en.Wedoany.com Reported - VIAVI Solutions Inc. (VIAVI) has launched the industry's first Ultra Ethernet Transport (UET) verification solution for AI architectures, providing traffic generation and analysis capabilities for hyperscale cloud providers, cloud and emerging cloud providers, and network equipment manufacturers.

This new solution for the VIAVI TestCenter platform adopts a GPU-free, full-fidelity design to verify scale-out and scale-up AI backend networks using the Ultra Ethernet Consortium (UEC) Ultra Ethernet protocol stack, accelerating the deployment of next-generation high-speed AI networks. The Ultra Ethernet protocol stack is specifically designed for large-scale AI and high-performance computing (HPC) workloads. UEC 1.0 introduces the new UET protocol, featuring advanced congestion control and massive scalability, aimed at driving multi-vendor adoption of Ethernet as the primary AI transport technology.

This UET verification solution enables customers to verify the performance and resilience of AI architectures without deploying dedicated GPU infrastructure. The solution emulates the UET transport layer, replicating realistic stateful AI traffic patterns at scale, including reliable ordered and unordered delivery (ROD/RUD), packet trimming, congestion control, and dynamic multipathing. It also supports full-fidelity simulation of AI workloads, such as collective communications (CCL) and large language model (LLM) traffic. The platform further supports verification of load balancing mechanisms in AI architectures, including ECMP, packet spraying, and flowlet switching.

Aniket Khosla, Vice President of Product Management for Optical Transport and High-Speed Ethernet at VIAVI, stated that AI clusters will soon scale to millions of endpoints, making it impractical to rely solely on physical GPUs for network behavior verification. While Ultra Ethernet brings high throughput and standardization benefits to AI and HPC, its success depends on rigorous, realistic testing. This GPU-free, full-fidelity solution enables customers to deploy scalable, high-performance AI architectures more quickly and cost-effectively.

Mahesh Subramaniam, Senior Director of Product Management for HPE AI Data Center, noted that as AI data centers grow in scale and complexity, traditional verification methods struggle to keep pace. The VIAVI TestCenter platform, through realistic traffic simulation and granular visibility, can verify congestion control and transport performance in large-scale environments. Collaborative UET transport verification using the Juniper QFX5240 platform and Junos Evolved (featuring advanced software capabilities such as packet trimming) demonstrates how advanced switching technology and realistic traffic testing can accelerate the deployment and reliable scaling of Ultra Ethernet in next-generation AI networks.

This article is compiled by Wedoany. All AI citations must indicate the source as "Wedoany". If there is any infringement or other issues, please notify us promptly, and we will modify or delete it accordingly. Email: news@wedoany.com