1.6T High-Speed Interconnect
Up to 1.6Tbps per port to meet next-generation AI cluster demands.
Up to 1.6Tbps per port to meet next-generation AI cluster demands.
Optimized for NVIDIA AI platforms including B300 and GB300.
High-bandwidth and ultra-low-latency connectivity for efficient GPU communication.
Rigorous compatibility and performance testing with leading switches and NICs.
Designed for NVIDIA B300 and GB300 AI infrastructure, this solution delivers high-performance, reliable interconnects across compute, storage, and management networks.
AICPLIGHT provides 1.6T / 800G optical transceivers, DAC / AOC cables, fiber connectivity products, and related accessories to help customers build scalable, efficient, and future-ready AI clusters.
Massive data throughput for demanding AI workloads.
Stable operation in large-scale cluster deployments.
Flexible expansion for future network upgrades.
delivers an end-to-end solution service process-from planning and design to validation, deployment,and operations-helping customers efficiently build stable and scalable network infrastructure.
The expert team conducts thoroughresearch to deeply understand yourbusiness needs.
The expert team conducts thoroughresearch to deeply understand yourbusiness needs.
The expert team conducts thoroughresearch to deeply understand yourbusiness needs.
The expert team conducts thoroughresearch to deeply understand yourbusiness needs.
With Al computing power demands growingexponentially, next-generation supercomputingarchitectureslike the NvDIA B300 have emerged has launched a 1.6T transceiver solution specifically designed for the NvIDlA B300,which can significantly improve the computingefficiency of Alclusters and reduce TCO, providing high-reliability and scalable interconnection support for Al data centers.
InfiniBand-ready topology with scalable interconnect planning.With Al computing power demands growingexponentially, next-generation supercomputingarchitectureslike the NvDIA B300 have emerged has launched a 1.6T transceiver solution specifically designed for the NvIDlA B300,which can significantly improve the computingefficiency of Alclusters and reduce TCO, providing high-reliability and scalable interconnection support for Al data centers.
With Al computing power demands growingexponentially, next-generation supercomputingarchitectureslike the NvDIA B300 have emerged has launched a 1.6T transceiver solution specifically designed for the NvIDlA B300,which can significantly improve the computingefficiency of Alclusters and reduce TCO, providing high-reliability and scalable interconnection support for Al data centers.
With Al computing power demands growingexponentially, next-generation supercomputingarchitectureslike the NvDIA B300 have emerged has launched a 1.6T transceiver solution specifically designed for the NvIDlA B300,which can significantly improve the computingefficiency of Alclusters and reduce TCO, providing high-reliability and scalable interconnection support for Al data centers.
Ensure your network performs flawlessly before deployment.
Tested with NVIDIA GPUs, IB switches, and NICs for seamless integration.
IR, LDD, eye diagram, temperature, and other key optical parameters.
End-to-end traffic and BER testing for link stability and reliability.
Custom labeling, packaging, 100% testing, and phased delivery support.
This solution includes 1.6T / 800G / 400G optical transceivers, DAC / AOC cables, MPO fiber patch cords, fiber connectivity products, and related accessories. Switches, NICs, and other hardware can also be matched according to project requirements.
Yes. This solution is designed for next-generation AI infrastructure such as B300 and GB300 platforms, supporting GPU interconnects, storage access, and management networks.
Share your requirements and our experts willrecommend the best InfiniBand solution for you.