Network Automation Nerds

NAN107: How AI is Changing the Networking Landscape

7 snips
Nov 26, 2025
Chris Kane, a seasoned network engineer and community leader with a background in AI infrastructure, joins the discussion. He emphasizes the rapid evolution of networking driven by AI technology. Chris advocates for network engineers to grasp adjacent technologies like compute and storage. He identifies the shift from 'selfish' networking to essential collaboration across disciplines. The conversation highlights the need for robust observability tools, power constraints, and the importance of intellectual curiosity in navigating this dynamic landscape.
Ask episode
AI Snips
Chapters
Books
Transcript
Episode notes
ADVICE

Learn Adjacent Domains

  • Do learn adjacent domains (compute, storage, power, facilities) before joining AI cluster projects.
  • Talk to compute and storage teams to make informed networking decisions.
INSIGHT

AI Clusters Force Overprovisioning

  • AI clusters force Ethernet and data center designs to "level up" in speed and reliability.
  • Over-provisioning and reduced oversubscription become necessary because GPUs must remain busy.
ADVICE

Audit Tools For RDMA Traffic

  • Do audit your tools for new traffic types like RDMA and RoCEv2 before deployment.
  • Ensure packet-capture and NMS tools understand those protocols.
Get the Snipd Podcast app to discover more snips from this episode
Get the app