96 B200 GPUs. 12 nodes. A load balancer distributing requests across them.
https://medium.com/google-cloud/1-million-tokens-per-second-qwen-3-5-27b-on-gke-with-b200-gpus-161da5c1b592
DE DevOps&SRE Library @devopslibrary · 19.9K subscribers 96 B200 GPUs. 12 nodes. A load balancer distributing requests across them.