SPOTLIGHT: Data Centers Need to be Multi-Tenant presented by Netris
About this episode
The episode discusses the necessity of multi-tenancy in data centers due to the rise of AI workloads and GPU clusters.
As the explosive rise of AI workloads brings massive GPU clusters back to on-premises environments and surging neo-clouds, the race to maximize hardware utilization has made multi-tenancy an absolute necessity. In this episode of the Tech Field Day Spotlight Podcast, host Tom Hollingsworth sits down with networking experts Scott Robohn, Ed Weadon, and Netris CEO and Cofounder Alex Saroyan to dissect the hidden complexities of modern AI infrastructure. Managing thousands of GPUs means navigating dense networking architectures like quad-planes, massive power restrictions, and the intricate blending of Ethernet and InfiniBand environments. Alex reveals how Netris solves these scaling bottlenecks from Day 0 using digital twin simulations and automated network abstraction, allowing operators to dynamically isolate cluster resources for different tenants through simple, unified API calls. Netris Representative: Alex Saroyan , Cofounder and CEO of Netris Scott Robohn , CEO of Solutional Ed Weadon , Network Engineer Follow the Tech Field Day Podcast on X/Twitter or on Bluesky and…
Topics covered
- AI workloads
- multi-tenancy
- networking architecture
- GPU clusters
- data centers
- digital twin simulations
More episodes of Tech Field Day Podcast
- 117. Multi-Terabyte GPUs the Only Answer for AI Scaling
- SPOTLIGHT: Networks Need to Drive Themselves presented by HPE
- SPOTLIGHT: Enterprise Networks Are Embracing Open Options Presented by Cisco
- 113. Optimizing Existing Hardware with Smart Software is the Only Way to Survive the AI Boom
- 112. Trillion Dollar AI Hype is Just Tech Groundhog Day
Explore listener stats, chart rankings, contacts and more on the Tech Field Day Podcast podcast page.