Cornelis Networks has deployed its CN5000 intelligent networking solution at the Lawrence Livermore National Laboratory (LLNL) as part of a major upgrade to the U.S. Department of Energy’s (DOE) National Nuclear Security Administration (NNSA) computing infrastructure. The CN5000 will power the new 952-node “Lynx” cluster under the NNSA’s Commodity Technology Systems-2 (CTS-2) program, which supports the agency’s Advanced Simulation and Computing (ASC) initiatives.
The CN5000 platform, launched in June 2025, is purpose-built to support AI and high-performance computing (HPC) environments at scale. It features a fully integrated stack with patented congestion management, ultra-low latency, and lossless data transfer to minimize bottlenecks and improve compute utilization. The deployment at LLNL stems from a three-year joint development effort under the Next-Generation High Performance Computing Network (NG-HPCN) project between Cornelis and the NNSA.
Cornelis offers an end-to-end solution including air- and liquid-cooled SuperNICs, 48-port and 576-port modular switches, high-speed cabling, and open-source management tools. These capabilities are expected to significantly enhance the performance of LLNL’s simulation and AI workloads, which are central to ensuring the safety and reliability of the U.S. nuclear stockpile.
- CN5000 deployed on LLNL’s 952-node Lynx cluster under CTS-2 initiative
- Purpose-built for AI and HPC workloads with low latency and congestion-free design
- Jointly developed with the NNSA under the NG-HPCN project
- System includes switches, SuperNICs, and open-source management software
- Supports critical U.S. national security and simulation research missions
“We’re pleased to partner with Cornelis on CN5000 and look forward to leveraging its performance and features to accelerate our mission-critical HPC and AI workloads,” said Matt Leininger, Senior Principal HPC Strategist at LLNL.
🌐 We’re tracking the latest developments in networking silicon. Follow our ongoing coverage at: https://convergedigest.com/category/semiconductors/
