专题：凝心聚力共奋进 十五五开局创未来——2026全国两会财经特别报道 新浪科技讯 3月2日上午消息，2026年全国两会召开在即，新浪科技获悉，今年两会，第十四届全国政 协委员、360创始人、董事长兼CEO周鸿祎将主要围绕AI安全、AI落地、AI应用、AI培训等以AI为核心 的多方面展开。 关于AI赋能安全，当前Anthropic等已通过AI编程、AI查找漏洞把很多原来安全上不能解决的问题给解 决了，因此，周鸿祎建议关注AI智能体。据他介绍，360已经做了几十种上万个AI安全智能体，这些智 能体能够实现AI挖掘软件漏洞、自动运行，同时能智能抵御其它国家的黑客智能体，包括用AI智能体 来解决AI的安全问题。现在中国有两百万家中小企业已经在用360的AI智能体来帮他们的企业安全做实 时防护。 关于AI赋能安全，当前Anthropic等已通过AI编程、AI查找漏洞把很多原来安全上不能解决的问题给解 决了，因此，周鸿祎建议关注AI智能体。据他介绍，360已经做了几十种上万个AI安全智能体，这些智 能体能够实现AI挖掘软件漏洞、自动运行，同时能智能抵御其它国家的黑客智能体，包括用AI智能体 来解决AI的安全问题。现在中 ...

英伟达整合LPU（语言处理单元）技术、OpenAI多线押注推理芯片，正在将AI算力竞争的主战场从训练切换至推理。申万宏源研究认为，2026年 算力产业的核心关键词将是推理，Token消耗总量与技术范式均将围绕这一主题深度重构。 2月28日，据《华尔街日报》报道，英伟达计划在下月的GTC开发者大会上发布一款整合了Groq"语言处理单元"（LPU）技术的全新推理芯片，英 伟达首席执行官黄仁勋称其为"世界从未见过"的全新系统。OpenAI已同意成为该处理器的最大客户之一，并将向英伟达购买大规模"专用推理产 能"。 与此同时，OpenAI上月还与初创公司Cerebras达成数十亿美元计算合作，后者称其推理芯片速度已超越英伟达GPU（图形处理器）。这一系列动 向表明，AI巨头正在从训练算力的军备竞赛，转向推理算力的多线布局。 申万宏源报告指出，Token经济时代，推理算力正迎来四大趋势：一是纯CPU（中央处理器）部署场景增多，低成本推理需求加速算力下沉；二 是LPU等专用架构崛起，挑战GPU在推理环节的主导地位；三是国产算力芯片加速突破，供应链多元化趋势明确；四是推理算力的需求结构从"单 次训练"向"海量Token消耗 ...

Core Viewpoint - Nvidia's stock has been stagnant, with a slight increase of 1.7% since Q4 of last year, underperforming the S&P 500 by 3.3% during the same period, raising concerns about whether strong earnings will be sufficient to meet market expectations [1][3] Group 1: Earnings Expectations - Nvidia is set to release its highly anticipated Q4 and annual earnings report, with Wall Street consensus expecting strong performance that may exceed analyst predictions [1] - Investors are facing a "expectation paradox," as historical data shows Nvidia's stock has faced sell-offs following its last two earnings reports, despite strong earnings [3] - Concerns exist that even if Nvidia's official earnings and guidance are solid, they may not meet heightened expectations, leading to potential stock declines [3] Group 2: Market Sentiment and Competition - The technology sector, particularly the "Magnificent Seven," has seen a nearly 1% decline since Q4 of last year, underperforming the S&P 500, reflecting cautious investor sentiment regarding AI capital expenditures translating into actual profits [4] - Nvidia's forward P/E ratio has dropped below 24, nearing a five-year low and significantly lower than its five-year average of 38, which may present a buying opportunity [7] - The competitive landscape is shifting as companies like AMD, Amazon, Broadcom, and Alphabet introduce chips for generative AI models, raising questions about Nvidia's ability to maintain market share [7] Group 3: Macro Environment - The macroeconomic environment for 2026 is uncertain, with geopolitical tensions and mixed economic data contributing to market volatility [5][6] - Recent economic indicators show a slowdown in growth and persistent inflation, leading traders to bet that the Federal Reserve will adopt a cautious stance on further rate cuts [5]

Core Insights - The rise of AI agents is expected to significantly impact the chip industry, with a focus on inference capabilities becoming paramount as companies like Nvidia invest heavily in this area [1][2] - Aixin Yuanzhi has emerged as a leading player in the edge and endpoint inference chip market, recently becoming the first Chinese edge AI chip company to be listed on the Hong Kong Stock Exchange [1][5] Industry Trends - The demand for AI chips is shifting from training to inference, with a projected compound annual growth rate (CAGR) of 31.0% for global AI inference chips from 2024 to 2030 [5][6] - The edge inference segment is expected to grow at a CAGR of 42.2%, indicating a substantial market opportunity [5] Company Positioning - Aixin Yuanzhi's unique "dual-track development model" focuses on both vertical IP core technology upgrades and horizontal application expansion, supported by its proprietary AXNeutron NPU and AXProton AI-ISP [3][4] - The AXNeutron NPU is designed to address the "impossible triangle" of performance, power consumption, and cost, achieving a throughput improvement of up to 10 times compared to traditional GPU-based solutions [4] Market Performance - Aixin Yuanzhi is projected to ship over 900 million chips in 2024, capturing a market share of 6.8%, and leading the mid-to-high-end chip segment with a 24.1% share [6][8] - The company ranks third in the domestic edge AI market, with an expected shipment of 100,000 units in 2024 and a market share of 12.2% [6] Future Outlook - The global market for edge inference chips is forecasted to reach 726.2 billion yuan by 2030, while endpoint inference chips are expected to reach 886.1 billion yuan, totaling over 1.5 trillion yuan [6] - Aixin Yuanzhi aims to leverage its high-performance, cost-effective platform capabilities to strengthen its position in the AI perception and edge computing sectors, potentially reshaping the global edge computing landscape [8]

Core Viewpoint - Qualcomm warns that rising memory prices will slow down growth in the smartphone industry, causing a significant drop in its stock price by 11% [2] Group 1: Financial Performance - Qualcomm reported record revenue of $12.3 billion for Q1 2026, driven by strong sales of high-end smartphones and growing interest in smart glasses, automotive, and IoT products [2] - The company forecasts Q2 revenue to be between $10.2 billion and $11 billion, down from $11 billion in the same quarter last year, with smartphone chip sales expected to decline from $6.9 billion to $6 billion [3] Group 2: Market Dynamics - The cautious approach of smartphone manufacturers does not indicate a decline in market demand but rather reflects concerns over insufficient memory supply, leading to reduced production plans [3] - Qualcomm's CEO believes that the current turmoil will result in short-term losses but does not foresee long-term difficulties in the market [3] Group 3: Future Prospects - Qualcomm is diversifying its revenue streams through developments in robotics, automotive, and patent licensing, aiming to reduce its dependence on smartphone revenue by 2029 [4] - The company is also venturing into AI chip development, with initial deliveries made to its confirmed customer, Humane, and expects revenue from AI chips to materialize next year [3]

Summary of Conference Call Notes Industry Overview - The document discusses the trend of price increases across various sectors, particularly in technology and cloud services, indicating a widespread inflationary environment driven by demand and supply dynamics [1]. Key Points - **Price Increases**: - CPU prices are rising, impacting companies like **Haiguang Information** [1]. - GPU rental prices are increasing, affecting firms such as **Xiechuang Data** [1]. - Cloud service prices are on the rise, with **Kingsoft Cloud** being a notable example [1]. - CDN (Content Delivery Network) prices are also increasing, with **Wangsu Technology** mentioned [1]. - **Market Sentiment**: - There is a consensus to remain optimistic and to follow the trend of price increases, suggesting a bullish outlook on the affected sectors [1]. - **Future Trends**: - Potential future price increases may extend to power supply and SaaS (Software as a Service) sectors [1]. - **Underlying Drivers**: - The root cause of these price increases is attributed to the booming demand for agents and application software [1]. - **Company Focus**: - Attention is drawn to **CloudWalk Technology**, which has shifted to self-developed inference chips since last year, indicating a strategic pivot in supply [1]. - There is an expectation that during the current cycle of rising computing resource prices, long-tail demand may partially shift towards second-tier companies [1]. - **Alpha Expectations**: - There is anticipation regarding the successful rollout of self-developed inference chips in the near future [1].

Core Insights - The integration of AI inference chips is becoming crucial for automotive intelligence as autonomous driving approaches [2][10] - The demand for inference chips is expected to significantly increase by 2026 due to the rapid growth of automotive intelligence needs [3] Inference Demand Surge - AI model training has been a key growth driver for the AI chip market, with high-end chips like NVIDIA's H100 and H200 being highly sought after, often resulting in multi-million dollar orders [4] - Inference chips have now surpassed training chips in demand, becoming the new mainstay for data center computing power and smart driving applications, as companies focus on translating large models into practical applications [4][5] Automotive Intelligence Key to Success - Autonomous vehicles are evolving into highly integrated "smart mobile terminals" that require real-time decision-making capabilities, supported by the powerful computing power of inference chips [6] - A Level 4 autonomous vehicle can generate data volumes of several gigabytes per second, necessitating rapid processing and analysis for effective driving decisions [6][7] Performance and Efficiency of Inference Chips - Inference chips are designed for edge computing, allowing for immediate data processing without relying on cloud transmission, which is critical for timely decision-making in autonomous driving [7] - New generation inference chips utilize advanced architectures and manufacturing processes, such as 7nm technology, to provide high performance while significantly reducing energy consumption [8] Customization for Autonomous Driving - Inference chips must be tailored for core tasks in autonomous driving, such as visual recognition and decision control, through customized neural network accelerators to enhance processing efficiency and accuracy [9] Industry Transformation with Inference Chips - Inference chips represent a pivotal point in AI industry development, acting as a bridge from research to market application and playing an essential role in automotive intelligence [10] - Achieving automotive-grade certification is a significant hurdle for inference chips, requiring rigorous environmental testing to ensure reliability and stability throughout the vehicle's lifecycle [10][11] Challenges and Future Outlook - Algorithm adaptation is a key challenge for inference chips in automotive applications, necessitating close collaboration between chip manufacturers and automotive companies to optimize performance [11] - The rise of inference chips marks a new phase in the AI and autonomous driving industry, addressing core issues such as cost, latency, and privacy, and enabling deeper integration of AI technologies into operational contexts [11][12] - As AI technology and automotive hardware converge, the future application prospects for inference chips will expand, with increasing competition among automotive companies to develop more competitive autonomous driving solutions [12]

Core Insights - Shanghai is emerging as a leading hub for the semiconductor and artificial intelligence industries in China, with significant growth projections for both sectors by 2025 [1][7][13] - The rapid expansion of the general GPU market in China is driven by the increasing demand for computing power, particularly due to the rise of AI models [3][6] - A number of innovative AI companies have recently gone public, highlighting the vibrant "Shanghai tech sector" [6][11] Industry Overview - The integrated circuit industry in Shanghai is expected to achieve a revenue scale exceeding 488 billion yuan by 2025, doubling in five years [1] - The artificial intelligence industry in Shanghai is projected to exceed 550 billion yuan, with a growth rate surpassing 30% [1] - Shanghai ranks fourth globally and first in China in the latest global integrated circuit industry competitiveness ranking [8] Company Developments - Wallen Technology, the first GPU company listed on the Hong Kong stock market, has set a record for the largest IPO since the implementation of the special technology company listing mechanism [3] - TianShu Intelligent Chip is expected to deliver approximately 15,000 general GPU units in the first half of 2025, covering various sectors including finance and healthcare [3] - MiniMax, a rapidly growing AI company, has achieved significant milestones in video model development, generating over 590 million videos [4][5] Investment and Ecosystem - Shanghai has established a comprehensive investment ecosystem, including a fund matrix with a scale of 105 billion yuan to support core technology breakthroughs and industry chain organization [11][12] - The city is fostering a collaborative environment for AI companies, with over 200 AI enterprises having settled in the "Mosu Space" innovation community [9][10] - The government is actively supporting innovation by facilitating pilot projects for new products, which is crucial for early-stage product iteration [12]

Core Viewpoint - The artificial intelligence industry is undergoing a transition from 2025 to 2026, with Yuntian Lifei focusing on reasoning chips and smart hardware in response to industry changes [1] Company Insights - Yuntian Lifei's Chairman and CEO, Chen Ning, highlighted the company's commitment to deepening its involvement in the reasoning chip and smart hardware sectors [1]

该行称，是否批准所有云服务商的H200芯片采购仍存在不确定性，因可能影响国产芯片普及，但预计 云服务提供商将结合H200与本地芯片使用，尤其在推理需求上。 摩根士丹利发布研报称，看好中芯国际(00981)，因其是"中国人工智能本地化的关键支持者"，且因先 进逻辑芯片需求将保持强劲，并对中国芯片设备企业持积极看法。 大摩预计H200有助于满足中国训练需求，并可能带来更多对推理芯片的需求，从而扩大市场需求。因 此，该行预计AI资本支出将有所增长，亦料中国芯片自给率预计将由2024年的24%提升至2027年的 30%。 ...