-

TACC Accepts Cornelis CN5000 Upgrade to Accelerate Scientific Computing on Stampede3

National Science Foundation-approved deployment upgrades more than 600 compute nodes and delivers up to 71% performance improvements for scientific computing workloads.

Summary:

  • TACC formally accepted a CN5000 fabric upgrade supporting the Stampede3 supercomputer.
  • Upgrade deploys Cornelis CN5000 networking technology across more than 600 compute nodes.
  • Testing using the Weather Research and Forecasting (WRF) model demonstrated performance gains ranging from 53% to 71%.
  • Deployment enhances performance and scalability for scientific computing workloads serving the U.S. open science and open research community.

WAYNE, Pa.--(BUSINESS WIRE)--Cornelis announced the formal acceptance of a CN5000 networking upgrade supporting a key compute partition within the Texas Advanced Computing Center's (TACC) Stampede3 supercomputer.

The deployment upgrades more than 600 compute nodes to provide improved performance and scalability for scientific computing workloads serving the U.S. open science and open research community. The upgrade helps researchers complete simulations faster and supports increasingly complex research across weather forecasting, engineering, and data-intensive analysis.

"TACC has long been one of the most respected research computing organizations in the world, and we're proud to continue supporting its mission," said Lisa Spelman, CEO of Cornelis. "The acceptance of this upgrade marks another milestone in our collaboration with TACC and reflects our shared commitment to helping researchers tackle increasingly complex scientific challenges."

The acceptance further expands Cornelis' support of one of the world's leading academic supercomputing centers. Stampede3 serves more than 5,000 active researchers each year and powers a wide range of scientific and engineering applications, including weather forecasting, computational science, engineering simulation, and data analytics.

Accelerating Scientific Research

As scientific and engineering workloads continue to grow in complexity, efficient movement of data between compute resources becomes increasingly important to overall application performance.

The CN5000 upgrade enhances a core Stampede3 compute partition, helping improve application performance and enabling more effective use of shared compute resources.

“Researchers rely on Stampede3 to support a wide range of scientific discovery efforts," said Dan Stanzione, executive director of TACC. "The successful acceptance of the CN5000 upgrade strengthens an important production partition within the system and helps ensure researchers have access to the performance needed for increasingly demanding workloads. Rigorous acceptance testing is how we ensure this system delivers for the thousands of researchers who depend on it each year, and it reflects the standard we hold for every technology partner we bring into the ecosystem.”

Measurable Application Performance Improvements

Testing using the Weather Research and Forecasting (WRF) model, one of the HPC community’s most widely used weather and climate applications, demonstrated performance improvements ranging from 53% to 71% across multiple scaling points.

The results highlight how modern networking infrastructure can improve the performance of large-scale scientific applications, enabling researchers to complete simulations faster and maximize available computing capacity.

The upgrade further demonstrates Cornelis’ ability to support leading HPC organizations with networking technologies designed to improve application performance, system efficiency, and scientific productivity.

About Cornelis

Cornelis delivers high-performance, scale-out networking solutions that accelerate AI and HPC workloads. Built on the powerful Omni-Path architecture, Cornelis technology enables lossless, congestion-free networking that reduces training time, improves inference, and maximizes compute utilization. From foundation model training to complex climate modeling and real-time analytics, Cornelis' solutions power the most demanding workloads across commercial, academic, and cloud environments. With a focus on performance, scalability, and efficiency, Cornelis helps organizations achieve faster insights and greater return on infrastructure investments. Learn more at cornelis.com.

FAQ

  1. What is Cornelis announcing?
    Cornelis is announcing the formal acceptance of a CN5000 networking upgrade supporting a significant compute partition within TACC's Stampede3 system.

  2. What was upgraded?
    More than 600 Sapphire Rapids-based compute nodes were upgraded to CN5000 networking technology.

  3. What benefits does the upgrade provide?
    The upgrade delivers increased bandwidth, lower latency, and improved application performance for scientific computing workloads running on the upgraded partition.

  4. What performance improvements were observed?
    Benchmarking using the WRF weather forecasting application demonstrated significant application performance improvements compared with the previous networking infrastructure.

  5. Why is this announcement important?
    TACC is one of the world's leading academic supercomputing centers and a highly visible HPC deployment. Formal acceptance of the upgrade demonstrates successful deployment of CN5000 technology in a major production research environment. 

 

Contacts

Matt Stubbs
mstubbs@voxuspr.com
Voxus PR

Cornelis


Release Versions

Contacts

Matt Stubbs
mstubbs@voxuspr.com
Voxus PR

More News From Cornelis

Cornelis and NextSilicon to Build Joint Reference Architectures for AI and HPC

HAMBURG, Germany--(BUSINESS WIRE)--Cornelis and NextSilicon today announced at ISC High Performance 2026 a collaboration to build and evaluate joint reference architectures for AI and high-performance computing. The work pairs the Cornelis CN5000 fabric with the NextSilicon Maverick-2 compute platform. Joint evaluation is already underway, with the goal of commercialization through joint OEM partners. The collaboration starts with the 400 Gbps CN5000 fabric, launched in 2025, paired with Maveri...

Cornelis and xFusion Deliver High-Performance Infrastructure for European Automotive and Industrial HPC

WAYNE, Pa.--(BUSINESS WIRE)--Cornelis, a leading provider of intelligent networking solutions for AI and high performance computing (HPC), and xFusion, a global provider of computing infrastructure and services, announced a collaboration to deliver high-performance infrastructure solutions for industrial HPC and AI environments, including automotive and advanced engineering workloads across Europe.The joint platform combines xFusion’s FusionServer infrastructure with Cornelis’ CN5000 intelligent...

Cornelis Announces Deployment of “Lynx” Supercomputer at Lawrence Livermore National Laboratory

WAYNE, Pa.--(BUSINESS WIRE)--Cornelis, a provider of high-performance networking solutions, today announced the successful deployment of the “Lynx” supercomputing cluster at Lawrence Livermore National Laboratory (LLNL). The 952-node Lynx cluster, featuring Dell PowerEdge servers, Intel® Xeon® processors, and the Cornelis CN5000 Omni-Path fabric, is part of the National Nuclear Security Administration’s (NNSA) Commodity Technology Systems (CTS-2) program. It will provide additional production c...
Back to Newsroom