NVIDIA MCX653106A-ECAT ConnectX-6 EN Adapter

NVIDIA MCX653106A-ECAT ConnectX-6 EN Adapter

Brand: NVIDIA | Category: Network Adapters & NICs

SKU: NVIDIAMCX653106AECATCONNECTX6ENADA | Part #: MCX653106A-ECAT | MPN: MCX653106A-ECAT

Contact for Pricing — Request a Quote

Request a Quote Contact Us

About the NVIDIA MCX653106A-ECAT ConnectX-6 EN Adapter

For low-latency workloads, NIC offload capabilities free critical CPU cycles—and the NVIDIA MCX653106A-ECAT ConnectX-6 EN Adapter delivers exactly that. This single-port 100GbE network adapter combines extreme throughput with hardware-accelerated offloads including TSO, LRO, and IP/TCP/UDP checksum offloading, enabling data center infrastructure teams to reduce latency-sensitive bottlenecks. Built on PCIe 4.0 x16 connectivity, the MCX653106A-ECAT supports up to 127 virtual functions per port via SR-IOV, making it ideal for virtualized and containerized environments requiring fine-grained network isolation.

The ConnectX-6 EN series supports RoCEv2 (RDMA over Converged Ethernet v2) and GPUDirect RDMA, allowing direct GPU-to-GPU data transfer without CPU intervention—essential for AI and machine learning clusters. Advanced flow control via IEEE 802.1Qbb PFC and IEEE 802.1Qaz ETS ensures deterministic performance, while tunneling offload for VXLAN, NVGRE, and MPLS keeps overlay networks performant. With RoHS compliance and an operating temperature range of 0°C to 55°C, deployment flexibility spans multiple data center environments. NVIDIA MLNX_OFED and OpenFabrics Alliance driver support covers Linux (RHEL, Ubuntu, SLES) and Windows Server, with typical power consumption around 18W. Network architects and infrastructure teams seeking enterprise-grade 100GbE performance should evaluate this solution.

Deployment Scenarios

  • High-performance computing clusters requiring sub-microsecond latency and GPU acceleration
  • AI training and inference infrastructure leveraging GPUDirect RDMA for distributed model parallelism
  • Multi-tenant cloud platforms benefiting from SR-IOV virtual function segmentation
  • Storage array backends and NVMe-over-Fabrics (NVMe-oF) networks using advanced RDMA protocols
  • Virtualized data center networks with overlay technologies (VXLAN/NVGRE) requiring stateless offload

For detailed specifications, compatibility validation, and deployment guidance on the MCX653106A-ECAT, contact Omnixon Global via RFQ.

Technical Specifications

BrandNVIDIA
CategoryNetwork Adapters & NICs
SKUNVIDIAMCX653106AECATCONNECTX6ENADA
Part NumberMCX653106A-ECAT
ConditionNew
Capacity200GB
Form FactorHHHL
InterfacePCIe Gen4
Manufacturer Part NumberMCX653106A-ECAT
Product SeriesConnectX-6 EN
Network Interface Speed100GbE
Number of Ports1 x 100GbE
Port Connector TypeQSFP56
Host InterfacePCIe 4.0 x16
RDMA SupportRoCEv2 (RDMA over Converged Ethernet v2)
GPUDirect RDMASupported
SR-IOV Virtual FunctionsUp to 127 VFs per port
Tunneling OffloadVXLAN, NVGRE, MPLS
Stateless OffloadsTSO, LRO, IP/TCP/UDP checksum offload
Flow ControlIEEE 802.1Qbb Priority Flow Control (PFC); IEEE 802.1Qaz ETS
Form FactorStandard height, half-length PCIe card
Operating Temperature0°C to 55°C
Software StackNVIDIA MLNX_OFED; OpenFabrics Alliance (OFED) drivers
OS SupportLinux (RHEL, Ubuntu, SLES), Windows Server
Power Consumption (typical)Approximately 18W
RoHS ComplianceRoHS compliant

Frequently Asked Questions about NVIDIA MCX653106A-ECAT ConnectX-6 EN Adapter

What does the NVIDIA MCX653106A-ECAT ConnectX-6 EN Adapter do?

The NVIDIA MCX653106A-ECAT ConnectX-6 EN Adapter is enterprise networking hardware for data-center, campus, and branch deployments. Common roles include core/distribution switching, server uplinks, top-of-rack fabric, and edge connectivity in mixed Cisco/Juniper/Arista environments.

What are the headline specs of the NVIDIA MCX653106A-ECAT ConnectX-6 EN Adapter?

Key specifications for the NVIDIA MCX653106A-ECAT ConnectX-6 EN Adapter: interface PCIe 4.0; new condition; adapter family ConnectX-6 VPI Adapter; network ports 2x QSFP56; supported speed 100GbE / HDR100 / EDR; host interface PCIe 4.0 x16; form factor HHHL; use case High-performance server networking, virtualization, storage, AI, and cloud. Manufacturer part number MCX653106A-ECAT. For the full datasheet with electrical, environmental, and compliance details, contact our pre-sales engineering team.

Is the NVIDIA MCX653106A-ECAT ConnectX-6 EN Adapter compatible with my infrastructure?

The NVIDIA MCX653106A-ECAT ConnectX-6 EN Adapter interoperates with standard switching protocols (LACP, MLAG/VPC, 802.1Q, OSPF/BGP). For specific cross-vendor scenarios — Cisco-to-Arista MLAG peering, fabric integration with Cumulus/SONiC, or migration from legacy Catalyst — our network engineers map a compatibility plan as part of the RFQ.