Please ensure Javascript is enabled for purposes of website accessibility

Why Ethernet for AI Fabric?

In this white paper, Utilizing Distributed Disaggregated Chassis (DDC) for Back-End AI Networking Fabric, we explore how the DDC-based solution, already proven as a robust and scalable solution in the telecom space, delivers the highest performance AI back end network.

In this white paper you will learn how DriveNets AI Fabric impacts performance:

  • 30% improvement in Job Completion Time (JCT)
  • Highest bisectional bandwidth
  • No vendor lock
  • 30% higher performance in AI training compared to ethernet Clos
  • Based on next-gen routers: Distributed Disaggregated Chassis (DDC) best serves AI workloads

Continue reading

Collateral

DriveNets AI Fabric

Full-stack networking portfolio for AI infrastructure. Cluster builders need a networking solution that delivers high pe ...

Read more

Collateral

AMD and DriveNets Reference Architecture

Building high-performance AI clusters is no longer a matter of adding GPUs.

Read more

Blog

The Hidden Mismatch Inside Every AI Cluster

AI infrastructure has, for most of its history, run on a one-size-fits-all assumption: pick a flagship GPU, build a flee ...

Read more