
Subscribe To Our Newsletter!
Subscribe to the newsletter to stay up to date with the latest news and most useful
Newsletter
↑
Back to Top
| Product Overview |
|---|
The MCX654105A-HCAT ConnectX-6 VPI adapter card enables high-speed networking through a single QSFP56 port using either HDR InfiniBand or 200Gb Ethernet. It fits into servers via a PCIe 3.0 x16 slot and is designed to handle demanding data center tasks where speed and flexibility are important. Its Socket Direct feature helps reduce latency by connecting directly to the CPU, making it suitable for workloads that require quick data movement. |
| General Information | |
|---|---|
| Brand | Mellanox |
| Part Number | MCX654105A-HCAT |
| Technical Information | |
|---|---|
| Chipset/Controller | Mellanox ConnectX-6 |
| Virtualization Support | SR-IOV, VXLAN, NVGRE, Geneve, NetQueue, vNIC |
| Data Rate (Speed) | 200GbE |
| Data Center Bridging | Yes |
| Storage Over Ethernet | iSCSI, SMB Direct |
| Interface (Bus) | PCIe 3.0 x16 |
| Number of Ports | 1-Port |
| OS Compatibility | Windows Server 2022, VMware ESXi 8.0, RHEL, VMware ESXi |
| Network Management | SNMP, RMON, NC-SI |
| Offload Support | RDMA, RoCE v2, PXE Boot, RoCE v1, iWARP, Jumbo Frames |
| Bracket Height | Full Height (Standard) |
| Connectivity & Physical | |
|---|---|
| Form Factor | Standard PCIe Card |
| Port Type | QSFP56 |
| Cabling Type | AOC |
| Weight | 5.00 |
| Condition | Refurbished |
| Miscellaneous | |
|---|---|
| Assembly Required | Yes |
| Eco Friendly | Yes |
| Compliance Standards | WEEE, RoHS, CE, FCC, UL, TUV, cULus |
| Product Description |
|---|
The Mellanox MCX654105A-HCAT is built for high-performance computing environments requiring fast, reliable networking. It often appears in data centers and enterprise servers where ultra-low latency and high throughput are essential. Professionals managing HPC clusters or storage networks benefit from its flexible dual-protocol support. Key Features
This adapter is commonly deployed in environments where data moves rapidly between nodes, such as HPC, AI workloads, and large-scale storage. It helps reduce bottlenecks and supports demanding applications by delivering consistent throughput with low latency. |