HPE P18457-B21

Ethernet 10 Gb 2‑Port 562FLR‑SFP+ Network Adapter (FlexibleLOM, SFP+ Ports)

The HPE P18457‑B21 is a FlexibleLOM Network Adapter  featuring two 10 Gb/s SFP+ ports integrated via a PCIe 3.0 ×8 interface. Powered by an Intel X710 chipset, it delivers near line-rate dual‑port performance (aggregate upto 40 Gb/s full duplex) in select HPE ProLiant Gen 9/Gen 10 servers. Designed for virtualization-heavy or throughput-intensive workloads, this adapter includes enterprise-level features such as SR‑IOV, RSS, VXLAN/NVGRE stateless offload, jumbo frame support, checksum and segmentation offloads, PXE boot, and precision time‑stamping (IEEE 1588). Ideal for high-density virtualization, HPC, cloud I/O, or converged networking, it boosts performance and scalability while offloading CPU workloads.

Features

Built on the Intel X710 controller, the 562FLR SFP+ adapter supports two 10 Gb/s full-duplex ports via SFP+ fiber or DAC connectivity. With PCIe 3.0 ×8 bus support, it enables aggregate throughput up to 40 Gb/s—20 Gb/s per port. It includes advanced network acceleration features such as SR‑IOV for VM I/O sharing, VXLAN and NVGRE stateless offloads for overlay networking, RSS for CPU-core balancing, and large receive/send offload for efficient TCP/IP packet processing.The adapter is validated on HPE ProLiant Gen9 and Gen10 server lines such as DL360, DL380, DL385, DL560, DL580, and Apollo 4000/2000 series platforms. It supports HPE Active Health System monitoring and integrates with Server Insight tools. The FlexibleLOM form factor lets it plug into supported HPE servers without consuming PCIe slots. Offered both as OEM-new (with remaining server warranty or lifetime support options depending on reseller) and refurbished (commonly ~90‑day warranty from resellers).

Advanced Options

Engineered for performance and scalability, this adapter minimizes host CPU load through hardware offloads and delivers predictable throughput across both ports. SR‑IOV support enables single physical adapter to serve multiple VMs with near-native performance. VXLAN/NVGRE offloading enables high-speed overlay networking without burdening CPU cores—ideal for multi-tenant or containerized deployments. Jumbo frame (up to 9.5 KB) support improves bandwidth efficiency for large I/O applications.