Astera Labs 25Q3 业绩表现：2025 年第三季度业绩表现全面超越市场预期，实现营收 2.306 亿美元，同比大幅增长 104%，环比亦提升 20%；盈利能力方面尤为突出，非 GAAP 每股收益（EPS）达 0.49 美元，显著高于市场预期，超出幅度达 276.92%，非 GAAP 毛利率进一步提升至 76.4%的高位，非 GAAP 经营利润率也达到 41.7%，反映其产品结 构优化与经营杠杆持续释放。期末公司现金及等价物与有价证券余额为 11.3 亿美元，为未来研发与战略布局提供充裕资金支持。尽管业绩强劲，盘后股价仍出现 5.91%的回调，或与市场对后续毛利率走势及竞争格局的短期担忧有关。对于 2025 年第四季度，公司给出营收 2.45–2.53 亿美元的指引，显示其增长势头有望延续。公司正积极扩展产品线与团队规模，计划在 2025 年底前实现员工总数增长约 60%，以支持未来在多互联标准与光技术领域的长期 布局。

我国半导体市场规模持续增长，但受制于行业周期波动、需求分化及结构性复苏程度不均衡等因素，增速或许放缓，2025 年第三季度，供给侧的国产替代成效正在加速显现，需求侧的 增长动力由传统消费电子逐步转化为算力与新能源。近期政策利好频出，“十五五”规划建议 的主要目标之一为科技自立自强水平大幅提高，行业政策继续强化“国货国用”，国内半导体产业链的持续完善使得风险抵御能力进一步加强。展望 2025 年第四季度，半导体行业仍位于信用扩张初期阶段，信用水平将持续加强，并呈现早期结构性分化特征。

The global AI in appliances and electronics market encompasses the integration of artificial intelligence technologies into consumer electronics and household appliances. This technology enables devices to learn from user behavior, automate tasks, and optimize performance for enhanced efficiency and user experience.

The global AI in semiconductor devices market encompasses the design, fabrication, and integration of specialized hardware engineered to execute artificial intelligence workloads efficiently. This includes a spectrum of silicon solutions, such as Graphics Processing Units (GPUs), Field Programmable Gate Arrays (FPGAs), and Application Specific Integrated Circuits (ASICs), including Neural Processing Units (NPUs). These devices are architected to accelerate machine learning and deep learning algorithms, enabling rapid data processing directly on the hardware.

The global AI server hardware market encompasses the design, manufacturing, and distribution of specialized computing systems engineered to execute artificial intelligence workloads. These servers are fundamentally different from general purpose servers, incorporating high performance components such as Graphics Processing Units (GPUs), Tensor Processing Units (TPUs), Field Programmable Gate Arrays (FPGAs), and other application specific integrated circuits (ASICs) known as AI accelerators.

AI in electronics and sensors encompasses the design, development, and integration of artificial intelligence capabilities directly within electronic components and sensor systems. This involves embedding machine learning algorithms into hardware like microcontrollers and system-on-chips, enabling on-device data processing, pattern recognition, and real time decision making a concept known as edge AI.

The market by value in this report is defined as the revenues that enterprises gain from goods and/or services sold within the specified market and geography. • Values in this market are ‘factory gate’ values, that is the value of goods sold by the manufacturers or creators of the goods, whether to other entities (including downstream manufacturers, wholesalers, distributors and retailers) or directly to end customers. The value of goods in this market includes related services sold by the creators of the goods.

核聚变技术产业化进程加速，多技术路线并行。随着全球能源转型迫切性提升，可控核聚变作为终极清洁能源解决方案，正从实验室研究迈向工程化与商业化临界点。当前，磁约束托卡马克路线主导产业化进程，惯性约束与混 合路线并行发展，各国通过政策协同与资本投入加速布局，科技巨头也在积极布局聚变能源赛道。中国以 EAST、BEST 等大科学装置为依托，在长脉冲等离子体控制、高温超导磁体等关键领域实现突破，供应链国产化率持续提升，预计 2027 年 BEST 装置将率先实现聚变能发电演示（Q>1）。近期，BEST 装置关键部件完成落位安装，标志着总投资近 200 亿元的主体建设全 面启动，这是我国核聚变商业化的重要里程碑。