IndustrySelected

Lambda deploys CPO networking across 10,000+ Blackwell Ultra GPUs

Why we picked it

Production evidence now exists that CPO cuts networking power by up to 5x at 10,000-GPU scale, which shifts datacentre power planning from optics budgets toward compute. Operators designing large training clusters should track CPO switch availability when planning next-generation builds. The deployment also sets a concrete reference point against conventional pluggable optics for anyone benchmarking interconnect efficiency at campus scale.

Lambda has completed the world's first large-scale deployment of co-packaged optics (CPO) networking, linking more than 10,000 NVIDIA Blackwell Ultra GPUs with NVIDIA Photonics networking switches. The deployment was announced at SC26.

Body

Compared with conventional pluggable optics, Lambda measured up to 10% lower latency and up to 5x better power efficiency, with error rates holding steady as the fabric scaled. In CPO designs, the optical engine is packaged directly with the switch ASIC, shortening electrical paths and cutting the power that conventional pluggable transceivers consume.

For AI datacentre operators, the gains translate directly into economics: more of the power budget is available for compute rather than networking, and steadier error rates mean fewer interruptions for teams running long training jobs. The result is an early signal that CPO is moving from vendor demos to production-scale AI clusters.

Original · Lambda

At an AI campus scale, the network puts the “super” in supercomputer.

Lambda has completed the world’s first large-scale deployment of co-packaged optics (CPO) networking, connecting 10,000+ NVIDIA Blackwell Ultra GPUs using NVIDIA Photonics networking switches.

Compared with conventional pluggable optics, we observed up to 10% lower latency, up to 5x better power efficiency, and error rates that held steady as the fabric grew.

That means more of the power budget goes to compute, and fewer interruptions for teams running long training jobs.

Read the announcement: https://t.co/yY8JPM7QEZ

@NVIDIA @NVIDIANetworking #SC26

Industry Coordinates & Source Verification

This report is categorized under AITOP's "Industry" desk, focusing on US frontier AI labs and Silicon Valley technology leaders. Within our cross-regional analytical framework, it directly maps to AI developer toolchains, coding agents, and engineering workflows. Key entities and technical topics under continuous tracking include Lambda, NVIDIA, Blackwell Ultra, CPO.

Source attribution: curated from Lambda (@LambdaAPI).