证券代码（A/H）：000063/00763 证券简称(A/H)：中兴通讯 公告编号：202544 中兴通讯股份有限公司 关于按照《香港上市规则》公布 2025 年 4 月份证券变动月报表的公告 本公司及董事会全体成员保证信息披露的内容真实、准确和完整，没有虚假记载、误导 性陈述或重大遗漏。 中兴通讯股份有限公司根据《香港联合交易所有限公司证券上市规则》（简 称"《香港上市规则》"）规定，在香港联合交易所有限公司披露易网站 （www.hkexnews.hk）刊登了截至 2025 年 4 月 30 日的证券变动月报表。 根据《深圳证券交易所股票上市规则》关于境内外同步披露的要求，特将有 关公告同步披露如下，供参阅。 特此公告。 呈交日期: 2025年5月7日 I. 法定/註冊股本變動 | 1. 股份分類 | 普通股 | 股份類別 | H | | | 於香港聯交所上市 (註1) | 是 | | | --- | --- | --- | --- | --- | --- | --- | --- | --- | | 證券代號 (如上市) | 00763 | 說明 | | | | | | | | | | 法定/註冊股份數目 ...

Group 1 - The Zhongzheng Guangdong-Hong Kong-Macao Greater Bay Area Development Theme Index (931000) increased by 0.25% as of May 7, 2025, with notable gains from Huali Group (300979) up 4.64% and Guangdong Hongda (002683) up 4.52% [2] - The Greater Bay Area ETF (512970) rose by 1.02%, marking its third consecutive increase, with the latest price at 1.18 yuan and a total scale reaching 67.06 million yuan, a new high in nearly a month as of May 6 [2] - As of April 30, 2025, the top ten weighted stocks in the Zhongzheng Guangdong-Hong Kong-Macao Greater Bay Area Development Theme Index accounted for 53.26% of the index, including BYD (002594) and China Ping An (601318) [7] Group 2 - As of the end of 2024, only 46.34% of the market value of listed state-owned enterprises in Shandong Province experienced growth, with only one enterprise exceeding a market value of 100 billion yuan showing an increase [4] - The Shandong State-owned Assets Supervision and Administration Commission signed a memorandum of cooperation with the Shandong Securities Regulatory Bureau to encourage state-owned enterprises to enhance investment value through share buybacks, increased dividends, and other measures [4] Group 3 - The state-owned enterprise co-win ETF (159719) includes various stocks with notable weightings such as China Petroleum (601857) at 15.58% and China Shenhua (601088) at 2.91% [6] - The performance of major stocks in the state-owned enterprise ETF shows slight increases, with China Petroleum up 1.13% and China Mobile (600941) up 0.39% [6]

沪指5月6日上涨1.13%，申万所属行业中，今日上涨的有30个，涨幅居前的行业为计算机、通信，涨幅 分别为3.65%、3.59%。通信行业位居今日涨幅榜第二。跌幅居前的行业为银行，跌幅分别为0.13%。 通信行业资金流出榜 | 代码 | 简称 | 今日涨跌幅（%） | 今日换手率（%） | 主力资金流量（万元） | | --- | --- | --- | --- | --- | | 688313 | 仕佳光子 | 11.14 | 13.50 | -11185.56 | | 600941 | 中国移动 | 0.72 | 1.06 | -10553.64 | | 300570 | 太辰光 | 6.82 | 11.22 | -8451.44 | | 600522 | 中天科技 | 1.08 | 1.17 | -3375.26 | | 300502 | 新易盛 | 4.28 | 6.14 | -2352.35 | | 603602 | 纵横通信 | 1.87 | 3.56 | -2308.74 | | 601728 | 中国电信 | 1.17 | 0.17 | -1399.92 | | 688182 | 灿勤科技 | ...

在DeepSeek带动下，业内普遍认为，今年行业将迎来AI推理加速发展，据统计，目前市场上已经涌现 百款接入DeepSeek的智算一体机，但背后仍面临落地应用挑战。如何从基础能力到平台落地过程中全 面应对AI新趋势正成为重要命题。 近日举行的2025中兴通讯中国生态合作伙伴大会上，中兴通讯系统阐述了其在人工智能领域的战略规 划、政企业务的发展路径及智算能力应用落地等进展。 据第三方机构IDC预测，2028全球AI市场规模将突破8000亿美元，五年复合增长率超32%；此外，国内 运营商表态，2025年算力投资将保持20%以上的增长，互联网头部企业未来三年AI投资将超5000亿元 ——这意味着ICT基础设施建设正迎来广阔发展空间。 中兴通讯首席发展官崔丽对21世纪经济报道记者分析，近些年间，AI行业经历了诸多变化，从2023年 的"百模大战"到DeepSeek等开源模型兴起。"其实从技术发展角度看，这是正常现象，AI技术仍处在发 展早期。但DeepSeek推动AI应用门槛大幅下降，我们认为，AI会在越来越多场景中产生实际价值，将 推动智算中心建设和消费端业务增长，同时带动网络建设的增长。" 中兴通讯执行副总裁、首 ...

Group 1 - The rapid development of AI is increasingly emphasizing its application in home scenarios, expanding the market space from "network" to "network-computing-screen" [2] - The four key components of AI smart home products are AI home network, AI home computing, AI home smart screens, and AI home robots, which create a closed loop of "perception-computation-interaction-execution" [2] - The transition from single-point mechanical devices in the feature phone era to IoT-connected devices in the smart phone era, and now to AI-native devices that provide proactive services, marks a significant evolution in home technology [2] Group 2 - ZTE has launched the Wentiant series WiFi 7 routers, which support dual 2.5G ports and dual WAN functions, addressing the need for data processing capabilities in home computing products [3] - The current focus is on weak computing power products, with a strategy to handle local privacy computing issues through a cloud-based approach, offering a better experience at a lower cost [3] - The introduction of FTTR all-optical smart screens is part of ZTE's strategy to enhance home computing capabilities [3] Group 3 - ZTE has released a two-in-one cloud laptop and a 5G free screen, which leverage cloud computing to overcome the limitations of personal data and hardware devices [4] - The cloud computer market is rapidly growing, with a potential sales space of over 10 million units in China, redefining traditional computers by integrating cloud, network, AI, and terminal components [5] - The focus on cloud computing is expected to lead to a significant transformation in the PC industry, with the potential for low-cost sharing and high network support in home environments [5] Group 4 - ZTE is concentrating on the companionship and safety sectors within the AI home robot market, which is still in its early stages [5]

Core Viewpoint - The article discusses the development of a novel method for generating high-quality Chain of Thought (CoT) data, focusing on the adaptive difficulty grading of questions for large language models (LLMs) to enhance the reasoning capabilities of smaller models [2][6][41]. Group 1: Research Motivation and Challenges - The emergence of large models like DeepSeek-R1 (671 billion parameters) has highlighted the challenges of deploying such models in real-time systems and edge devices [6]. - There is a pressing need for research on smaller models with fewer than 7 billion parameters, particularly in complex reasoning tasks such as mathematical problem-solving and code generation [7]. - Current CoT data generation methods face challenges, including high computational and annotation costs associated with large-scale data-driven approaches and limited performance gains from high-quality sample-driven methods [8][9]. Group 2: Proposed Methodology - The article introduces a new method called "LLM Adaptive Question Difficulty Grading," which aims to improve the quality of CoT data by dynamically matching model capabilities with data difficulty [12][13]. - The method includes four key innovations: establishing a question difficulty grading system based on inherent model reasoning capabilities, creating an adaptive question bank, designing a difficulty distribution sampling strategy, and generating high-quality CoT data using DeepSeek-R1 [15][18]. Group 3: Experimental Results - The proposed method has shown significant improvements in reasoning performance across various model sizes, with accuracy increases ranging from 6.66% to 26.7% on the AIME24 mathematics competition dataset compared to traditional non-adaptive strategies [18][20]. - Detailed experimental results indicate that models trained with the adaptive CoT data outperform baseline models in multiple mathematical reasoning benchmarks, achieving up to 94.6% accuracy on MATH500 [37]. - The ZCode-32B model demonstrated superior performance across different difficulty levels, indicating that smaller models can achieve competitive results through adaptive data training [38]. Group 4: Conclusion and Future Work - The article concludes that the proposed framework for generating high-quality CoT data is efficient and effective, requiring only about 2,000 high-quality samples to significantly enhance model performance while reducing data and computational costs [41]. - Future work will focus on further integrating reinforcement learning to explore deeper reasoning capabilities and extending applications to more complex cross-domain tasks such as communication fault diagnosis [42].

香港交易及結算所有限公司及香港聯合交易所有限公司對本公告的內容概不負責，對其準確性或完整 性亦不發表任何聲明，並明確表示，概不對因本公告全部或任何部份內容而產生或因倚賴該等內容而 引致的任何損失承擔任何責任。 ZTE CORPORATION 中興通訊股份有限公司 （於中華人民共和國註冊成立的股份有限公司） （股份代號：00763） 須予披露交易 本公司第九屆董事會第四十一次會議及二零二四年度股東大會審議通過《二零二五年度擬使用自有 資金進行委託理財的議案》，同意授權本集團 2025 年度使用自有資金在銀行、證券公司、基金公司 等金融機構購買安全性高、流動性好、中低風險的理財產品，資金總額不超過 300 億元人民幣，在 此額度內可滾動使用，期限內任一時點的交易金額不超過上述投資額度，授權有效期自本公司二零 二四年度股東大會審議通過之日起至二零二五年度股東大會召開之日止。 茲提述本公司於 2024 年 9 月 9 日發佈的《須予披露交易-認購理財產品》，內容有關於 2024 年 5 月 9 日至 2024 年 9 月 9 日期間，本公司及附屬公司中興康訊共認購 6 筆杭州銀行理財產品，認購金額 合計為 70 億元 ...

Investment Rating - The report maintains a "Buy" rating for the company, indicating an expected stock price increase of over 15% relative to the industry index in the next six months [4]. Core Viewpoints - The company is experiencing steady revenue recovery, with a projected revenue of 124.25 billion yuan in 2023, slightly increasing to 150.12 billion yuan by 2027, reflecting a compound annual growth rate (CAGR) of approximately 7.5% from 2025 to 2027 [1]. - The net profit attributable to the parent company is expected to decline from 9.33 billion yuan in 2023 to 8.43 billion yuan in 2024, before gradually increasing to 9.46 billion yuan by 2027, indicating a focus on long-term growth despite short-term challenges [1]. - The company is undergoing a structural transformation, shifting from a primary focus on operator networks to a collaborative development model involving "network + computing power + terminals," with over 35% of revenue coming from the second curve business [2][3]. Financial Summary - Revenue is projected to decrease by 2.4% in 2024, followed by a recovery with a growth rate of 7.0% in 2025 and maintaining similar growth rates through 2027 [1]. - The gross margin is under pressure during the transformation phase, with a reported gross margin of 34.27% in Q1 2025, down 7.75 percentage points year-on-year, primarily due to intense competition in the computing power and terminal markets [2]. - The company anticipates that the computing power business will become a major growth driver, particularly with the launch of DeepSeek, which is expected to shift demand from training to inference, thus expanding the market size [2]. Earnings Forecast - The forecasted net profit for the company is 8.58 billion yuan in 2025, 8.88 billion yuan in 2026, and 9.46 billion yuan in 2027, with corresponding price-to-earnings (P/E) ratios of 17.5, 16.9, and 15.9 respectively [3][4].

Core Insights - The development of large AI models is driving a significant increase in demand for computing power, with a projected annual growth rate of 10 times as AI evolves from version 1.0 to 2.0 [1] - ZTE Corporation has established a dual-driven strategy of "Connectivity + Computing Power" to become a leader in the "network connectivity + intelligent computing power" sector, emphasizing the importance of AI integration in various industries [3][4] - The global AI market is expected to exceed $800 billion by 2028, with a compound annual growth rate (CAGR) of over 32%, indicating substantial investment opportunities in AI technologies [4] Industry Developments - ZTE's revenue for Q1 this year reached 32.97 billion yuan, a year-on-year increase of 7.8%, with over 35% of revenue coming from its second curve business focused on computing power and terminals [5] - The company is focusing on three key areas in AI: enhancing ICT infrastructure solutions, developing AI terminal applications, and promoting AI innovation for broader application [5][6] - ZTE has launched the AiCube DeepSeek intelligent computing integrated machine, which is already being applied in sectors such as education, healthcare, steel, and automotive [5] Challenges and Solutions - Industry clients face challenges such as data pollution, high costs of data collection, and long business cycle times for AI technology updates [9] - ZTE has upgraded its digital cloud platform to address these challenges, facilitating better integration of AI into business operations [9] - The introduction of the "Xingyun Intelligent Agent" marks a new phase in the deployment of enterprise-level intelligent agents, aimed at lowering the technical barriers for AI model commercialization across various industries [10]

Core Viewpoint - The forum focused on how artificial intelligence (AI) can empower the digital transformation of education and explore innovative development paths for "AI + Education" [1] Group 1: AI and Education Transformation - AI is reshaping the educational ecosystem, with digitalization being a key engine for high-quality educational development [3] - The Ministry of Education and nine other departments have issued opinions to accelerate educational digitalization, guiding collaborative innovation between schools and enterprises [3] - ZTE Corporation has planned three main paths: establishing an AI + Education special fund, creating a digital resource pool for education, and deepening industry-education integration [3] Group 2: Innovation and Competitions - The 2025 Second Education Information Technology Application Innovation Competition is set to begin, with ZTE hosting the AI-focused "Target Intelligent Detection Technology Application Practice Competition" [5] - This competition aims to address technical challenges in target detection, combining academic research with practical needs to accelerate technology application [5] Group 3: Industry Collaboration and Solutions - Representatives from leading companies discussed AI education, digitalization, and industry-education integration, launching the "Zhihai AI Education Integrated Machine" for smart teaching solutions [6] - ZTE officially launched the ZTE AI Certification, emphasizing its role as a builder of digital economy and its commitment to educational digitalization [6] - ZTE's "three-in-one" solution focuses on technology foundation, industry-education integration, and ecological co-construction [6] Group 4: Future Directions and Goals - ZTE is actively collaborating with the Ministry of Education on various initiatives, including talent cultivation and international professional standards [7] - The successful forum marks a significant step in promoting the integration of education and technology, contributing to the "Digital China Strategy" [7] - ZTE aims to accelerate the implementation of industry-specific large models and enhance AI talent development to achieve educational reform goals [7]