NVIDIA 920-9B110-00FH-0D0 | 40x QSFP56 200Gb HDR IB | 2x PSU | P2C airflow | Rail kit | Unmanaged AI and in-network computing acceleration

and in-network computing acceleration

Each port uses a 12‑fiber MPO/APC interface with 4 transmit and 4 receive fibers

and high-performance computing environments requiring cost-effective

Integrated optics and fiber simplify deployment and reduce power compared to discrete transceivers

With 64 QSFP28 ports delivering 100GbE connectivity

NVIDIA 920-9B110-00FH-0D0 | 40x QSFP56 200Gb HDR IB | 2x PSU | P2C airflow | Rail kit | Unmanaged AI and in-network computing accelerationHighdensity 1U 200Gb s HDR InfiniBand switch for AI HPC leaf or spine deployments. Delivers linerate throughput with ultralow latency and advanced congestion control to maximize application performance. This configuration uses portside exhaust (P2C) airflow, includes dual hotswap PSUs and rack rails, and is the unmanaged variant requiring an external Subnet Manager (e. g., UFM or OpenSM).