Core Viewpoint - The article introduces the embedded GPU (e-GPU), a configurable RISC-V GPU platform designed specifically for ultra-low-power edge devices (TinyAI), addressing the challenges of power consumption and area constraints in traditional GPU implementations [1][6]. Group 1: Introduction and Background - The increasing demand for real-time computing driven by machine learning is propelling the rapid development of edge computing, which enhances privacy and energy efficiency by processing data locally rather than relying on cloud servers [4]. - Specialized hardware architectures are required to meet the performance, real-time response, and power consumption limitations of these workloads, with heterogeneous architectures integrating CPUs and domain-specific accelerators being an effective solution [4][5]. - Traditional GPUs have not been thoroughly studied for their trade-offs in ultra-low-power edge devices, which typically operate under strict power constraints in the tens of milliwatts range [5][6]. Group 2: e-GPU Architecture and Features - The e-GPU architecture is designed to minimize area and power consumption while being adaptable to TinyAI applications, featuring a configurable design that allows for optimization of area and power [24][25]. - The memory hierarchy employs a unified architecture that maps the host's main memory and e-GPU global memory to the same physical memory, enhancing programmability and reducing data transfer complexity [26][27]. - A dedicated controller manages e-GPU operations, integrating power management features to monitor and control the power state of computation units [29]. Group 3: Performance Evaluation - The e-GPU configurations were tested using two benchmark tests: General Matrix Multiplication (GeMM) and TinyBio, demonstrating significant performance improvements and energy savings [48][49]. - The e-GPU system achieved speedups of up to 15.1 times and energy reductions of up to 3.1 times compared to baseline systems, while maintaining a power budget of 28 mW [2][58]. - The area of the e-GPU system ranged from 0.24 mm² to 0.38 mm², proving its feasibility for deployment in TinyAI applications, which typically have strict area constraints [50]. Group 4: Industry Context - Commercial edge GPU solutions, such as Qualcomm's Adreno and ARM's Mali GPUs, are not specifically designed for TinyAI applications, often exceeding the power requirements needed for these applications [11]. - Academic GPU research focuses on developing programmable and configurable architectures suitable for various computing domains, with the e-GPU proposed as a suitable solution for TinyAI workloads [12][13]. - The e-GPU platform is positioned as an open-source, configurable RISC-V GPU platform that addresses the programming limitations and energy efficiency needs of the TinyAI domain [12][13].

凤凰网财经讯 6月3日，三大指数集体低开，沪指低开0.22%，深成指低开0.34%，创业板指低开0.33%， EDA、汽车拆解、无人车辆等板块指数跌幅居前。 | | | | | 户深京重要指数 | | | | | | --- | --- | --- | --- | --- | --- | --- | --- | --- | | 名称 *● | | 最新 涨幅% | | 涨跌 | | 总手 | 现手 | 金额 | | 上证指数 | -0.22 | 3340.07 | -7.42 | 449/1547 | -0.07 | 4687 | 4687 | 54.63 乙 | | 深证成指 | -0.34 | 10006.13 | -34.50 | 488/2124 | -0.13 | 843万 | 843万 | 88.03 乙 | | 北证50 | | 1398.00 -0.76 | -10.69 | 74/170 | -0.44 | 13.6万 | 13.6万 | 2.74 乙 | | 创业板指 | | 1986.54 -0.33 | -6.65 | 254/1026 | -0.16 | 1957 | | 195万 ...

5.【LG电子在印度开建第三座制造工厂，未来四年将投资约6亿美元】LG电子印度公司5月8日宣布开始 在该国建设第三座制造工厂，尽管该公司推迟了在印度规模逾10亿美元的上市计划。这座位于安得拉邦 的工厂将生产冰箱、洗衣机和空调等多种产品，预计明年投产。声明称，LG电子计划未来四年内投资约6 亿美元吸引配套企业落户该地区，从而在安得拉邦打造白色家电制造生态系统。（科创板日报） 更多智能制造产业资讯 …… 扫码订阅 智能制造 产业日报， 精选行业新闻，帮你省时间！ 1.【成功构建三星星座，我国开启地月空间探索新纪元】近日，在地月空间DRO（远距离逆行轨道）探索 研究学术研讨会上，我国宣布成功构建国际首个基于DRO的地月空间三星星座。由我国部署研制的DRO- A/B两颗卫星在抵达并驻留地月空间远距离逆行轨道后，与先前发射的DRO-L近地轨道卫星建立起星间测 量通信链路。这标志我国正式开启地月空间探索新纪元。（经济日报） 2.【湖南山河SA750U大型无人运输机再创中国第一】5月7日，由湖南山河华宇航空科技有限公司自主研 制、山河星航实业股份有限公司战略协同推进的SA750U大型无人运输机，在陕西靖边成功完成国内首次 ...

Group 1 - The core viewpoint is that A-shares are expected to continue showing a warming risk appetite and thematic rotation, focusing on low institutional holdings and thematic trading opportunities [1] - The economic landscape is anticipated to face new variables by the end of Q2, particularly in the context of Sino-US economic relations [1] - Three major trends are highlighted: the unwavering trend of enhancing China's independent technological capabilities, the reconstruction of European defense and energy infrastructure, and the necessity for China to accelerate domestic demand through the "dual circulation" strategy [1] Group 2 - Short-term factors affecting A-shares include the resolution of negative Q1 reports, the TMT sector reaching a lower response model, and ongoing advancements in AI applications by major domestic and international companies [2] - In the medium term, a focus on neutral dividend combinations is recommended until significant rebounds in real estate or technological applications occur [2] - The current market is likely to favor a rotation and thematic investment approach due to uncertainties in reported earnings across various listed companies [2] Group 3 - The end of the performance verification period is expected to enhance the outlook for technology stocks, with a high probability of a short-term rebound led by the tech sector [3] - Consumption and technology are both seen as areas where expectations for growth are strengthening, with a current high profitability effect in consumption and a relatively low position in technology [3] - The report indicates a preference for investment opportunities in AI computing and embodied intelligence in the medium term [3] Group 4 - Despite ongoing trade tensions, Chinese assets are viewed as having better value, with a focus on sectors benefiting from domestic demand, such as tourism, food, and retail [4] - Recommendations include resource products and capital goods that will benefit from the restructuring of global economic order [4] - Low-valuation financial sectors are also suggested as a hedge against potential external shocks [4] Group 5 - The appreciation of the RMB is expected to boost Chinese assets, with AI becoming a key focus for investment in May [5][6] - The report emphasizes the potential of the domestic AI industry and applications, supported by high capital expenditure from overseas firms [5][6] Group 6 - The A-share market is expected to experience increased volatility, with a shift from small-cap growth to large-cap value stocks [7] - Recommendations include reducing exposure to AI sectors with low penetration rates and increasing allocation to structural tech growth areas with performance contributions [7] - The report suggests focusing on sectors like infrastructure and consumption that are expected to benefit from growth dividends [7] Group 7 - The technology growth style is returning as the market begins to shift following the resolution of prior performance and tariff disruptions [8] - The technology sector is seen as having reached a favorable valuation range, making it an attractive area for investment [8] - The report indicates that as pessimism fades, the tech sector is regaining its position as a focal point for capital [8] Group 8 - A-shares and Hong Kong stocks are becoming more resilient to external shocks, with macro policies expected to support market stability [9] - Key areas of focus include high-margin assets, the tech sector as a long-term investment, and consumer sectors benefiting from policy support [9] - The report suggests that Hong Kong stocks are currently undervalued and may benefit from expanding domestic demand policies [9] Group 9 - The A-share market is expected to demonstrate independence and resilience, with opportunities in technology, consumption, and certain cyclical sectors [10] - The report highlights the importance of performance improvement and policy alignment in the tech sector, particularly in TMT [10] - It also emphasizes the potential of sectors benefiting from rising domestic consumption expectations [10] Group 10 - The market may experience a controlled pullback due to tariff impacts, but the overall outlook remains positive with favorable domestic policies [11] - The report anticipates that the market will stabilize and potentially return to a bullish state by the latter half of the year [11] - A focus on value-oriented investments is recommended, particularly in themes related to growth and domestic substitution [12]

Industry Overview - The esports industry in China has matured, with significant breakthroughs in technologies such as 5G and AIGC. The domestic game market achieved a sales revenue of 302.96 billion yuan in 2023, marking a year-on-year growth of 13.95% and surpassing the 300 billion yuan mark for the first time. The market is projected to reach 324.64 billion yuan in 2024, indicating steady growth potential [1][20][25]. Industry Chain - The esports industry chain has formed a complete ecosystem covering hardware, content, events, and commercialization. The demand for GPUs is driven by the increasing use of high-end computing systems for graphics-intensive gaming applications. The GPU market in China reached 80.7 billion yuan in 2023, growing by 32.78% from 2022, and is expected to reach 107.3 billion yuan in 2024 and 120 billion yuan in 2025 [2][34][39]. Company Focus - Thunderobot Technology, the first esports equipment company listed on the Beijing Stock Exchange, focuses on integrating AI technology. The company has launched the flagship AI gaming laptop ZERO 18 Pro, equipped with high-performance hardware. In 2024, Thunderobot is expected to achieve a revenue of 2.955 billion yuan, reflecting a year-on-year growth of 15.65% [3][23][59]. Market Dynamics - The esports market is benefiting from the robust growth of esports games, the development of professional and amateur esports events, and the flourishing esports ecosystem. The number of registered esports-related companies in China exceeded 120,000 by the end of 2023, indicating intense competition within the industry [25][56].

② 财经网，从多个独立信源处获悉/今年一季度/腾讯向字节跳动购买了价值 约20亿元的GPU算力资源/这批资源以英伟达H20卡和服务器为主/腾讯元宝 目前的更新主要使用来自字节的卡/ ① 主板领涨， 天娱数科（+10.07%）/大位科技（+10.06%）/德生科技（+10.02%）/中电鑫龙 （+10.00%）/瀛通通讯（+10.00%）/ ②创业板领涨， 盛天网络（+9.47%）/先进数通（+9.14%）/优博讯（+8.40%）/赛意信息 （+5.83%）/硕贝德（+5.81%）/ ③科创板领涨， 迅捷兴（+11.18%）/神工股份（+6.37%）/中望软件（+6.32%）/杰华特 （+5.18%）/伟测科技（+5.02%）/ ④活跃子行业， SW 教育出版（+1.58%）/ SW 游戏（+1.30%）/ SW 数字芯片设计 （+0.57%）/ 国内新闻 ① 新浪财经，触觉感知与人形机器人企业帕西尼感知科技(PaXini Tech)拿 下中国新能源汽车龙头比亚迪超亿元战略投资/其6D霍尔阵列式触觉传感技 术被重点看好/在推动具身智能快速落地与多场景智能应用有无限潜力/这是 比亚迪本年度抛出的首笔重磅股权投资/ ...

Core Insights - Intel has made significant progress in the high-performance graphics sector, with the Xe3 hardware design completed and software development ongoing, indicating a strong commitment to future advancements [1][21]. GPU Organization - Modern GPUs are built on a hierarchical structure that allows for scalability in performance, power consumption, and pricing [3]. - The Xe3 architecture features a new topology that can support up to 16 Xe cores per render slice, compared to only 4 in previous generations [4][7]. Performance Metrics - Xe3 can theoretically support a configuration of 256 Xe cores with 32,768 FP32 channels, surpassing Nvidia's RTX 5090, which has 21,760 FP32 channels [7]. - The Xe3 architecture aims to enhance flexibility in scaling computational power without being constrained by fixed-function hardware [8]. XVE Changes - The Xe Vector Engine (XVE) in Xe3 can run up to 10 threads simultaneously, improving thread-level parallelism and latency hiding compared to previous generations [10]. - Xe3 introduces a more flexible register allocation system, allowing for better utilization of resources and higher occupancy rates [11]. Instruction Modifications - Xe3 supports new instructions for floating-point conversions and sparse matrix operations, indicating a focus on optimizing performance for specific workloads [14]. - The architecture includes enhancements for ray tracing, such as Sub-Triangle Opacity Culling (STOC), which aims to reduce unnecessary shader work during rendering [15][17]. Future Outlook - The Xe3 architecture represents a significant step forward for Intel's graphics capabilities, with improvements in latency tolerance and dynamic register allocation [21]. - Intel's ongoing research and development efforts suggest a commitment to advancing its GPU architecture, potentially positioning it more competitively against Nvidia and AMD [22].