Cornelis details AMD-based AI network architecture

Key Takeaways

• Cornelis has introduced a new reference architecture that combines its CN6000 SuperNIC with AMD's 6th Gen EPYC processors and Instinct MI400 series GPUs
• The architecture is designed for high-performance AI inference, large-scale training, and HPC simulation
• It provides a blueprint for building integrated compute and networking platforms
• The CN6000 SuperNIC supports multiple transport protocols and offers 800 Gbps of bandwidth through a native PCIe 6.0 x16 host interface

Introduction to Cornelis' New Reference Architecture

Cornelis has unveiled a new reference architecture that pairs its CN6000 SuperNIC with AMD's 6th Gen EPYC processors and Instinct MI400 series GPUs. This architecture is designed to provide infrastructure operators with a blueprint for building high-performance systems on an integrated compute and networking platform. The combination of Cornelis' CN6000 SuperNIC and AMD's latest processors and GPUs enables fast and efficient processing of large amounts of data, making it ideal for applications such as AI inference, large-scale training, and HPC simulation.

Engineering High-Performance AI Networks

AI infrastructure requires a network that can keep pace with the compute power of the system. The new reference architecture from Cornelis is designed to meet this need, with the CN6000 SuperNIC supporting multiple transport protocols, including Omni-Path, RoCEv2, and Ultra Ethernet. This allows operators to standardize their network strategy without locking into one protocol. The CN6000 also offers 800 Gbps of bandwidth through a native PCIe 6.0 x16 host interface, matching the latest offerings from AMD.

Comparison of Key Components

Performance Capabilities

The Cornelis CN6000 architecture is designed to handle high volumes of data and provide fast processing times. It is capable of processing more than 1.6 billion bidirectional messages per second, keeping time to first token low and token throughput steady as concurrent requests scale. This makes it ideal for applications that require fast and efficient processing of large amounts of data.

Bottom Line

In summary, Cornelis' new reference architecture provides a blueprint for building high-performance systems on an integrated compute and networking platform. The combination of the CN6000 SuperNIC and AMD's latest processors and GPUs enables fast and efficient processing of large amounts of data, making it ideal for applications such as AI inference, large-scale training, and HPC simulation. With its high-performance capabilities and support for multiple transport protocols, the Cornelis CN6000 architecture is well-suited for a wide range of applications that require fast and efficient processing of large amounts of data.