全球领先的新经济产业第三方数据挖掘与分析机构
关于“寒武纪-U”的报告
-
艾媒咨询 | 2025年中国算力概念产业发展状况及重点企业大数据监测报告
2025年中国算力概念产业已进入“头部引领、两极分化”的成熟发展阶段，资源、技术、资金正加速向产业链核心企业集聚，行业集中度将持续提升并趋于稳定。全球领先的新经济产业第三方数据挖掘和分析机构iiMedia Research（艾媒咨询）最新发布的《2025年中国算力概念产业发展状况及重点企业大数据监测报告》数据显示，2025年中国智能算力规模达1590 EFLOPS，核心产业规模预计突破1.2万亿元，较2024年725.3 EFLOPS的规模实现大幅跃升，行业进入规模化、高质量发展新阶段。在此背景下，算力概念产业经营表现分化加剧，头部效应显著：2025年前三季度，营收排名前三的企业合计贡献TOP10总营收的72.90%，研发投入集中度达64.69%，凸显产业链关键环节的高度集聚特征。艾媒咨询分析师认为，算力概念产业头部企业凭借技术、规模及资金优势，在市值、营收、研发等核心指标上占据主导地位。2025年算力概念产业发展确定性较强，但同时也面临产业内企业表现分化、技术研发投入压力较大等挑战，未来产业竞争将聚焦于核心技术突破与高效服务能力提升，具备技术壁垒与场景优势的企业将实现持续突围，推动中国算力产业逐步跻身全球领先行列。
In 2025, Chinese computing power concept industry has entered a mature development stage characterized by "leading by head enterprises and polarization". Resources, technologies, and capital are accelerating their agglomeration to core enterprises in the industrial chain, and the industry concentration will continue to increase and tend to stabilize. According to the latest report "Report on the Development Status of China's Computing Power Concept Industry and Big Data Monitoring of Key Enterprises in 2025" released by iiMedia Research, a world-leading third-party data mining and analysis institution for the new economy industry, China's intelligent computing power scale reached 1590 EFLOPS in 2025, and the core industry scale is expected to exceed 1.2 trillion yuan, achieving a substantial jump compared with the scale of 725.3 EFLOPS in 2024, indicating that the industry has entered a new stage of large-scale and high-quality development.Against this background, the operating performance of the computing power concept industry has become increasingly differentiated, with a prominent head effect: in the first three quarters of 2025, the top three enterprises in terms of revenue contributed 72.90% of the total revenue of the top 10 enterprises, and the concentration of R&D investment reached 64.69%, highlighting the high agglomeration characteristics of key links in the industrial chain. Analysts from iiMedia Research pointed out that, relying on their advantages in technology, scale and capital, the head enterprises in the computing power concept industry occupy a dominant position in core indicators such as market value, revenue and R&D.The computing power concept industry is expected to maintain strong development certainty in 2025, but it also faces challenges such as differentiated performance of enterprises in the industry and great pressure of technological R&D investment. In the future, industrial competition will focus on the breakthrough of core technologies and the improvement of efficient service capabilities. Enterprises with technological barriers and scenario advantages will achieve continuous breakthroughs, promoting China's computing power industry to gradually rank among the world's leading levels. -
艾媒咨询|2020上半年中国人工智能芯片行业研究报告
本报告研究涉及企业/品牌/案例：寒武纪、知存科技。
iiMedia Research(艾媒咨询)数据显示，2019年中国AI芯片市场规模约为115.5亿元，在5G商用的普及和政策、技术等各因素的推动下，AI芯片有望在云计算、安防、消费电子、机器人等领域实现大规模商用，预计2021年AI芯片市场规模将达到436.8亿元。但值得注意的是，随着人工智能技术的加速普及，下游应用领域对AI算力和能耗的要求越来越高，传统冯诺依曼架构式芯片的瓶颈逐渐显露，AI芯片将朝着存算一体化方向发展。艾媒咨询分析师认为，存算一体AI芯片的发展前景虽受到广泛认可，但整体仍处在发展的起步阶段。从实现计算与存储的融合设计，到技术的落地、量产、规模化商用，还有较长阶段。能够率先实现技术、产品突破的企业将更容易获得资本、人才、市场的支持。
According to data from iiMedia Research, the scale of China’s AI chip market in 2019 is about 11.55 billion yuan. Driven by the popularization of 5G commercial use, policies, and technology, AI chips are expected to be used in cloud computing, security, and consumer electronics. Robotics and other fields achieve large-scale commercial use, and the AI chip market is expected to reach 43.68 billion yuan in 2021. However, with the accelerated popularization of artificial intelligence technology, downstream application fields have higher and higher requirements for AI computing power and energy consumption. The bottleneck of traditional Von Neumann architecture chips has gradually emerged, and AI chip will develop towards the integration of storage and computing. iiMedia Consulting analysts believe that although the development prospects of the storage-computing integrated AI chip are widely recognized, the overall development is still in the initial stage. From the realization of the integrated design of computing and storage, to the implementation of technology, mass production, and large-scale commercial use, there is still a long stage. Companies that can take the lead in achieving technological and product breakthroughs will be more easily supported by capital, talent, and the market. -
-
-
-
-
-
-
- 1
- 2
- 3
- 4
- 5
