Cornelis Partners with U.S. Department of Energy on the Genesis Mission

Building the Backbone of American AI Leadership

Front view of the United States capitol building

CN6000 800 Gbps Ethernet SuperNIC

An 800 Gbps networking adapter for AI and high performance computing

Cornelis and Synopsys Achievement

Best-in-Class Ansys Fluent Performance Powered by AMD

The World's First Lossless and Congestion-Free Scale-Out Network

Cornelis is solving one of the world’s biggest compute efficiency challenges - unlocking application performance with network-led acceleration at any scale. From faster AI training and ultra-responsive inference to the most predictable, high-throughput HPC simulations, Cornelis delivers results where legacy networks fall short. Built on the proven Omni-Path architecture, our solutions provide maximum performance, efficiency, scalability, resiliency and interoperability - empowering the next generation of AI and HPC infrastructure.

Introducing the CN5000 Omni-Path
End-to-End Network

400G SuperNIC

Complete 400G Switch Portfolio

48-Port Switch	Up to 576-Port Director Switch

Optimized Open-Source Host and Management Software

Delivering Industry-Leading HPC Performance

>2X Higher Message Rates
35% Lower Latency
30% Higher Application Performance

Revolutionizing AI Performance - 6X Faster AI Collectives Communication Performance

>30% Higher HPC Performance
>2X Higher Bandwidth on AI Sync Messages
500K+ Endpoint Scale

Talk to an expert

Message Rate, Latency, and HPC claims based on CN5000 relative to NDR 400G AI Collectives Performance
claim based on internal simulations vs 400G Ethernet

Why Cornelis Networks?

Highest Performance

Credit Based Flow Control, Dynamic Adaptive Routing, Enhanced Congestion Control
Reduced Checkpointing

AI & HPC at Scale

Near-linear Training Performance
Unlocks advanced inference with extended reasoning

Seamless Integration, Universal Interoperability

Universally interoperable with all leading GPUs, CPUs, and accelerators
BYOP (Bring your own platform)

Enhanced Resiliency

Ensuring Uptime Even During Failures
No Error-Driven Retries

Boost Datacenter Efficiency

No Packet Drops
Optimal Path Selection
Prevent Congestion Trees
Greater power availability to GPUs, sustaining workloads at peak efficiency

Making Your Network Work Smarter—Together

We have network experts ready to help you make the most of your infrastructure.
Plan your network, optimize your current infrastructure, ask questions, or get support. We are here to make it easy.

Accelerate Enterprise AI Applications

From training large language models and powering agentic AI frameworks, to real-time fraud detection in finance and computer vision in autonomous systems - CN5000 Omni-Path AI solutions deliver the high-performance, lossless networking foundation that next-generation AI demands.

Explore AI Solutions

HPC Networks, Redefined

Push the boundaries of scientific discovery and engineering innovation with CN5000 Omni-Path-based HPC solutions. From advancing climate modeling and seismic analysis, to accelerating simulation and design in aerospace and automotive, and enabling complex genomic research in life sciences, CN5000 Omni-Path solutions power breakthrough results—faster, smarter, and at scale.

Explore HPC Solutions

Experience Optimized AI Today!