德国研究人员推出了一种由极薄半导体层制成的超快光开关，其运行速度比当今的电子晶体管快约 10,000 倍。 由奥尔登堡大学物理学家团队设计的新型光学器件，可用作光开关或光晶体管。研究人员表示，超快光开关为光数据处理提供 了广阔的应用前景。 这种被描述为纳米结构"活性超材料"的原型由银和原子级薄的半导体层制成。它能够在飞秒（即千万亿分之一秒）的时间尺度 上控制光。 "如果我们想在纳米尺度上制造超快光开关，我们的发现就特别有意义，"奥尔登堡大学实验物理学教授、该研究的主要作者克 里斯托夫·利瑙博士表示。 纳米尺度光学 利瑙和他的研究团队着手寻找一种材料，利用聚焦激光束，可以在短短几个飞秒内操控或切换其反射特性。作为参考，一飞秒 等于十亿分之一秒的百万分之一。 在这个项目中，研究团队使用了一种超薄的银纳米狭缝阵列。然后，他们在其表面铣削出一个平行沟槽网格，每个沟槽的宽度 和深度约为 45 纳米（十亿分之一毫米）。 参与这项研究的英国剑桥大学的科学家们随后将一层仅有三个原子厚的半导体晶体二硫化钨单层涂覆到该结构的表面上。 莫里茨·吉廷格博士（左）和丹尼尔·蒂默博士（右）站在这项研究所依据的实验装置前。图片来源：奥尔登 ...

Core Viewpoint - A global storage crisis is emerging, driven by a significant shortage of memory chips, particularly DRAM, which is impacting profits across various sectors, including consumer electronics and automotive industries [2][3][4]. Group 1: Causes of the Crisis - The shortage of DRAM is primarily attributed to the increasing demand from artificial intelligence data centers, which require high bandwidth memory (HBM) for processing large datasets [3][6]. - Companies like Alphabet Inc. and OpenAI are purchasing millions of memory chips for AI applications, leading to a competitive squeeze on supply for consumer electronics manufacturers [4][5]. - The price of DRAM has surged dramatically, with some prices increasing by 75% from December to January, reminiscent of hyperinflation [4][23]. Group 2: Impact on Companies - Major semiconductor manufacturers, including SK Hynix, Samsung, and Micron, control over 90% of global memory chip production and are experiencing record valuations due to the demand surge [4][5]. - The shift in production focus from traditional DRAM to HBM has resulted in reduced availability of standard memory chips for consumer electronics, affecting companies like Apple and Tesla [2][5]. - The memory crisis is causing significant disruptions in product release schedules, with companies like Sony considering delays for their next-generation gaming consoles [19][23]. Group 3: Market Predictions - TrendForce predicts that DRAM and NAND flash prices will rise by 90% to 95% and 55% to 60%, respectively, in the first quarter alone [4][6]. - The demand for HBM is expected to grow by 70% year-on-year by 2026, indicating a long-term shift in the memory market dynamics [6][22]. - The supply-demand imbalance is projected to persist throughout the year, with potential declines in smartphone shipments and increased prices for low-end devices [23][19].

Group 1 - The article discusses the compliance challenges faced by automotive manufacturers due to U.S. regulations limiting the use of Chinese technology in connected vehicles, requiring tracking of embedded software sources before a key deadline in March [2][3] - The U.S. Department of Commerce's new rules mandate that as of March 17, automakers must prove that core vehicle systems connected to the cloud do not contain software written by Chinese companies, impacting advanced driver-assistance systems and extending to connected hardware by 2029 [3] - The regulations aim to mitigate risks associated with sensitive data transmission through systems like cameras and GPS, complicating the supply chain for automakers who may struggle to trace embedded code from Chinese joint ventures or subcontractors [3] Group 2 - The regulatory shift presents opportunities for U.S. suppliers, such as Eagle Wireless, which is positioning itself as an alternative supplier for cellular modules, previously dominated by Chinese manufacturers with an 87% global market share in the first half of last year [4] - Eagle Wireless is collaborating with original equipment manufacturers to migrate software before the compliance deadline while building domestic manufacturing capabilities, indicating a significant shift in software development and manufacturing back to the U.S. [5] - European suppliers are also affected, with companies like Pirelli evaluating options due to their major shareholder being a Chinese group, potentially leading to divestment or separation of U.S. operations [5]

Core Viewpoint - The LM317, introduced by National Semiconductor in the mid-1970s, revolutionized linear voltage regulation by integrating adjustable voltage functionality into a three-terminal module, allowing for a wide range of output voltages with minimal components [2][3]. Group 1: Product Features and Design - The LM317 is an adjustable positive voltage regulator capable of delivering over 1.5A of current, with output voltage adjustable between approximately 1.25V and 37V using two resistors [2]. - Unlike traditional regulators that compare output voltage to ground, the LM317 maintains a reference voltage of about 1.25V between its output and adjustment pins, allowing for a wide voltage range without internal recalibration for each voltage value [4][7]. - The device features built-in current limiting, thermal shutdown, and safe operating area protection, ensuring reliability even if the adjustment pin is accidentally disconnected [7]. Group 2: Performance and Practical Applications - The LM317 exhibits excellent performance as a linear regulator, with a typical line regulation of about 0.01% per volt and load regulation of approximately 0.1%, with ripple rejection rates nearing 80 dB when using appropriate capacitors [8]. - It has found extensive applications in laboratory power supplies, battery chargers, and audio equipment, providing clean power without the noise associated with earlier DC-DC converters [12]. - The LM317 has influenced the design of later adjustable regulators and modern switching controllers, establishing a standard where voltage selection is determined by circuit design rather than product catalogs [12]. Group 3: Longevity and Market Presence - Despite its age and relatively low efficiency by modern standards, the LM317 continues to be produced by multiple manufacturers, demonstrating its durability and reliability in various applications [12][13]. - The success of the LM317 is attributed to its straightforward solution to practical problems, establishing a trusted model for engineers and maintaining performance close to ideal conditions [13].

Core Viewpoint - TSMC's decision to exit the GaN foundry service by July 2027 is reshaping the GaN industry landscape, transitioning from reliance on advanced foundries to a focus on specialty process foundries [2][19]. Group 1: TSMC's Exit and Industry Impact - TSMC has been a crucial player in the GaN industry, being the only foundry capable of providing both high and low voltage GaN solutions [2]. - The exit of TSMC is prompting second-tier foundries to accelerate their capacity to fill the void left behind, leading to a reconfiguration of GaN manufacturing capabilities [2][5]. - GlobalFoundries (GF) has signed a GaN technology licensing agreement with TSMC, aiming to establish itself as a strategic GaN production center in the U.S. with over $80 million in federal funding [3]. Group 2: New Entrants and Strategic Moves - World Advanced (VIS), a TSMC subsidiary, is also entering the GaN market by expanding its GaN-on-Si capabilities, targeting mid-to-low margin orders previously handled by TSMC [3]. - Navitas, a major GaN customer of TSMC, is diversifying its supply chain by partnering with PSMC for 200mm GaN-on-Si production and strengthening ties with GF to mitigate manufacturing risks [4]. - ROHM is shifting from relying on TSMC to producing GaN devices in-house, establishing a new 8-inch wafer production line in Japan [7]. Group 3: Market Dynamics and Growth Projections - The GaN market is expected to grow significantly, with projections indicating a market size of approximately $3 billion by 2030 and a compound annual growth rate (CAGR) of 42% from 2024 to 2030 [10]. - The demand for GaN is expanding beyond consumer applications into high-reliability sectors such as data centers and electric vehicles, with automotive applications projected to grow at a CAGR of 73% from 2024 to 2030 [13]. - The shift in GaN applications is moving from consumer electronics to critical systems in data centers and automotive power supplies, emphasizing the need for reliability and efficiency [19]. Group 4: Structural Changes in the GaN Industry - The exit of TSMC is not a sign of declining GaN demand but rather a transition towards a decentralized manufacturing model, where multiple foundries share the production load [19]. - The industry is witnessing a redistribution of power, with IDM manufacturers regaining control over core processes and Fabless companies gaining more flexible manufacturing options [19]. - The GaN industry is evolving into a more independent and scalable sector, moving away from dependence on a single advanced foundry [19].

公众号记得加星标⭐️，第一时间看推送不会错过。 NOR闪存图 NAND 存储器以页为单位访问单元，而不是单个比特。要读取一个 NAND 单元，必须将序列中的其 他单元导通并充当传输晶体管。一个 NAND 块包含多个页，并且写入操作只需一次即可完成。 人工智能数据中心蓬勃发展带来的供应短缺已经影响到DRAM和NAND市场。现在，NOR闪存也受 到了波及。 据报道称，随着人工智能服务器需求的增长，这些服务器中使用的NOR闪存的需求也随之增长。这 些产品以及其他产品，例如嵌入式应用服务器，都使用NOR闪存来存储引导代码和固件、元数据配 置以及其他低延迟代码，这些代码可以直接执行，无需复制到DRAM。 《 商 业 时 报 》 称 ， 一 个 GPU 服 务 器 机 架 中 现 在 可 以 容 纳 多 达 30 个 NOR 设 备 ， 每 个 Nvidia GB200 NVL72系统的成本高达600美元，并预测几年内这一价格可能会达到900美元。 NOR 和 NAND 闪存以相同的方式将位存储在单元中，但访问单元的方式不同。 NOR 单元并联连接。每个单元的漏极直接连接到一条位线，源极共用一条公共线。这使得可以直接 随 ...

Core Viewpoint - MicroVision aims to reduce the production cost of its solid-state automotive lidar sensors to below $200, with a long-term goal of reaching $100, which would significantly broaden the application of lidar technology beyond high-end autonomous vehicles to advanced driver-assistance systems (ADAS) [2][3][4] Group 1: Cost Reduction and Market Impact - The current price of mechanical lidar systems ranges from $10,000 to $20,000, down from approximately $80,000, indicating a nearly tenfold decrease in costs [2][5] - The professor from Michigan State University suggests that further cost reductions of one to two orders of magnitude are feasible as demand shifts from fully autonomous vehicles to driver-assistance applications [5][6] - MicroVision's Movia S sensor features a fixed field of view of 180 degrees and can detect objects up to 200 meters away, which is less than the 300 meters typical of mechanical lidar [4][6] Group 2: Design and Integration Challenges - Solid-state lidar systems often have a smaller field of view compared to mechanical systems, necessitating the deployment of multiple sensors to achieve full coverage [5][6] - The integration of multiple sensors increases complexity, requiring precise alignment, calibration, and synchronization to ensure accurate data fusion [5][6] - MicroVision emphasizes that automotive manufacturers are designing complete perception systems rather than purchasing individual sensors, making overall system cost more critical than the cost of a single sensor [6][7] Group 3: Competitive Landscape and Industry Trends - Other companies, including Chinese firms and major suppliers like Luminar and Velodyne, are also targeting long-term cost goals below $500, but MicroVision's commitment to keeping prices below $200 is notable [7][8] - The cautious approach of some competitors reflects the structural challenge of achieving consumer-level pricing, which requires large and predictable demand to justify the necessary production investments [8] Group 4: Performance Evaluation and Future Considerations - There is a need for universally accepted metrics to evaluate the safety improvements provided by specific sensor configurations in ADAS and autonomous driving systems [9] - The focus is shifting from purely performance enhancements to economic benefits, with the potential for lidar prices to stabilize below $200 possibly altering the design decisions of automotive manufacturers [9][10] - If cost is no longer a primary objection, manufacturers must decide whether to forgo lidar based on technical or strategic considerations, which will significantly influence the integration of lidar into vehicle safety systems [9][10]

Core Viewpoint - Reliability has become a system-level issue, encompassing various aspects from materials and packaging to back power testing [2] Group 1: Chiplet and 3D-IC Architecture Challenges - The rapid adoption of chiplet-based architectures in data centers is driving significant changes in design, leading to rising costs and increasing reliability issues [3] - Reliability is the biggest challenge for chiplets and 3D integrated circuits (3D-ICs), as each chiplet may fail individually, necessitating a redesign to meet defect rate targets [4] - The introduction of chiplets and 3D-ICs brings new thermal-mechanical stresses that can affect overall system reliability [5] - Traditional departmental barriers are being broken down, forcing design teams to address material selection issues previously handled by foundries [5] Group 2: Thermal-Mechanical Stress and Reliability - Thermal-mechanical stress, caused by differing coefficients of thermal expansion (CTE) of materials, is a critical concern during chip assembly [12] - The complexity of reliability issues has increased with the introduction of new materials and packaging techniques, requiring chip designers to engage more in material management [13][14] - The reliability of chiplets and 3D-ICs is further complicated by diverse packaging processes and interconnect methods [10] Group 3: Process Technology and Circuit Reliability - Circuit reliability begins with process technology, focusing on the uniformity of FinFETs and nanosheet transistors [15] - Local and global variations in manufacturing processes can lead to significant reliability challenges, necessitating careful design considerations [15] - The need for lower defect parts per million (DPPM) is more critical than ever, especially in the design of standard cells and modules [15] Group 4: System-Level Integration and Verification - The integration of multiple chips in a system requires early consideration of packaging factors, differing from traditional SoC project lifecycles [17] - New EDA solutions and tools are needed to verify interconnect reliability, as failures can occur at any connection point [17] - As system complexity increases, the importance of a holistic approach to reliability management becomes evident [17] Group 5: Future Considerations and Industry Collaboration - The potential of chiplets to transform the semiconductor industry is significant, but they also introduce complex challenges that must be addressed from the early development stages [21] - Continuous collaboration and innovation are necessary to tackle cost and interface intellectual property concerns [21] - Robust verification methods must be prioritized to ensure seamless integration and long-term functionality of chiplet-based systems [21]

Core Viewpoint - The global semiconductor industry is crucial for modern technology, with TSMC, Samsung Foundry, and Intel Foundry as the three major players driving innovation and competition in advanced chip manufacturing [2][4][6]. Group 1: Company Profiles - TSMC is the absolute leader in pure foundry services, focusing solely on wafer manufacturing and avoiding competition in chip design, which has fostered strong relationships with companies like NVIDIA, AMD, Apple, and Qualcomm [2]. - Samsung Foundry represents a vertically integrated alternative, producing both chips and consumer electronics, and is investing heavily in advanced process nodes like 2nm GAA technology [3]. - Intel Foundry is a strategic newcomer, transitioning from a vertically integrated model to opening its fabs to external customers, with a roadmap that includes advanced nodes like Intel 18A [3]. Group 2: Competitive Dynamics - Competition among these companies drives technological advancement, as the need for innovation in transistor architecture and manufacturing processes is fueled by competitive pressure [5]. - The presence of multiple strong competitors enhances supply chain resilience, reducing risks associated with geopolitical tensions and natural disasters [5]. - Customers benefit from increased choice and bargaining power, allowing for better pricing and capacity allocation, which encourages foundries to respond actively to customer needs [5]. Group 3: Industry Implications - The competition among TSMC, Samsung, and Intel is not merely a commercial issue but has structural significance for the semiconductor ecosystem, ensuring innovation, resilience, and sustainable growth [4][6]. - The semiconductor industry is recognized as one of the most strategically important sectors globally, underscoring the necessity of these three companies for its success [6].

当全球能源转型进入深水区，功率电子作为能源转换与高效利用的核心载体，正迎来以宽禁 带半导体（SiC/GaN）为核心的技术革命。 2026年8月26日，深圳国际会展中心（宝安新馆）举办的PCIM Asia深圳展会，将成为这场革 命的重要见证者——由 半导体行业观察 倾力打造的 "破局与共生——宽禁带半导体引领功率 电子产业升级与智能应用革新" 专场论坛，不仅汇聚全球产业链顶尖力量，更将以技术深潜、 产业协同、趋势预判的多维视角，解构宽禁带半导体从技术突破到规模落地的核心命题，为 行业提供可落地、高价值的战略参考。 战略锚点： 宽禁带半导体重构全球功率 电子产业格局 01 在"双碳"目标与新能源革命的双重驱动下，功率电子的应用场景正从传统工业控制向新能源汽车、 储能系统、高压快充、AI数据中心等高端领域全面延伸，对器件的耐高压、耐高温、低损耗、高 密度特性提出了极致要求。 传统硅基半导体受限于物理特性，已难以满足800V高压平台、kW级功率密度、超高频开关等新一 代 应 用 需 求 ， 而 宽 禁 带 半 导 体 凭 借 其 优 异 的 材 料 本 质 （ SiC 禁 带 宽 度 3.26eV ， GaN 禁 带 ...