[SMM Tin News Flash: MIIT: Publicly Soliciting Comments on 2 Mandatory National Standards Including "Safety Requirements for Combined Driving Assistance Systems of Intelligent Connected Vehicles"]

On April 15, China Mobile took the lead in releasing the Single-Lane 400G Ethernet Physical Layer Report at the 2026 Cloud-Network Intelligent Connectivity Conference, systematically proposing a technical framework for single-lane 400G Ethernet physical layer. Today, as the parameter scale of AI large models continues to increase, the demand for high-speed interconnection among GPUs in intelligent computing centers is growing rapidly, urgently requiring greater interconnection bandwidth. GPU interconnection bandwidth has reached 1.6 Tbps. To achieve 3.2 Tbps intelligent computing center interconnection bandwidth, there are two paths: "increasing single-lane rate" and "increasing the number of lanes." With the existing 200 Gbps single-lane approach, 16-lane aggregation is required, imposing significant pressure on ports, heat dissipation, and cabling, making deployment difficult. With single-lane 400 Gbps, only 8-lane aggregation is needed, making it a superior and more feasible solution. Single-lane 400 Gbps is not simply a doubling of the rate, but rather a systematic innovation that breaks through the limits of physical transmission, facing multiple challenges in theory, materials, integration, and industry collaboration.

MeiG Smart announced that the company signed the "MeiG Smart AI R&D and Advanced Manufacturing Industry Project Cooperation Agreement" with the Management Committee of Nantong Chibei High-tech Industrial Development Zone, Jiangsu Province on April 15. The company or its wholly-owned subsidiary plans to purchase approximately 46 mu of industrial land in Nantong Chibei High-tech Industrial Development Zone to establish an R&D center and an advanced process pilot testing and manufacturing base, and to invest in the construction of the MeiG Smart AI R&D and Advanced Manufacturing Industry Project, with an estimated total investment of 300 million yuan.

Alibaba Cloud announced that, in order to continuously ensure a stable supply of underlying hardware, improve platform operation and maintenance service quality, and respond to changes in computing power market costs, it would moderately adjust the service prices of certain MU (Model Unit) model units. The affected product was the Model Studio large model service platform, with price raises ranging from 2% to 7%.