48 x 25Gb SFP28 Data Center Leaf/Spine Switch

Open NOS, full L2/L3 and a consistent operating experience

The switch runs AsterNOS, Asterfusion's enterprise distribution of open-source SONiC, with a single software binary spanning the entire CX-N-V2 family — no per-feature license tiers, no vendor lock-in. On top of it sits a complete Layer 2/Layer 3 feature set: VXLAN, EVPN Type 1–5, distributed/anycast gateway, VRF and MC-LAG for multi-tenant overlays, plus full BGP, OSPF v2/v3, IS-IS, ECMP and policy-based routing for the underlay. Operators get a consistent management experience across every box — industry-standard CLI, REST API, NETCONF, Zero-Touch Provisioning and Syslog — so the same automation and tooling carry from one switch to the whole fabric.

High-precision time synchronization

Timing is a first-class capability on the CX308P-48Y-N-V2. The switch provides IEEE 1588v2 PTP with Class C accuracy and supports the full range of industry timing profiles — G.8275.1, G.8275.2, SMPTE 2059-2 and AES67 — across flexible clock roles including GrandMaster (GM), Boundary Clock (BC) and Ordinary Clock (OC). Combined with SyncE for physical-layer frequency distribution, it delivers the deterministic, low-jitter timing that AVoIP, broadcast, financial and industrial time-sensitive networks depend on.

Comprehensive multicast and RoCEv2 transport

For one-to-many media and data distribution, the switch offers a complete multicast toolset — IGMP v1/v2/v3, IGMP Snooping, IGMP Querier, PIM-SM and multicast routing — to keep AVoIP and broadcast traffic efficient and loss-free across large fabrics. For storage and AI/ML workloads, it adds RoCEv2 with PFC, PFC Watchdog, ECN, DCBX, Fast CNP and one-step Easy RoCE configuration, so a lossless RDMA fabric can be brought up in a single command.

Built-in observability

The CX308P-48Y-N-V2 is designed to be transparent in production. A native AsterNOS Prometheus exporter streams real-time switch metrics to your monitoring stack and renders them on ready-made Grafana dashboards, while sFlow, SPAN/ERSPAN and SNMP v1/v2/v3 provide flow sampling, traffic mirroring and standards-based management. Together they give operators per-port, real-time visibility for fast troubleshooting and capacity planning.

Carrier-grade reliability

Built for always-on data centers, the switch uses hot-swappable 1+1 redundant power supplies and 3+1 (N+1) redundant, hot-swappable fans to eliminate single points of failure. Flexible AC, DC and HVDC input options and selectable front-to-rear or rear-to-front airflow let it fit cleanly into hot/cold-aisle containment and mixed power environments — making the CX308P-48Y-N-V2 a dependable open-networking foundation for cloud data centers, HPC, distributed storage and AVoIP deployments.

Architecture & performance

• 9th-generation, 16nm Marvell Prestera Falcon — 2.0 Tbps to 12.8 Tbps wire-speed switching across the 98CX85xx family
• Large shared tables: 128K MAC, 288K IPv4 / 144K IPv6 prefixes, high-scale multi-tenant L2/L3
• Globally shared packet buffer with dynamic thresholds for microburst (µBurst) absorption
• Cut-through switching and dynamic load balancing (DLB) across the fabric

Virtualization & overlays

• eBridge unified virtualization architecture with hardware-based interface/domain virtualization
• Programmable header editor: hardware-accelerated VXLAN, VXLAN-GPE, Geneve and NSH
• MPLS and IPv6 segment routing for flexible transport

RDMA, telemetry & openness

• RoCEv2-aware congestion management for lossless RDMA storage and AI/ML traffic
• In-band network telemetry built into the data path — sFlow, IPFIX, remote port analysis
• Open Switch Abstraction Interface (SAI) + Telemetry & Monitoring (TAM) APIs — merchant-silicon foundation for SONiC/AsterNOS