AI Data Center Networking Test
Validate the high-speed Ethernet fabrics, Ultra Ethernet Transport (UET) deployments, and AI inference pipelines powering next-generation AI data centers and applications, from 800G and 1.6T TestCenter testing to AI Inference load testing with CyberFlood.
AI workloads are reshaping data center networking. Hyperscalers, cloud providers, and AI-native companies demand test solutions that mirror real-world AI traffic patterns, from large-scale model training across 800G and 1.6T Ethernet fabrics using the new UEC 1.0 specification, to real-time AI inference behind production applications. VIAVI delivers an end-to-end test portfolio spanning the full AI data center stack.
-
High-speed fabric validationTestCenter emulates advanced RoCEv2 and CCL traffic, GPUs, NICs, and Servers from 400G to 1.6T
-
Lifecycle & Regression Support128 GPU/NIC emulation per port. Native CCL and LLM training traffic at production scale
-
Ultra Ethernet Transport (UET)First field-demonstrated UEC 1.0 solution. AI workloads, congestion control, and load balancing at 800G to 1.6T
-
Advanced AI Performance MetricsJCT, tail latency, and advanced packet-level stats for AI fabric benchmarking and performance isolation
-
AI Inference testingCyberFlood stress-tests LLM inference endpoints at production scale, measuring TTFT, throughput, and latency
-
DPU and SmartNIC AssessmentCyberFlood validates DPU and SmartNIC offload performance under realistic AI and cloud-native traffic mixes
TestCenter, High-speed Ethernet & AI fabric testing
Industry-leading Ethernet test platform for validating AI training fabrics. Delivered on the A1 400G, B3 800G and D2 1.6T appliances, supporting RoCEv2, CCL AI workload emulation, congestion control, and multi-path load balancing at wire-rate speeds.
Ultra Ethernet Transport (UET), UEC 1.0
The first field demonstrated UET testing solution. A software license on the B3 or D2 appliance adds full UEC 1.0 transport validation: packet spraying, packet trimming, Programmable Congestion Management (PCM), Reliable Ordered & Unordered Delivery (ROD/RUD), ECMP and Adaptive Packet Spraying load balancing, and AI fabric behavior emulation with CCL/LLM traffic at scale.
Key capabilities:
- AI fabric emulation, a library of native CCL algorithms, mixed-algorithm and background traffic support
- LLM traffic emulation for ChatGPT, Llama, DeepSeek, RestNet, AlexNet
- Tail latency metrics: JCT, P95/P99/P99.9, BusBW, out-of-sequence packets
- Congestion control: NSCC and RCC; ROD/RUD on UD/IP and IP
- Multi-rate: 100G, 200G, 400G, 800G, 1.6T on a single platform
- Scale to multi-terabit router validation for hyperscale environments
- UET software license available: B3 800G and D2 1.6T Appliances
TestCenter D2 1.6T 4-port Appliance
Purpose-built 1.6T test appliance with four OSFP ports of line-rate traffic generation in a compact form factor, ideal for switch and transceiver validation labs, and the platform for UET 1.6T testing.
Key capabilities:
- 4 × 1.6T OSFP ports, line-rate wire-speed traffic
- Multi-rate: 1.6T, 800G, 400G, 200G operation
- UET software license available
- Integrated with TestCenter software ecosystem
CyberFlood AI inference testing
Purpose-built AI inference load testing for LLM serving infrastructure. Emulates realistic prompt traffic, measures latency under load, and stress-tests GPU clusters and inference endpoints at scale.
Key capabilities:
- LLM prompt emulation with configurable token lengths
- Time-to-first-token (TTFT) and end-to-end latency measurement
- Throughput benchmarking: tokens/sec and requests/sec
- Multi-model, multi-endpoint concurrency testing
- OpenAI-compatible API support for broad ecosystem coverage
CyberFlood, DPU & SmartNIC assessment
Validate DPU and SmartNIC offload performance under realistic AI and cloud-native traffic. Characterize throughput, latency, and offload efficiency across security, storage, and networking functions.
Key capabilities:
- Offload function benchmarking: crypto, compression, vSwitch
- Realistic multi-tenant traffic mix emulation
- Latency characterization under load
- Support for major DPU and SmartNIC vendors
VIAVI AI Workload Traffic Emulation
Testing the AI Fabric
Bracing for AI’s Impact on Data Center Networking
Understand AI’s potential impact on data center networking, telecoms and enterprise digital transformation efforts, and gain actionable insights to better align your business with the AI revolution.
我们倾力相助
我们倾力相助，为您获得成功加油。