INSIGHT

Why GPUs wait - Rethinking data infrastructure for the AI era

Name: Why GPUs wait - Rethinking data infrastructure for the AI era
Uploaded: 2026-10-14
Description: Explore data infrastructure strategies that keep large AI environments productive with throughput, scalability, and resilience for training, inference, and multi-tenant services while maximizing GPU utilization.

Heute

Explore data infrastructure strategies that keep large AI environments productive with throughput, scalability, and resilience for training, inference, and multi-tenant services while maximizing GPU utilization.

Sprecher

Devon Helms

Devon Helms is a thirty year data management industry veteran and survivor of multiple startups. He gets energized thinking and writing about data and all its uses, sustainability, and increasing health span. A California native, Devon grew up surfing the swells of Redondo Beach as well as the early internet through bulletin board systems. But his main passion is rock climbing, a skill he honed on the razor sharp crags of Joshua Tree, CA. Devon now splits his time between his home office in the kitchen and his home office in the dining room because he loves to travel.

Mallikarjun Chadalapaka

Group Product Manager, Google

Mallikarjun is the Group Product Manager at Google Cloud. His team of PMs is responsible for all storage cloud services that ship on Google Distributed Cloud. In this role, Mallikarjun & his team have closely partnered with NetApp product and engineering teams. Mallikarjun's prior roles at other high-tech companies covered a gamut of products spanning AI infrastructure services, enterprise storage systems, distributed systems, Kubernetes, a variety of storage protocols, and management software platforms & technologies.

Related resources

Session presentation