3月26日，商汤科技发布2024年财报，总体收入同比增长10.8%，亏损同比大幅收窄33.7%，其中生成式 AI增长103.1%，营收占比从2023年的35%来到64%[1]。 考虑到时下生成式AI"退潮裸泳"的大背景，这种表现令人侧目。新旧势力纷纷遭遇"落地墙"围堵，商汤 却像是每个班都有的学霸，别人还在回顾知识点，ta已经开始刷题了。 在AI公司中，商汤的学霸光环向来耀眼。提到中国AI产业，商汤是绝不会忽略的一家公司。出道至今研 发实力从未掉出过第一梯队，研发成果如数家珍。但光环背后，一度是面包和理想的长久拉锯。 2018年—2022年间，商汤开始投入基础设施建设，研发费用率由此飙升，背负起巨大的盈利压力，但也 由此生长出了独一无二的商业模式。如今守得云开见月明，是科技产业对长期主义者的又一次褒奖。 从"大装置"开始 上海临港，距离特斯拉超级工厂3公里，是商汤SenseCore AI大装置所在地——商汤智算中心AIDC，后 者是当前亚洲最大的超算平台之一[2]。 商汤智算中心AIDC AIDC正式启动于2021年，SenseCore AI大装置的打造则从更早之前就开始了，是商汤埋的最深的一个伏 笔。 20 ...

目前，平台已与上海三大运营商以及多家云算力企业达成初步合作意向，并探索接入西部对口合作地区优质算力。 上海智能算力科技有限公司总经理孙跃介绍称，上海市智能算力资源统筹调度服务平台以社会闲散算力纳管交易为核心，通过"一云多芯、一云多池"灵 活调度的云服务平台，为垂类行业应用、开源社区、AI研究者、个人开发者等提供高性价比普惠算力服务。目前，平台已首批接入上海仪电、上海电信、 上海移动、上海联通、商汤科技、碳和科技等互联网企业算力、云计算服务商、电信运营商等。 现场还颁发了2024年度上海市算力网络高质量发展标杆应用案例、上海市智算中心综合能效卓越奖和绿色技术单项成就奖。其中，商汤临港智算中心获 得标杆应用案例一等奖。 近日，2025年"智算申城"高峰论坛在上海召开。会上，上海市智能算力资源统筹调度服务平台上线。 上海市经济和信息化委员会主任张英在致辞中指出，上海正顺应智能化变革机遇，将人工智能作为重点发展的三大先导产业之一，引领制造业数智化转 型，带动构建具有竞争力的智算产业生态。面向未来，上海将充分利用超大城市的综合优势，加快建设更具国际影响力的人工智能"上海高地"。 现场，上海市智能算力资源统筹调度服务平台 ...

Investment Rating - The report maintains a "Buy" rating for the company [1] Core Insights - The company reported a revenue of 3.772 billion CNY for 2024, representing a year-on-year growth of 10.8%, while the net profit attributable to shareholders was a loss of 4.278 billion CNY, narrowing from a loss of 6.440 billion CNY in the previous year [4][6] - The growth in generative AI revenue was 103.1% year-on-year, slightly exceeding the previous expectation of 100%, while traditional AI revenue declined by 39.5% [6] - The company is transitioning successfully to generative AI, with this segment accounting for 63.7% of total revenue [6] - The company is expected to release version 6.0 of its generative AI model in Q2 2025, focusing on enhancing multimodal understanding and interaction capabilities [6] Financial Data and Profit Forecast - The company forecasts revenues of 5.138 billion CNY for 2025, with a year-on-year growth rate of 36% [5] - The adjusted net loss for 2025 is projected to be 3.540 billion CNY, with a gradual reduction in losses expected in subsequent years [5][7] - The company maintains a high R&D investment, with R&D expenses accounting for 109.5% of sales in 2024 [6]

在去年高增长、高基数的基础上，2024年生成式AI业务依然保持了极高的增速。这不仅彰显了以大模 型技术为基础的业务所具有的强劲生命力，也验证了商汤"AI基础设施(大装置)-大模型(日日新)-应用"三 位一体战略的可行性。该战略构建了完整的商业闭环：AI大装置为应用落地提供算力支撑，应用反馈 又指导基础设施升级；大模型能力与基础设施相互促进，实现算力效率的持续优化；同时，多元化的技 术路径加速了应用场景渗透，而应用实践又推动大模型向精细化方向发展。 特别值得关注的是，2024年生成式AI业务在总收入中的占比从34.8%提升至63.7%，是商汤集团最大的 业务收入。这标志着公司彻底完成从视觉AI向生成式AI战略转型的全面胜利，其收入结构变革已形成 不可逆的良性发展态势。 一、"三位一体"战略持续巩固技术领先地位，训练效率优于DeepSeek公开报告 商汤科技依托"大装置-大模型-应用"的深度协同创新，形成核心技术差异化优势，并持续巩固技术领先 地位。 2024年，商汤生成式AI业务收入突破24亿元，同比增长103.1%，连续两年保持三位数增长。这一强劲 表现推动商汤2024年全年收入同比增长10.8%，达到37. ...

Investment Rating - The report maintains a "Buy" rating for the company [6]. Core Insights - The company achieved a significant revenue growth of 10.8% in 2024, with total revenue reaching 3.77 billion RMB, compared to a decline of 10.6% in 2023 [2]. - The revenue from generative AI surged by 103% to 2.4 billion RMB, increasing its share of total revenue from 35% in 2023 to 64% in 2024 [2]. - Adjusted net loss narrowed by 33.7% to 4.31 billion RMB, while EBITDA loss decreased by 42.3% to 3.14 billion RMB [2]. - The company has optimized its cash flow management, reducing the cash flow cycle from 334 days to 228 days, despite a net cash outflow of 3.93 billion RMB from operating activities [3]. Summary by Sections Financial Performance - In 2024, the company reported a total revenue of 3,772.1 million RMB, with a growth rate of 10.8% [7]. - The adjusted net loss for 2024 was 4,278.4 million RMB, showing improvement from the previous year [11]. - The gross margin slightly decreased to 42.9% from 44.1% in 2023 [2]. Business Highlights - The company is focusing on a synergistic development model of "large devices + large models + AI applications," enhancing resource allocation [4]. - The AI applications are gaining traction, with a sixfold increase in customer willingness to pay in 2024 compared to 2023 [5]. - The company is positioned third in China's GenAI IaaS market with a 13.3% market share [4]. Future Projections - Revenue forecasts for 2025 and 2026 have been adjusted downwards by 15% and 17% to 4.88 billion RMB and 6.24 billion RMB, respectively [6]. - The net profit projections for 2025 and 2026 are revised to -2.73 billion RMB and -0.66 billion RMB, with a new forecast for 2027 showing a potential profit of 140 million RMB [6].

Investment Rating - The report maintains a "Buy" rating for the company, expecting a price increase of over 15% in the next 6-12 months [3][10]. Core Insights - The company reported a revenue of 3.77 billion yuan for 2024, a year-on-year increase of 10.8%, with generative AI revenue reaching 2.4 billion yuan, growing by 103.1% [1]. - Adjusted net loss narrowed to 4.25 billion yuan from 5.41 billion yuan in the previous year [1]. - Revenue projections for 2025-2027 are estimated at 5.069 billion yuan, 6.920 billion yuan, and 9.538 billion yuan, reflecting growth rates of 34.38%, 36.51%, and 37.84% respectively [3]. Financial Performance - The company’s operating income for 2024 is projected to be 3.772 billion yuan, with a growth rate of 10.75% [6]. - The adjusted net profit for 2025 is expected to be -3.783 billion yuan, improving from -4.278 billion yuan in 2024, indicating a growth of 11.57% [3][6]. - The research indicates a decrease in sales, management, and R&D expense ratios, with R&D expense ratio at 109.5%, up by 7.8 percentage points year-on-year [2]. Business Strategy - The company is advancing a "large device - large model - application" strategy, focusing on AI visual and intelligent driving businesses, which saw revenue declines of 39.5% and 33.2% respectively [2]. - The AI visual customer repurchase rate increased by 31 percentage points, and the company is expanding into Southeast Asia and the Middle East [2]. - The company delivered 1.67 million units of its product, a 9.2% increase year-on-year, and plans to mass-produce a collaboration with Dongfeng Intelligent Driving in 2025 [2]. Valuation Metrics - The projected price-to-sales (P/S) ratios for 2025, 2026, and 2027 are 11.61, 8.50, and 6.17 respectively [3][6]. - The diluted earnings per share (EPS) is expected to improve from -0.12 yuan in 2024 to -0.06 yuan in 2027 [6].

Group 1 - The core viewpoint of the report is that the company maintains an "outperform" rating for SenseTime-W (00020) and is optimistic about the mid-term demand for AIGC, switching to a 10.8x P/S for 2026 [1] - The company expects a revenue growth of 10.8% year-on-year in 2024, reaching 3.772 billion yuan, with generative AI and visual AI revenues increasing by 103.1% and 39.5% respectively [2] - The company has achieved a computing power scale of 23,000P by the end of March 2025, a 92% year-on-year increase, and plans to adopt a dual-driven model of self-built and light asset operations in the future [2] Group 2 - The company has initiated a "1+X" new structure in 2024, focusing on its core business while encouraging independent operations and financing for ecosystem enterprises [3] - The accounts receivable turnover days have improved from 399 days in 2024 to 231 days, with cash inflow increasing by 19% year-on-year to 4.623 billion yuan [3] - The company has a total cash reserve of 12.75 billion yuan at the end of 2024, indicating sufficient liquidity [3]

Financial Performance - Total revenue for the year ended December 31, 2024, reached RMB 3,772.1 million, representing a year-on-year increase of 10.8%[5] - The annual loss narrowed to RMB 4,306.6 million, a reduction of 33.7% compared to the previous year[5] - Adjusted EBITDA for the year was RMB (3,089.2) million, showing a year-on-year improvement of 29.3%[5] - Revenue increased by 10.8% from RMB 3,405.8 million in 2023 to RMB 3,772.1 million in 2024, driven primarily by the expansion of generative AI[33] - The annual loss for 2024 was RMB 4,306.6 million, a reduction from RMB 6,494.7 million in 2023, reflecting a focus on improving operational efficiency[46] - The company reported a net loss of RMB 4,278,383 thousand for the year 2024, compared to a loss of RMB 6,440,159 thousand in 2023, indicating an improvement of approximately 33%[80] Revenue Breakdown - The revenue from generative AI exceeded RMB 2,404.0 million, with a remarkable growth rate of 103.1%, accounting for 63.7% of total revenue[9] - Smart vehicle revenue decreased by 33.2% from RMB 383.7 million in 2023 to RMB 256.2 million in 2024, due to a strategic shift towards smart cockpit and autonomous driving[34] - Visual AI revenue accounted for 29.5% of total group revenue, with a 31 percentage point increase in customer repurchase rate in 2024 compared to 2023[20] - Visual AI revenue fell by 39.5% from RMB 1,838.4 million in 2023 to RMB 1,111.9 million in 2024, as the company focused on high-quality customers[34] User Engagement and Product Development - Monthly user engagement for productivity and interaction tools increased eightfold compared to 2023[8] - The company launched the first domestic large model surpassing GPT 4-Turbo performance in April 2024[8] - The monthly invocation of a leading humanoid companion app increased by 833% over the year, with a next-day retention rate of 50% and an average daily usage time of 130 minutes[18] - The number of users for the office assistant and code assistant products surpassed 1.5 million, with daily invocation exceeding one million and processing over 3.5 billion tokens daily[19] - The company plans to release the next version of its multi-modal model, expected to significantly enhance multi-modal understanding and reasoning capabilities, targeting international standards[17] Operational Efficiency - Management and sales expenses decreased by 9.1% year-on-year, reflecting improved operational efficiency[9] - The company achieved a 90% reduction in virtual cluster management costs and increased training efficiency for large models through automated multi-dimensional parallel strategies[14] - The efficiency of heterogeneous training reached 95% of homogeneous training, with resource utilization for large-scale domestic chip clusters improved to 80%[15] - Operating loss improved from RMB 6,646.6 million in 2023 to RMB 4,529.3 million in 2024, indicating a reduction in losses[32] Research and Development - R&D expenses grew from RMB 3,465.8 million in 2023 to RMB 4,131.9 million in 2024, reflecting the company's commitment to innovation[32] - R&D expenses increased by 19.2% from RMB 3,465.8 million in 2023 to RMB 4,131.9 million in 2024, primarily due to investments in training and fine-tuning base models and developing generative AI applications[37] Cash Flow and Financing - The company achieved a cash reserve of RMB 12,752.2 million as of December 31, 2024[9] - Cash and cash equivalents as of December 31, 2024, were RMB 8,888.0 million, down from RMB 9,423.5 million as of December 31, 2023[53] - The net cash used in operating activities was RMB 3,926.7 million, an increase attributed to higher prepayments for expanding generative AI business capabilities[56] - The net cash generated from financing activities was RMB 6,259.9 million, mainly from the issuance of ordinary shares and net borrowings[58] Market Position and Partnerships - The company maintained a leading position in the AI infrastructure sector, ranking first in comprehensive capability assessments in the GenAI technology stack market[12] - The company holds a 13.8% market share in the Chinese large model application market, ranking third behind Baidu and Alibaba Cloud[17] - The company has established partnerships with major telecom operators and asset management platforms, optimizing capital efficiency and responding flexibly to market changes[13] Infrastructure and Energy Efficiency - The Shanghai Lingang AIDC was recognized as the first 5A-level intelligent computing center in China, marking a significant achievement in infrastructure[12] - The PUE of the Lingang Intelligent Computing Center was optimized to 1.28 in 2024, achieving a significant reduction in energy consumption[27] Corporate Governance and Compliance - The company has complied with corporate governance codes, except for the separation of the roles of Chairman and CEO, which are currently held by the same individual[160] - The group did not declare or pay any dividends for the year ended December 31, 2024, consistent with 2023[121] Future Outlook - The company plans to focus on core AI cloud business and expand its ecosystem of AI applications in various sectors, including smart vehicles and healthcare[30] - The company expects that 26% of the transaction price allocated to unsatisfied performance obligations will be recognized as revenue in the next year, down from 56% in 2023[103]

数据统计显示，中证沪港深张江自主创新50指数近一个月下跌6.62%，近三个月上涨7.11%，年至今上 涨10.05%。 据了解，中证沪港深张江自主创新50指数从沪港深三地市场中选取50只业务涉及生物医药、集成电路、 人工智能、高端装备制造、节能环保、新材料等产业的张江高新区上市公司证券作为指数样本，以反映 沪港深三地市场张江高新区新兴产业上市公司证券的整体表现。该指数以2014年12月31日为基日，以 1000.0点为基点。 中证沪港深张江自主创新50指数上涨0.03%，前十 大权重包含中芯国际等 从中证沪港深张江自主创新50指数持仓样本的行业来看，信息技术占比67.98%、医药卫生占比 20.15%、工业占比6.75%、通信服务占比2.12%、原材料占比1.69%、房地产占比0.89%、可选消费占比 0.42%。 金融界3月24日消息，上证指数上涨0.15%，中证沪港深张江自主创新50指数 (SHS张江50，931596)上涨 0.03%，报1485.13点，成交额293.43亿元。 从指数持仓来看，中证沪港深张江自主创新50指数十大权重分别为：韦尔股份（11.04%）、中芯国际 （9.85%）、中微公司（ ...

Core Viewpoint - SenseTime's R-UniAD end-to-end autonomous driving solution is set to be unveiled at the Shanghai Auto Show in April, with real vehicle deployment completed and expected delivery by the end of the year [1][3]. Group 1: Company Developments - Wang Junping, former senior director of intelligent driving at Nezha Auto, joined SenseTime's autonomous driving division in February 2023, previously being part of Baidu's intelligent driving team [2]. - SenseTime has been collaborating with Nezha Auto since September 2021, focusing on intelligent driving and smart cockpit technologies [2]. - Wang Weibao, who took over from Shijianping as the head of intelligent driving, joined SenseTime at the end of 2023 and has a background in Apple's autonomous driving team and as CTO at New Stone Unmanned Vehicle Company [3]. Group 2: Industry Context - The autonomous driving sector is experiencing intensified competition, particularly for companies not in the top tier, highlighting the challenges faced by solution providers [3]. - SenseTime collaborates with over 30 automotive companies, including GAC, BYD, Honda, and NIO, with solutions already deployed in models like the Haobo and Nezha's super sedan [3].