Shenzhen, China – Huawei Technologies Co. announced new SuperPoD Interconnect technology at its Huawei Connect conference on Thursday, designed to enhance compute power for artificial intelligence (AI) systems. The unveiling comes as China has reportedly restricted domestic tech companies from purchasing advanced AI hardware from rival chipmaker Nvidia Corp.
The new SuperPoD Interconnect infrastructure is engineered to link up to 15,000 graphics cards, including Huawei’s own Ascend AI chips, to collectively increase computational capacity. This technology aims to provide the necessary compute power for training and scaling complex AI systems, which are increasingly vital across various industrial sectors from manufacturing automation to supply chain optimization.
Industry observers note that the SuperPoD Interconnect technology appears to be positioned as a direct competitor to Nvidia’s NVLink infrastructure, which facilitates high-speed communication between AI chips to boost performance. While individual Huawei AI chips are generally considered less powerful than those offered by Nvidia, the SuperPoD system seeks to overcome this disparity by enabling extensive clustering of processors.
The announcement by Huawei follows closely on recent developments in the technology trade landscape. Just one day prior to Huawei's keynote, reports indicated that China had prohibited domestic technology firms from acquiring certain Nvidia hardware, specifically mentioning Nvidia’s RTX Pro 600D servers, which were designed for the Chinese market. This regulatory measure underscores the increasing focus on domestic AI chip development and infrastructure within China.
Huawei’s strategic introduction of this new AI infrastructure highlights ongoing shifts in the global AI hardware market, particularly in light of trade restrictions affecting the availability of advanced semiconductor technology. The company has not disclosed specific deployment timelines or initial industrial partners for the SuperPoD Interconnect system, though its stated purpose is to bolster AI capabilities for large-scale applications.
