Huawei CloudEngine 9860 V2

The Huawei CloudEngine 9860 V2 is a high-end modular core switch engineered for large-scale data center and campus network backbones, delivering terabit-level switching capacity through a distributed architecture. Built on Huawei's advanced ASIC technology, it supports high-density 100GE and 400GE interfaces, enabling organizations to consolidate spine-layer connectivity, east-west traffic aggregation, and north-south gateway functions within a single platform. The CE9860 V2 supports Huawei's iLossless intelligent lossless network algorithm, making it well suited for RDMA over Converged Ethernet (RoCE) traffic inherent in AI training clusters and high-performance computing environments.

The chassis employs a fully redundant design with hot-swappable line cards, fabric modules, and power supplies, supporting non-stop forwarding (NSF) and graceful restart (GR) to deliver carrier-grade availability. It is compatible with both traditional VXLAN-based overlay networking and segment routing (SR-MPLS / SRv6), enabling seamless integration into both legacy three-tier architectures and modern spine-leaf fabrics. Telemetry streaming via gRPC and support for NETCONF/YANG-based programmability allow operations teams to integrate the switch into modern intent-based network automation pipelines.

The CloudEngine 9860 V2 is managed through Huawei's iMaster NCE controller, which provides unified network-wide visibility, policy orchestration, and AI-driven traffic optimization. Its support for FlexE (Flexible Ethernet) and hard-slicing capabilities allows network administrators to carve deterministic low-latency channels across shared infrastructure, a critical requirement for mixed workloads spanning storage, compute, and AI fabric traffic. The platform is a primary choice for hyperscale enterprise data centers, cloud-native service providers, and organizations building GPU cluster interconnect fabrics across the UAE, GCC, EMEA, and APAC regions.

Ideal for

• Spine-layer switching in large-scale data center spine-leaf fabrics requiring high-density 100GE/400GE connectivity and lossless RoCE transport for AI and HPC workloads
• Core aggregation node for enterprise campus networks consolidating multi-building distribution layers into a single high-availability chassis platform
• GPU cluster interconnect fabric backbone supporting RDMA traffic with iLossless congestion control for distributed deep learning training jobs
• Cloud data center gateway providing VXLAN BGP EVPN-based overlay with hardware-accelerated routing at multi-terabit scale
• Multi-tenant network slicing infrastructure using FlexE hard-slicing to deliver deterministic SLA-bound channels across shared physical switching infrastructure
• Carrier and service provider edge aggregation requiring SRv6 programmable forwarding planes and high-precision streaming telemetry for real-time network operations

Technical specifications

Available from Omnixon Global. Submit an RFQ and our team will confirm configuration and availability for your order.