advancing the open-standard RISC-V instruction set architecture. The initiative seeks to reduce dependency on proprietary chip technologies and foster innovation in hardware and software across Europe.

Nvidia CEO Jensen Huang announced the company will invest $150 billion annually in Taiwan, calling the island the "epicentre" of the global AI revolution. The investment will support chip production and AI infrastructure, underscoring Taiwan's critical role in the semiconductor supply chain.

SK Hynix's market capitalization surpassed $1 trillion, making it the third memory chipmaker to reach this milestone after Samsung Electronics and Micron Technology, driven by the artificial intelligence chip boom.

Micron Technology, an Idaho-based chip maker, surged to a $1 trillion market capitalization in just 48 days, more than doubling its value amid a boom in AI-driven demand for memory chips. The rapid ascent highlights the growing influence of semiconductor companies in the global stock market.

Huawei claims a breakthrough in chip design, aiming to overcome U.S. export restrictions that have limited its access to advanced semiconductors and technology.

Chinese manufacturers are rapidly increasing production of DRAM and NAND flash memory, potentially leading to a global oversupply that could drive down prices for SSDs and RAM modules.

Huawei is developing a chip technology called 'LogicFolding' that aims to narrow the performance gap with competitors, as the company continues to face restrictions from US sanctions limiting its access to advanced semiconductor manufacturing.

Huawei has announced plans to develop a 1.4nm equivalent density semiconductor process by 2031, despite ongoing US trade restrictions limiting its access to advanced chipmaking equipment. The company aims to achieve this through innovations in advanced packaging and stacking technologies, bypassing traditional EUV lithography reliance.

Huawei has proposed a new "Tau Scaling Law" as a replacement for traditional geometric scaling in the semiconductor industry. The company presented the concept at the IEEE International Symposium on Circuits and Systems (ISCAS), suggesting it could address the physical and economic limitations of conventional transistor shrinking.

The article argues that the U.S. chip export restrictions on China are not a coherent long-term strategy but a temporary measure that creates a countdown for China to achieve semiconductor self-sufficiency. It contends that without a broader industrial and innovation policy, the ban will only accelerate China's domestic chip development and ultimately weaken U.S. technological leadership.

The article discusses the market for custom silicon chips, explaining that companies like Apple, Google, and Amazon design their own chips to optimize performance and power efficiency for specific products, rather than using off-the-shelf components from traditional suppliers.

IBM has established what it claims is the first pure-play quantum chip foundry, leveraging a $2 billion CHIPS Act investment. The facility uses 300mm superconducting silicon technology to produce quantum processors, aiming to advance the commercialization and scalability of quantum computing hardware.

Bolt Graphics has announced the Zeus GPU, a new graphics processor aiming to compete with Nvidia. The Zeus focuses on high-performance graphics and ray tracing for workstations and data centers, claiming significant speed improvements over existing hardware in specific tasks.

Epoch AI analysis shows that memory now accounts for nearly two-thirds of AI chip component costs, up from roughly half in 2019. The increasing size of AI models drives demand for high-bandwidth memory like HBM, making it the dominant cost factor over compute logic.

On-device AI requires rebuilding electronic products around a new chip layer. Specialized processors handling AI workloads locally will determine the next hardware wave, making chip-level design decisions critical for future devices.

The EU is reportedly planning a "CHIPS Act 2.0" to boost its semiconductor sector, shifting focus from self-sufficiency to technological "indispensability" in the global chip supply chain.

MatX is a startup developing high-throughput chips specifically designed for large language models (LLMs), aiming to improve efficiency and performance for AI workloads.

Reiner Pope presents a talk on chip design, approaching the topic from foundational principles and building upward through the various layers of abstraction involved in creating modern semiconductors.

Reiner Pope discusses chip design from the bottom up, covering how modern processors are built from transistors to architectures, the challenges of hardware design, and insights into the future of computing and AI hardware development.

This video explains how specialized AI chips, such as GPUs, TPUs, and NPUs, differ from traditional CPUs by handling parallel computations more efficiently, enabling faster processing of machine learning models and neural networks.