MCX653106A-HDAT ConnectX-6 200GbE Dual-Port QSFP56 NIC
MCX653106A-HDAT ConnectX-6 200GbE Dual-Port QSFP56 NIC is a high-performance NVIDIA Mellanox VPI network adapter designed for AI servers, GPU clusters, high-performance computing systems, enterprise data centers, and high-speed storage environments.
Based on NVIDIA ConnectX-6 technology, the adapter features dual QSFP56 ports and supports both Ethernet and InfiniBand networking. The MCX653106A-HDAT ConnectX-6 200GbE Dual-Port QSFP56 NIC supports Ethernet speeds up to 200GbE and InfiniBand speeds up to HDR 200Gb/s, providing high-bandwidth and low-latency connectivity for demanding AI and HPC infrastructure.
Key Features
- NVIDIA Mellanox ConnectX-6 VPI architecture
- Part Number: MCX653106A-HDAT
- Dual-port QSFP56 connectivity
- Up to 200GbE Ethernet
- HDR InfiniBand up to 200Gb/s
- Supports HDR100, EDR, FDR, QDR, DDR, and SDR InfiniBand
- Ethernet and InfiniBand VPI capability
- PCIe 4.0 x16 host interface
- Tall bracket configuration
- Suitable for AI servers and GPU clusters
- Designed for HPC and enterprise data centers
- High-speed storage and server networking
The MCX653106A-HDAT ConnectX-6 200GbE Dual-Port QSFP56 NIC is suitable for organizations requiring high-throughput Ethernet or InfiniBand connectivity for large-scale compute and storage environments.
Product Specifications
| Item | Specification |
|---|---|
| Brand | NVIDIA Mellanox |
| Model | MCX653106A-HDAT |
| Series | ConnectX-6 VPI |
| Product Type | InfiniBand / Ethernet Network Adapter |
| Ethernet Speed | Up to 200GbE |
| Supported Ethernet Speeds | 1/10/25/40/50/100/200GbE |
| InfiniBand Speed | Up to HDR 200Gb/s |
| InfiniBand Support | SDR / DDR / QDR / FDR / EDR / HDR100 / HDR |
| Number of Ports | 2 |
| Port Type | Dual QSFP56 |
| Host Interface | PCIe 4.0 x16 |
| Bracket Type | Tall Bracket |
| Application | AI Server, GPU Server, HPC, Data Center |
| Typical Use | AI Clusters, HPC, Storage Networking, Cloud |
AI Server and HPC Applications
The MCX653106A-HDAT ConnectX-6 200GbE Dual-Port QSFP56 NIC is designed for high-bandwidth AI training clusters, GPU computing nodes, HPC systems, distributed storage platforms, and enterprise data centers.
Its 200GbE Ethernet capability is suitable for large-scale data movement between GPU servers, storage systems, and compute nodes. HDR InfiniBand support also makes the adapter suitable for low-latency interconnect fabrics used in HPC and AI cluster environments.
Typical Applications
The adapter can be deployed in:
- AI training clusters
- GPU server interconnects
- HPC computing environments
- Distributed storage systems
- High-speed storage networks
- Enterprise data centers
- Private and public cloud platforms
- Virtualization environments
- Server network upgrades
- Ethernet and InfiniBand converged infrastructure
The MCX653106A-HDAT ConnectX-6 200GbE Dual-Port QSFP56 NIC is particularly suitable for infrastructure where 100GbE networking no longer provides sufficient bandwidth for compute, storage, or AI workloads.
Ethernet and InfiniBand VPI Support
MCX653106A-HDAT is based on NVIDIA VPI technology, allowing the adapter to support both Ethernet and InfiniBand networking.
For Ethernet environments, it supports network speeds up to 200GbE. For InfiniBand deployments, it supports speeds up to HDR 200Gb/s.
This flexibility allows system integrators and data center operators to deploy the same adapter platform across AI, HPC, storage, and cloud environments with different network architectures.
PCIe 4.0 x16 Host Interface
The MCX653106A-HDAT ConnectX-6 200GbE Dual-Port QSFP56 NIC uses a PCIe 4.0 x16 host interface.
PCIe 4.0 x16 provides substantially more host-side bandwidth than PCIe 3.0 x16, making it better suited to 200Gb/s-class networking applications and high-throughput server workloads.
For maximum performance, the adapter should be installed in a server platform that supports PCIe 4.0 x16 and provides sufficient cooling and power.
MCX653106A-HDAT vs MCX653106A-ECAT
Both models belong to the ConnectX-6 VPI family and use dual QSFP56 ports, but their maximum network speeds differ significantly.
| Model | Ethernet | InfiniBand | Ports | PCIe |
|---|---|---|---|---|
| MCX653106A-ECAT | Up to 100GbE | Up to HDR100 | Dual QSFP56 | PCIe x16 |
| MCX653106A-HDAT | Up to 200GbE | Up to HDR 200Gb/s | Dual QSFP56 | PCIe 4.0 x16 |
MCX653106A-HDAT is the higher-bandwidth option for AI, HPC, and data center environments requiring 200Gb/s-class networking.
About Us
Why Choose MCX653106A-HDAT?
The MCX653106A-HDAT ConnectX-6 200GbE Dual-Port QSFP56 NIC combines dual-port 200GbE Ethernet, HDR InfiniBand, PCIe 4.0 x16, and NVIDIA ConnectX-6 VPI technology.
It is suitable for customers requiring:
- Dual-port high-speed networking
- 200GbE Ethernet connectivity
- HDR 200Gb/s InfiniBand
- AI and GPU cluster networking
- HPC interconnects
- High-performance storage networks
- PCIe 4.0 server platforms
FAQ
Q: What is MCX653106A-HDAT?
A: MCX653106A-HDAT is an NVIDIA Mellanox ConnectX-6 VPI network adapter featuring dual QSFP56 ports and support for 200GbE Ethernet and HDR InfiniBand up to 200Gb/s.
Q: Does MCX653106A-HDAT support 200GbE?
A: Yes. It supports Ethernet speeds up to 200GbE.
Q: Does MCX653106A-HDAT support InfiniBand?
A: Yes. It supports InfiniBand protocols up to HDR 200Gb/s.
Q: How many ports does MCX653106A-HDAT have?
A: It features two QSFP56 network ports.
Q: What PCIe interface does MCX653106A-HDAT use?
A: It uses a PCIe 4.0 x16 host interface.
Q: Is MCX653106A-HDAT suitable for AI servers?
A: Yes. It is suitable for AI training servers, GPU clusters, HPC systems, high-speed storage networks, and large-scale data center infrastructure.
Q: What is the difference between MCX653106A-HDAT and MCX653106A-ECAT?
A: MCX653106A-HDAT supports up to 200GbE Ethernet and HDR 200Gb/s InfiniBand, while MCX653106A-ECAT is limited to 100GbE and HDR100.
Q: Can CubeCore supply MCX653106A-HDAT in bulk?
A: Yes. CubeCore supports tested hardware supply, model verification, bulk orders, project-based procurement, and global delivery.
