264001
eyJ0eXAiOiJKV1QiLCJhbGciOiJIUzI1NiJ9.eyJpc3MiOiJid2V2ZW50cy1wdWJsaWMtYXBpIiwiYXVkIjpbInB1YmxpYy13ZWIiXSwic3ViIjoiaHR0cHM6Ly93d3cuZ29vZ2xlY2xvdWRldmVudHMuY29tIiwiaWF0IjoxNzc1NDE1MzgyLCJuYmYiOjE3NzU0MTUzODIsImV4cCI6MTc3NTQxODk4MiwiY29udGV4dCI6eyJwYXlsb2FkIjp7ImlkIjoiIiwicmVnVHlwZU5hbWUiOiIiLCJldmVudF9pZCI6IjI2NDAwMSIsImNfMzI5MDQxNiI6IiJ9fX0.XKhmpf6azD_Qx_BWyhQedpN2NgWWn54sPnJqeNLzwTo
phase 5 session info modal
Day 2 – April 23, 2026
Breakouts
BRK2-115 • Kubernetes, Cloud Runtimes • Technical
Reve AI platform playbook: Scaling multimodal foundation models with Ray on GKE
location_on South Seas H
schedule 1:15 PM - 2:00 PM

Scaling AI workloads across heterogeneous compute clusters requires solving complex issues across the entire infrastructure stack. This session will detail how Reve manages the challenges of building state of the art diffusion models and scaling end to end development. Learn how to use Ray on GKE to orchestrate multi-region training, maximize hardware efficiency, and run unified reinforcement learning. You'll leave with an actionable blueprint for building highly capable AI platforms.

 Read more
This agenda widget contains logic for handling same sessions available at multiple times
GTM Tracking Header name present
Session
contains logic for gtm tracking when modal open