The Aurora X2-81 has a fully programmable 12.8T Xsight Labs X2 chip at its heart. It delivers ultra‑low‑power, low‑latency connectivity to scale AI, storage, and compute clusters, custom packet-broker/network-orchestration, and forensic infrastructure with zero trade‑offs.

With 32x400G ports, the Aurora X2-81 supports 2x200G and 4x100G breakout configuration for leaf or ToR cloud deployments. 2x SFP56 X2 management ports and 2x 10G SFP+ control plane ports give flexibility.

Featuring Xsight Instruction Set Architecture (ISA), the Aurora X2-81 enables tailored packet processing, routing protocols, and dynamic network policies. Such flexibility enables advanced features without performance penalties, paving the way for Ultra Ethernet Transport (UET) protocol support to scale out massive AI training clusters and High-Performance Computing (HPC) networks over standard Ethernet.

The Aurora X2-81 supports a timing option with 1 PPS and 10MHz inputs for Synchronized, Deterministic Networking.

The system supports Netberg SONiC as the complete switching package, Ubuntu Linux and X-SDK (Software Development Kit) for custom application development, and InfiniCore Visibility Studio for network visibility and management solutions.

Use cases

• Scale the future of networking with open, programmable performance.
  
  Scale AI, compute, and storage at will. Stay efficient at scale. Extend capabilities without replacing hardware.
  
  AI and GPU clusters, cloud and data center networks, and distributed and edge deployments greatly benefit from consistent results under real-world conditions.
• Precise Control over Packets
  
  Network Packet Brokers (NPB), Broadband Network Gateways (BNG), Load Balancers, and AI and Machine Learning apps benefit immensely from the flexible X-Switch parallel execution architecture. Unlike traditional match-action pipelines, X-Switch executes instructions across programmable micro-engines simultaneously, maintaining line rate without recirculation penalties.
• UEC-ready for emerging AI networking standards.
  
  X-IQ™ (Intelligent queuing) programmable congestion measurement and signaling to facilitate dynamic load balancing and fast failure detection and recovery. It dynamically optimizes packet processing to reduce "tail latency" and improve Flow Completion Times (FCT), ensuring that AI training jobs aren't stalled by a few slow packets.
• Real-time telemetry.
  
  X-View™ provides real-time telemetry for best-in-class network visibility. It allows operators to identify congestion, packet loss, or latency issues instantly, enabling proactive management of AI fabrics.
• Advanced timing options.
  
  IEEE 1588-2008 PTPv2 and SyncE support greatly benefits time-sensitive financial applications, 5G networks, broadcast streaming, and network packet brokering/orchestration NOS.
• Designed and manufactured in Taiwan.

Elite performance and efficiency

• 32x 400GbE QSFP112 ports in 1 RU for high-density applications.
  16 top ports support 24W transceivers.
  2x200G and 4x100G breakout configurations per port.
  
  2× 10/25/50GbE SFP56 X2 management ports
  2x 10G SFP+ CPU ports
• 12.8 Tbps Bandwidth with a fully shared packet buffer.
  Dynamic buffer allocation across all ports. No stranded memory; maximum utilization under variable traffic patterns.
• Under 200W power consumption at full bandwidth for reduced power and cooling demands.
• Under 700 ns first-bit-in-first-bit-out latency.
• Forwarding rate 5.8 Gpps
• Intel Xeon D-1713NT for robust management and application deployment.

Network OS (NOS) options

• SONiC Distribution by Netberg is a hardened, open-source distribution of SONiC (Software for Open Networking in the Cloud) that runs on Netberg Aurora switches. It meets the needs of AI backend fabrics, cloud, enterprise, and campus network environments.
• Ubuntu Linux for application development.
  
  Drive continuous innovation through fully programmable, software-defined switching. XISA gives you assembly-level access to programmable engines. No vendor lock-in.
  
  X2 integrates with industry-standard network operating systems and provides a complete development toolkit for custom implementations.
• InfiniCORE Network Visibility Studio transforms the Aurora X2 series into a visibility powerhouse by leveraging the programmability of the Xsight Labs X2 chipset. Visibility Studio provides a full data path design environment, known as the NextIO Design Environment, directly from the Web-based UI.