NVIDIA Mellanox® MCX345A-BCQN — ConnectX Network Adapter

Specifications

OEM: NVIDIA Mellanox
Part Number (MPN): MCX345A-BCQN
Category: ConnectX Network Adapter
ASIC / Family: ConnectX-3
Port Configuration: Single-Port
Connector Type: QSFP
Speed: 40GbE
Protocol Support: Ethernet
Condition: New, factory-sealed


Identity
The NVIDIA Mellanox® MCX345A-BCQN is a ConnectX-3 single-port QSFP 40GbE ConnectX Network Adapter. New, factory-sealed by T.E.S IT-SOLUTIONS.

What This Is NOT

NOT a switch. Host-side PCIe adapter card — requires a server PCIe slot.
NOT a different ConnectX generation. This is ConnectX-3. Verify driver and firmware match this exact generation.
NOT a VPI dual-protocol card unless explicitly stated. Check protocol mode in specifications above.
NOT bundled with cables, transceivers, or low-profile bracket.


Capabilities & Features

RDMA over Converged Ethernet (RoCE v1/v2) and native InfiniBand verbs.
NVMe over Fabrics (NVMe-oF) support via RoCE v2 or InfiniBand transport.
ASAP² (Accelerated Switching and Packet Processing) for hardware-accelerated Open vSwitch.
SR-IOV virtualization for VM and container deployments.
Encapsulation/decapsulation acceleration: VXLAN, NVGRE, Geneve, GTP.
T10-DIF data-integrity offload for storage workloads.


Compatibility & Constraints

Drivers: NVIDIA MLNX_OFED (Linux), in-box mlx5_core (upstream Linux), WinOF-2 (Windows), nmlx5 (VMware ESXi).
Cabling: QSFP DAC / AOC / transceivers matching 40GbE.
Compatible switches: NVIDIA Spectrum / SwitchX-2.


Deployment & Use Cases

HPC clusters with RDMA / RoCE communication.
AI / ML training fabrics with GPUDirect-capable adapters.
NVMe-oF storage targets and initiators.
Virtualized data centers with SR-IOV.
SDN / NFV with ASAP2 hardware OVS acceleration.


Why T.E.S IT-SOLUTIONS
Every unit is bench-tested by T.E.S IT-SOLUTIONS: link establishment, port-by-port traffic validation, and full cosmetic inspection. Professional anti-static packaging, global B2B shipping from EU stock, and technical consultation for HPC / AI cluster deployments.