With the continuous increase in bandwidth demand from applications such as AI and machine learning, network architectures must constantly break through power consumption bottlenecks and reduce system construction costs. The development needs of the internal network in the artificial intelligence data center are particularly urgent, where underlying optical technologies such as optical interconnects and optical switching are becoming key research areas, expected to provide faster and more reliable network connections for artificial intelligence data centers. As the high-speed technology of artificial intelligence data center continues to evolve, the requirements for device performance and power consumption are becoming increasingly stringent. This has prompted the continuous upgrading of technologies such as SerDes, silicon photonics, DSP, CPU/GPU/DPU, CPO/NPO/LPO, and others, to meet the needs of artificial intelligence data centers. The fierce competition in the GPU market last year further drove technological iteration, bringing significant improvements in storage capacity, bandwidth, and optical module processing capabilities. Meanwhile, as artificial intelligence data centers continue to be built and developed, the complexity of algorithms and models is also increasing, leading to higher requirements for intelligent management and control. One of the industry's current focuses is how to promote the evolution of intelligent computing center networks towards self-intelligent networks, advancing towards a higher level of automation and intelligent operation and maintenance.

Based on this backdrop, CIOE invites internet service providers, datacenter equipment manufacturers, and core module and component vendors to jointly host “Evolution Trends of Optical Technologies for Artificial Intelligence Data Center Forum”, focusing on discussing the future trends of high-speed transmission network interconnection in these centers.