华为徐直军谈芯片三年规划，努力打造“超节点+集群”解决方案

Core Viewpoint - Huawei expresses confidence in providing sustainable and sufficient computing power for the long-term rapid development of artificial intelligence, leveraging its supernodes and clusters with the world's strongest computing capabilities [1][6]. Group 1: Chip Development - Huawei has announced the development progress of several self-developed chips, including the Ascend 950PR, 950DT, and Ascend 960 and 970, with the 950PR set to launch in Q1 2026 [1]. - The company is focusing on creating a "supernode + cluster" computing solution to meet the continuously growing demand for computing power, based on China's available chip manufacturing processes [3]. Group 2: Supernode Infrastructure - The supernode has become a new norm in AI infrastructure construction, with over 300 Cloud Matrix 384 supernodes deployed [3]. - Huawei has released the Atlas 950 SuperPoD and Atlas 960 SuperPoD supernodes, supporting 8192 and 15488 Ascend cards respectively, showcasing global leadership in key metrics such as card scale, total computing power, memory capacity, and interconnect bandwidth [5]. Group 3: Technological Innovations - Huawei has introduced the world's first general-purpose computing supernode, TaiShan 950 SuperPoD, which, combined with the GaussDB distributed database, aims to replace various large and small machines as well as Exadata database integrated machines [6]. - The company has overcome significant challenges in large-scale supernode interconnect technology, launching the UnifiedBus interconnect protocol, with plans to open the Lingqu 2.0 technology specifications in the future [6].