Cisco and NVIDIA join forces to standardize 400G-per-lane pluggable optics, targeting increased bandwidth density and improved scalability for next-generation AI factory infrastructure.
NVIDIA is partnering with Cisco to advance the physical layer of AI infrastructure through the development of high-density 400G-per-lane pluggable optics. This collaboration is essential for NVIDIA to maintain its leadership in AI hardware by ensuring its GPUs can be interconnected with the necessary bandwidth density.
The collaboration aims to develop an open industry specification for 400G-per-lane pluggable optics to increase bandwidth density for AI factories.
Cisco is positioning itself as a critical infrastructure provider for AI-driven data centers by focusing on resilient, open Ethernet foundations. By collaborating on 400G-per-lane pluggable optics, the company aims to address the scalability requirements of massive, distributed AI factories.
Cisco is focused on delivering resilient, open Ethernet foundations to support the scaling of massive, distributed AI factories.
Cisco Systems and NVIDIA are collaborating to develop an open industry specification for 400G-per-lane pluggable optics. This initiative aims to enhance bandwidth density and scalability for massive, distributed AI factories, reinforcing the Ethernet foundation required for high-performance computing environments.
The companies aim to develop an open industry specification for 400G-per-lane pluggable optics to improve bandwidth density and scalability for AI factory networks.
As AI factories grow in scale, they require more resilient and open Ethernet foundations to manage massive data throughput, which this new optical design seeks to provide.