HPE P11671-001

StoreFabric SN2700M 32‑Port 100 GbE Managed Switch (QSFP28, Power‑to‑Connector Airflow)

The HPE P11671‑001, also known as the StoreFabric SN2700M, is a compact 1U rack-mount Ethernet switch geared for ultra-high performance deployments. Featuring 32 fixed QSFP28 ports each capable of 100 GbE, and with breakout support (50 GbE, 25 GbE, 10 GbE), it provides flexible migration paths. With non-blocking 6.4 Tb/s switching capacity and 9.52 Bpps throughput, the SN2700M is ideal for NVMe-oF, HPC clusters, AI workloads, and storage fabrics. True cut-through design delivers sub‑300 ns port-to-port latency, while the power-to-connector airflow architecture and dual hot-swappable PSUs/fans ensure reliability in high-density rack setups. ([turn0search1]turn0search6turn0search0)

Features

The SN2700M P11671‑001 supports 32 QSFP28 ports, each operating at 100 GbE native or breaking out into 2×50 GbE, 4×25 GbE, or 4×10 GbE links using appropriate optics or DAC—allowing maximum flexibility across network overlays. Its robust cut-through switching fabric guarantees zero packet loss even under mixed-speed traffic, while maintaining ultra-low latency (~300 ns). At only ~150 W typical power draw, it offers one of the most efficient 100 GbE platforms in its class. ([turn0search1]turn0search6turn0search4)

Advanced Options

The switch supports management via Mellanox Onyx or ONIE with CLI, WebUI, and automation-friendly telemetry tools such as sFlow, queue histograms, and error tracking. Routing and Fabric features include VXLAN, MLAG, BFD, VRRP, ECMP, DCB, OpenFlow, IGMP snooping, and STP variants. A TAA compliant build ensures suitability for government environments. Memory specs include 8 GB RAM, 32 MB flash, and 16 MB packet buffer. Management ports include 2 × 1 Gb RJ45, 1 × console RJ45, and 1 × USB port. Fans (n+1) and redundant PSUs support both power-to-connector and connector-to-power airflow modes. ([turn0search5]turn0search9turn0search1)Designed for performance-critical environments, the SN2700M ensures line-rate switching even under diverse traffic loads, without packet loss. Its cut-through architecture delivers stunningly low latency of ~300 ns—ideal for storage fabrics, AI training clusters, and microservices architecture. ([turn0search1]turn0search4)