The emergence of generative AI applications marks a new era for AI. The challenge, however, is that the characteristics of AI traffic are unique and quite distinct from the rest of the traffic carried by today’s data center network. The resulting data center build out is causing the need for a rethink of the data center network and strategies for deploying AI workloads in hybrid clouds. Once built predominantly in hyperscaler clouds, enterprises are now investing in on-prem, private AI data centers for local AI processing. This webinar will explore different network fabric options and network design approaches to meet these new on-premises and hybrid cloud AI driven requirements and discuss the size of the potential market opportunity for AI-tuned DC networks.
Resources:
https://www.juniper.net/us/en/solutions/data-center/ai-infrastructure.html
https://www.juniper.net/us/en/solutions/solution-briefs/2024/open-ethernet-based-ai-data-center-networking-solution-brief.html, https://blogs.juniper.net/en-us/ai-native-networking/the-most-flexible-way-to-deploy-and-manage-high-performing-networks-for-ai-workloads
