Core Viewpoint - NI is leading the way towards a data-driven and AI-enabled future in intelligent testing, showcasing its smart testing strategy and significant updates to core products like LabVIEW, PXI, and DAQ at the recently concluded NI Connect 2025 [1][2]. Group 1: Smart Testing Strategy - NI's smart testing strategy focuses on platform-based testing solutions, emphasizing open software, modular hardware, and ecosystem collaboration to tackle next-generation testing challenges [2]. - The company plans to enhance user efficiency and cross-team collaboration through continuous investment in LabVIEW+ suite and SystemLink software platform [2]. - NI aims to push the performance limits of its hardware products like PXI and DAQ while reducing costs to meet diverse needs [2]. Group 2: Product Innovations - The newly launched 3rd generation PXIe-5842 Vector Signal Transceiver (VST3) doubles the instantaneous bandwidth from 2GHz to 4GHz, with potential for up to 7GHz using proprietary spectrum stitching technology [3][4]. - NI plans to release over 25 new data acquisition products in 2025, including the FieldDAQ designed for harsh environments with IP67 protection and wide temperature adaptability [4][6]. - The USB DAQ series NI mioDAQ, launched last year, features 1 million samples per second and 20-bit resolution, gaining popularity for its ease of use [6]. Group 3: Software Enhancements - NI LabVIEW 2025 introduces optimized debugging features and enhanced interoperability with .NET 8.0 and Python 3.11/3.12, improving collaboration among diverse technical teams [8]. - The new version of SystemLink software simplifies setup and licensing for small labs, enabling low-cost deployment and advanced data strategies [10]. - AI advisor Nigel™ has been integrated into LabVIEW and TestStand, providing significant convenience for users [12]. Group 4: Future Directions - NI is committed to building an intelligent testing platform that integrates AI, cloud computing, and advanced technologies to manage complexity and enhance system performance [15][16]. - The company emphasizes the importance of a structured data strategy to realize the value of AI in testing systems [15]. - Future testing systems will focus on AI-driven advancements, enhancing productivity and speed through integrated data management and visualization [16].

Core Viewpoint - Nvidia holds a dominant position in the AI semiconductor field, with an estimated market share exceeding 80% in data center chips and products like ChatGPT and Claude [1] Group 1: Nvidia's Market Position - Nvidia's leadership in AI computing can be traced back nearly two decades, with the development of its CUDA software stack [1] - Despite being in a loss-making position for much of its early years, Nvidia's CEO Jensen Huang recognized the potential of GPUs for AI [1] - Nvidia's products now dominate a significant portion of global AI applications [1] Group 2: Competitors - AMD is Nvidia's largest competitor in the data center AI computing market, launching its MI300 GPU in 2024, over a year later than Nvidia's second-generation data center GPU [3] - Analysts predict AMD's market share to be less than 15%, but the company is focused on improving its software capabilities [3] - Custom-designed chips (ASICs) are emerging as a challenge to Nvidia, with a projected market size doubling by 2025 [4] - Major companies like Amazon and Google are designing their own chips, such as Amazon's Trainium and Google's TPU, to provide cheaper alternatives for AI workloads [6] Group 3: Emerging Threats - Huawei is considered a significant competitor to Nvidia, with reports indicating that its AI chip innovations are catching up [9] - Numerous startups are also challenging Nvidia with new chip designs and business models, including companies like Cerebras and Groq [11]

Core Viewpoint - The article discusses the advancements in photonic chips as a potential replacement for traditional electronic microchips, particularly in the context of increasing demands for computational power driven by artificial intelligence (AI) [1][2]. Group 1: Photonic Chips Advantages - Photonic chips utilize light (photons) instead of electricity (electrons) for information processing, promising higher speed, greater bandwidth, and improved efficiency due to the absence of electrical resistance and heat loss [1]. - They are particularly well-suited for matrix multiplication, a fundamental operation in AI [1]. Group 2: Challenges in Photonic Computing - Converting photons to electrical signals can slow down processing times, and photonic computing relies on analog rather than digital operations, which can reduce precision and limit the types of computations [2]. - The current inability to manufacture large-scale photonic circuits with sufficient precision complicates the transition from small prototypes to scalable solutions [2]. Group 3: Recent Research Developments - A new photonic processor called the Photonic Arithmetic Computing Engine (Pace) was developed by Lightelligence, featuring over 16,000 photonic components and demonstrating low latency and practical application viability [2][3]. - Another photonic processor from Lightmatter was shown to operate with precision comparable to traditional electronic processors, successfully executing AI tasks such as text generation and game playing [3]. Group 4: Future Potential - Both research teams believe their photonic systems could become part of scalable next-generation hardware to support AI applications, although further improvements in materials and design are necessary [3].

Core Viewpoint - TSMC's expansion in the U.S. is driven by increasing demand from major clients like Apple, NVIDIA, AMD, Qualcomm, and Broadcom, influenced by geopolitical factors and the need for alternative production sites [1][2]. Group 1: TSMC's U.S. Expansion - TSMC's new U.S. factories are seeing strong demand, with reports indicating that the capacity of the upcoming third factory has already been booked by clients [1]. - The company plans to invest a total of $165 billion in its U.S. operations, which will include six new wafer fabs and two advanced packaging facilities [1]. - By 2028, TSMC's overseas capacity is expected to reach about 20% of its total capacity, with a significant portion coming from its U.S. and Japanese facilities [1]. Group 2: Client Demand and Revenue Projections - TSMC's revenue from Apple is projected to reach between NT$800 billion and NT$1 trillion this year, with a year-on-year growth rate of up to 60% [4]. - Apple is TSMC's largest client in the U.S., with plans to procure over 19 billion chips across multiple states by fiscal year 2025 [4]. - TSMC's advanced process technology is crucial for Apple's innovation, as the company transitions from Intel processors to its own M-series chips, with expectations for the new M5 chip to utilize TSMC's N3P process [5]. Group 3: Advanced Process Technology - TSMC's advanced process technology now accounts for over 73% of its revenue, with significant contributions from 7nm, 5nm, and 3nm processes [5]. - The company is helping clients like Apple reduce costs through innovative semiconductor solutions, enhancing competitiveness in the market [6].

Core Viewpoint - The article discusses advancements in lithography, mask, and Optical Proximity Correction (OPC) technologies, emphasizing their critical role in improving power-performance-area-cost (PPAC) metrics for advanced logic chips. It highlights the shift from 2-D Manhattan layouts to 1-D designs and the introduction of curvilinear shapes to enhance manufacturing efficiency and reduce costs [1][2][6]. Summary by Sections Lithography and Design Advances - The evolution of lithography techniques, including immersion lithography and extreme ultraviolet (EUV) lithography, has significantly improved resolution, which is essential for meeting the demands of advanced logic chip design [2][4]. - The transition from 2-D to 1-D Manhattan layouts allows for denser representations but introduces challenges such as increased costs and longer current paths due to additional vias [2][10]. Curvilinear Design Concepts - The introduction of curvilinear shapes in mask writing aims to reduce errors during the transfer from design to wafer, addressing the inherent distortions caused by current lithography methods [6][7]. - imec proposes integrating curvilinear geometries at the design stage, which could lower manufacturing costs while enhancing electrical performance, representing a significant innovation in the semiconductor industry [7][10]. Use Cases for Curvilinear Design - **Use Case 1**: Curvilinear design can simplify middle-of-line (MOL) and back-end-of-line (BEOL) layers, potentially reducing wafer costs by 7% and improving performance by approximately 5% by eliminating unnecessary vias [10][12]. - **Use Case 2**: In CMOS devices, curvilinear designs can connect source/drain contacts and gates without additional metal layers, leading to a 20% reduction in area for standard cells [12][13]. - **Use Case 3**: Curvilinear geometries in layout routing are expected to have the most significant impact, requiring substantial industry investment but promising to enhance power-performance-area-cost (PPAC) metrics across future logic technology nodes [13][14]. Challenges and Future Directions - Implementing curvilinear designs poses challenges, including the need for specialized design rules and verification methods to manage increased data volumes in the manufacturing ecosystem [14][17]. - The potential of curvilinear designs to optimize high numerical aperture EUV lithography could benefit various applications, including image sensors and automotive chips, by reducing manufacturing costs [17].

Core Viewpoint - The Philippines possesses significant potential in the global semiconductor industry but risks missing opportunities due to slower development compared to regional competitors like Vietnam and Thailand [1][2]. Group 1: Market Growth and Opportunities - The global semiconductor market is projected to reach $167.7 billion in sales by Q1 2025, marking an 18.8% increase from the same period in 2024 [4][5]. - By 2025, the total revenue of the semiconductor market is expected to exceed $600 billion, with a year-on-year growth rate of 10-15%, driven by demand in AI, 5G, automotive, and cloud computing sectors [5][6]. - The memory semiconductor market alone is anticipated to surpass $250 billion by 2025, fueled by the growing need for data storage and processing [5]. Group 2: Challenges Facing the Philippines - The Philippines is lagging behind regional competitors, particularly Vietnam, which has been actively supporting its domestic industry to explore new investment opportunities [1][6]. - The country's share in assembly, testing, and packaging (ATP) has been declining, and the previous advantage of low labor costs is diminishing due to rising living costs [6][6]. - There is a significant shortage of over one million technical talents in the global semiconductor industry, presenting an opportunity for the Philippines to fill this gap by developing local engineering talent [4][5]. Group 3: Government Support and Education - The Philippine government has identified semiconductors as a priority industry but has yet to demonstrate effective actions to support it [2][3]. - Outdated educational curricula are a critical bottleneck, as they do not align with industry needs, and there is a lack of awareness regarding the potential of the integrated circuit design sector [2][3]. - Strengthening government support for the semiconductor industry could enhance capabilities in key areas such as military and infrastructure, ensuring continuity even during global supply chain disruptions [2][3].

Core Viewpoint - The article highlights the upcoming automotive electronics technology seminar hosted by Naxin Micro on May 15 in Guangzhou, focusing on various applications in the smart electric ecosystem, including body control, lighting, intelligent cockpit, and safe driving systems [1]. Group 1: Event Details - The seminar will take place on May 15 from 13:30 to 18:00 at the Guangzhou Zhengjia Plaza Marriott Hotel [2]. - There will be two sub-forums: one focusing on body control and lighting system applications, and the other on intelligent cockpit and safe driving system applications [2][3]. Group 2: Sub-Forum Focus Areas - The first sub-forum will address innovations in lighting efficiency, integrated SoC applications, domestic motor drives, and magnetic sensors in various scenarios within the body domain [2]. - The second sub-forum will cover key solutions such as one-stop power management, Class D audio amplifiers, SerDes high-speed transmission, ultrasonic sensors, and operational amplifiers with power interface systems [3].

最近些年。RISC-V引起了全球关注。这款革命性的 ISA 凭借其持续的创新，以及无数的学习 和工具资源以及来自工程界的贡献，像潮水般席卷了市场。RISC-V 最大的魅力在于它是一款 开源 ISA。 在本文中，我（指代本文作者Mitu Raj，下同）将介绍如何从零开始设计一款RISC-V CPU ，我们将讲解定义规格、设计和改进架构、识别和解决挑战、开发 RTL、实现 CPU 以及在 仿真/FPGA 板上测试 CPU 的流程。 以下为文章正文： 从命名开始 如果您希望可以时常见面，欢迎标星收藏哦~ 为你的想法命名或打造品牌至关重要，这样才能激励你不断前进，直至达成目标！我们打算构建一 个非常简单的处理器，所以我想出了一个花哨的名字" Pequeno "，在西班牙语中是"微小"的意 思；完整名称是：Pequeno RISC-V CPU，又名PQR5。 RISC-V 的 ISA 架构有多种风格和扩展。我们先从最简单的RV32I开始，它又称为 32 位基本整数 ISA。该 ISA 适用于构建支持整数运算的 32 位 CPU。因此，Pequeno 的第一个规格如下： Pequeno 是一款 32 位 RISC-V C ...

英特尔DeepLink，是什么？ 如果您希望可以时常见面，欢迎标星收藏哦~ 来源：本文 编译自tomshardware ，谢谢。 英特尔已停止对其 Deep Link 技术套件的支持，GitHub上的一位代表通过 X 用户Haze 证实了这 一点。继英特尔悄然停止在 Battlemage 等新产品中推广该功能后，该公司现已确认 Deep Link 的 积极开发已停止。虽然您仍然可以使用 Deep Link，但英特尔已澄清，未来将不会提供任何更新， 其客户服务渠道也不会提供官方支持。 Deep Link 于 2020 年底推出。它允许您充分利用 Intel CPU 和 Arc GPU 的强大功能，从而提升 流媒体播放、AI 加速和整体效率。要使用 Deep Link，您需要 Intel 第 11、12 或 13 代 CPU 和 专 用 的 Arc Alchemist GPU 。 该 套 件 提 供 四 个 关 键 实 用 程 序 ： 动 态 功 率 共 享 (Dynamic Power Share)、流辅助 (Stream Assist)、超级编码 (Hyper Encode) 和超级计算 (Hyper Com ...

Core Viewpoint - The "12th Automotive Electronics Innovation Conference and Automotive Chip Industry Ecosystem Development Forum (AEIF 2025)" will be held in Shanghai on May 14-15, focusing on cutting-edge, key, and disruptive technological breakthroughs in the automotive sector [1]. Group 1: Conference Overview - The conference will feature 1 summit forum, 1 supply-demand matching session, 3 thematic forums, and 1 product exhibition, with over a thousand attendees expected [2]. - The organizing committee aims to enhance connections between upstream and downstream players, expanding the coverage of vehicle manufacturers and parts suppliers, and providing more diverse showcasing opportunities for capable automotive chip companies [2]. Group 2: Agenda Highlights - The agenda includes various presentations on automotive chip solutions, such as high-performance AI cockpit systems, vehicle storage introductions, and applications of DSP chips in cabin audio [4][6][7]. - Notable speakers include industry leaders from companies like 瑞芯微电子, 普冉半导体, and 苏州国芯科技, discussing topics ranging from domestic chip solutions to advanced automotive technologies [4][6][7]. Group 3: Thematic Forums - The thematic forums will cover topics such as the automotive electronics industry ecosystem, smart connected and electric vehicles, and AI and autonomous driving [11][15][17]. - Key discussions will include the challenges and opportunities in the software-defined vehicle era, chip testing for automotive electronic quality, and the role of RISC-V architecture in promoting chip autonomy [12][14][18].